THE UNIVERSITY OF MICHIGAN INDUSTRY PROGRAM OF THE COLLEGE OF ENGINEERING LABORATORY PERFORMANCE STUDY OF COMMERCIALLY MANUFACTURED CONCRETE MASONRY UNITS MADE WITH LIGHTWEIGHT AGGREGATES A. Fattah Qasim.lhalabi I *.S I I I.... I I I.; r This dissertation was submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the University of Michigan. May, 1956 IP-157

ACKNOWLEDGEMENT We wish to express our appreciation to the author for permission to distribute this doctoral thesis under the Industry Program of the College of Engineering. ii

ACKNOWLEDGEMENTS The author acknowledges his indebtedness to Professor Legg for the encouragement and guidance so generously and graciously given throughout the experimental and research work of this project and for his careful checking and frank criticisms of the manuscript; to Professor Sherlock for his encouragement, genuine interest and valuable suggestions; to Professor Porter for his kind encouragement and for providing laboratory equipment and help for certain parts of the project. The co-operation of Professors Menefee and Legatski in reading and criticising the paper is deeply appreciated. The author also wishes to express his sincere appreciation to the Concrete Products Association of Michigan, especially to its executive secretary, Mr. C. A. Sirrine, for the samples, and to the loyal members of the association for their co-operation in furnishing their products for the test and the necessary field data. Thanks are also due to the Department of Mechanical Engineering and the State Highway Department for providing their laboratory facilities, to the Department of Civil Engineering for its financial support, and to the Industry Program of the College of Engineering for the publication of the dissertation. iii

ABSTRACT The tremendous growth in the use of concrete masonry units made with lightweight aggregates has made desirable the accumulation of data on their performances in order that intelligent decisions can be made regarding their proper use in building. This study attempts to fulfill this need in the Michigan area for a cross section of the products presently available. Evaluation of the performance of the lightweight products is made by comparison with that of dense products, all under the same condition of test. Materials were sampled from the regular production of ten commercial plants and included concrete units manufactured with several lightweight aggregates and three dense aggregates. The lightweight aggregates were: two cinders, one cinders mixed with fly ash, one cinders mixed with sand, one cinders mixed with expanded slag, three expanded slags including Waylite and Celocrete, one Waylite mixed with Beslite (expanded clay) and one expanded shale (Haydite). The dense aggregates were two natural sands and gravel and one aircooled slag, The tests conducted include: dimensional properties, unit weight, moisture content and absorption, compressive strength, flexural strength, elastic properties, freeze-thaw durability, drying shrinkage, thermal expansion and thermal conductivity. The tests were conducted on standard, hollow, 3-core, 8-inch block and on specimens cut from solid 4-inch slabs. Comparative studies were also iv

made of the 'physical properties of the aggregates themselves sampled at the plants involved. Methods and results of tests are given in detail and overall averages are presented. The more important conclusions are: (1) Adoption of the system of the modular dimensions by the industry is evident. (2) Weights of 8-inch thick masonry walls containing the lightweight units are 29 to 36 psf and for dense units, 43 to 48 psf. (3) Compressive strength of lightweight units varied between 58 and 88 per cent of the strength of dense units, (4) Lightweight aggregate and air-cooled slag products are more absorbent than sand and gravel products. (5) The moisture content of six of the products when sampled exceeded the 40 per cent maximum allowed by the ASTM specification. This requirement is considered of value in limiting shrinkage and subsequent cracking. (6) There was no sharp line of demarcation between the dense and lightweight aggregate products with regard to the total shrinkage. Autoclaving resulted in lower shrinkage. (7) Staining and popouts were shown by cinder products only. (8) Flexural strength of dense aggregate products are 1.6 to 3 times higher than those of lightweight aggregate. (9) Young's moduli for lightweight aggregate products range between 33 and 80 per cent of the lower value of the two sand and gravel aggregate products. v

(10) Completely saturated lightweight aggregate products of highly cellular structure, unlike sand and gravel products, are vulnerable to a limited number of accelerated freeze-thaw cycles. However, this condition of full saturation is rarely met in service. At moisture contents of 40 per cent (maximum allowed by ASTM specifications) lightweight aggregate products withstood 500 cycles of freezing and thawing well. (11) The coefficient of expansion of cinder products is lower than that of slag or sand and gravel products. (12) The coefficient of thermal conductivity of slag and cinder products is approximately 1/3 to 1/2 that of the sand and gravel products. Good correlation exists between this thermal coefficient and unit dry weight. It is concluded that lightweight aggregate masonry units possess characteristics that are desirable in construction materials for certain types of structures (residential buildings, etc.) such as their appealing surface texture, comparative lightness, which permits economy of construction, and lower thermal conductivity, which permits economy in heating and cooling. Methods have been devised and used with considerable success for reducing the shrinkage of certain lightweight aggregate products, such as autoclaving, etc. It is concluded that the continuing growth in the use of lightweight aggregate concrete masonry units for housing and similar types of construction is based on sound engineering judgment and is well justified. vi

TABLE OF CONTENTS Page ACKNOWLEDGEMENTS.......... ABSTRACT............. LIST OF ILLUSTRATIONS....... LIST OF TABLES........... INTRODUCTION............ 0 e & 0 0 e 0 a e 0 0 0 0 a ~ ~ ~ ~ I ~ e a ~ ~ o. * * * a 0 a * * * * * * 0 0 0 a e 6 9 6 0 0 0 0 iii iv xi xv 1 2 2 A. Aggregates.......... *... B. Concrete Units................... CHAPTER I. HISTORY.................... A. Lightweight Aggregate... B. Precast Building Units... C, Advancement in Technology D. Research on Lightweight Concrete, and Aggregates.......... 1. Precast Concrete Masonry.. 2. Plastic Concrete... f 0 0 0 9 0 ~..... 6 4 0 *... 8 Precast Masonry e ~ ~ I* ~ ~ ~ 14 15 22 ~ ~ * * ~ ~ (Made with Lightweight Aggregate) CHAPTER II. TEST SPECIMENS................. A. Sampling....................... B. Plant Operation................. CHAPTER III. AGGREGATES........... A. General Discussion of the Characteristics of Commercially Produced Lightweight Aggregates for Concrete................... 1. Cinders................. 2. Slags............, o 3. Expanded Clay, Shale and Slate...... B, Physical Properties of Aggregates o....... 1. Grading,................... 2. Unit Weight............ 3. Specific Gravity and Absorption...... 4. Deleterious Inclusions........... 27 27 28 34 34 34 38 41 44 48 49 56 vii

TABLE OF CONTENTS (Cont.) Page CHAPTER IV, DIMENSIONS OF THE MODULAR UNIT....... A. Definitions................... B. Procedure of Test. o. o o o.. o.. C. Discussion of Test Results o o. o.. o. o o o 62 62 63 64 CHAPTER V, UNIT WEIGHT, MOISTURE CONTENT AND ABSORPTIONo. o o 0 A. Procedure for Determining Unit Weight, Moisture Content and Absorption. o....... B. Discussion of Test Results. o o...... 68 68 1. Unit Weight................. 2. Moisture Contents and Absorption...... 68 71 CHAPTER VI, COMPRESSIVE STRENGTH. o o o o. o. o... 75 A. Procedure of Test.o o. o o.. o. o..o B. Discussion of Test Results.... o. o. 76 77 87 CHAPTER VIIo FLEXURAL STRENGTH........ A. Test Specimen............. 88 B. Apparatus.... o o. o.. o o o o... 88 Co Procedure of Test.. o....... o.. 91 D. Calculations........... o o o 91 E, Discussion of Test Results o. o o.. o.... 92 CHAPTER VIII', ELASTIC PROPERTIES....... 100 Ao Test Specimen....... o..... o... 102 Bo Dynamic (Sonic) Method.............. 103 1. Theoretical Discussion..... o. o o 103 20 Test Apparatus.............. o.. 110 3. Procedure of Test.............. 112 4. Calculations................. o. 116 Co Static Method.. o o............. o 122 1. Test Apparatus........... 2o Preliminary Investatgations........... 30 Procedure of Test. o......... 4. Calculations. o o. o o o. o...... 122 126 127 129 D, Discussion of the Results of Elastic Constants as Determined by Dynamic and Static Methods,... viii o. 132

TABLE OF CONTENTS (Cont.) Page CHAPTER IX. FREEZING AND THAWING DURABILITY.......... A. B. C. D. Specimen... 0.. 0 Apparatus............ Procedure of Test..... * * Discussion of Test Results....... * * * * *0 *0 e ~ ~ ~ ~I ~ ~ ~ ~ ~ ~ ~ ~ a ~ ~ ~ ~~~~ 138 141 141 142 145 CHAPTER X. STAINING AND POPOUTS OF CONCRETE MASONRY UNITS.. A. Procedure of Test........ B. Discussion of Test Results...* *. CHAPTER XI. DRYING SHRINKAGE............ A. Procedure of Test....... B. Discussion of Test Results........ CHAPTER XII. THERMAL EXPANSION.............. A. Methods of Test................... 154 155 156 16o 164 167 180 185 1. Extensometer Method.. 2. The SR-4 Gage Method. 0 0 r 0 0 0 0 4. 3 0 0 0 0 0 0 186 186 B. Procedure of Test.......... 190 1. Saturated Specimen.............. 190 2. Dry Specimen - Extensometer Method..... 190 3. Dry Specimen - SR-4 Gage Method....... 191 C. Discussion of Test Results.,.. *. 191 CHAPTER XIII. THERMAL CONDUCTIVITY............ 196 A. Brief Theoretical Analysis............. 204 B. Test Specimen................... 206 C, Test Apparatus............207 D. Procedure of Test...*......... 209 E. Calculations.... *....215 F. Discussion of Test Results........ 216 CHAPTER XIV. SUMMARY.................... 221 ix

TABLE OF CONTENTS (Cont.) Page APPENDIX I. AGGREGATES............ 229 A. Specific Gravity and Absorption...........229 1. Coarse Aggregate................229 2. Fine Aggregate.................230 B. Materials Causing Popouts of Concrete........231 APPENDIX II, DIMENSIONS OF THE MODULAR UNIT....... 236 APPENDIX III. UNIT WEIGHT, ABSORPTION AND MOISTURE CONTENT. 240 A. Unit Weight.................... 240 B. Absorption.................... 240 C. Moisture Content.................. 244 APPENDIX IV. COMPRESSIVE STRENGTH.........246 APPENDIX V. FLEXURAL STRENGTH............. 49 APPENDIX VI. ELASTIC PROPERTIES*....... o...... 251 A. Dynamic Method..................... 251 1. Identification of the Mode of Vibration.... 251 2. Calculations of the Elastic Constants.....252 B. Static Method.............. *. 252 1. Preliminary Investigations..........252 2. Calculations............... 252 APPENDIX VII. THERMAL EXPANSION.............. 301 APPENDIX VIII, THERMAL CONDUCTIVITY............. 306 A. Hot Plate Apparatus................ 30 B. Thermocouple Location................ 306 C. Discussion of Test Results.............308 LIST OF REFERENCES......................312 x

LIST OF ILLUSTRATIONS Figure Page 1 Cinder Aggregate Processing Equipment....... 9 2 Magnet Pulley for Separating Metallic Particles from Aggregate................9.. 9 3 Pipe Conveyor Transporting Aggregates to Over-Head Bins for Distribution to Batchers......... 10 4 Open Yard Storage of Products............ 10 5 Automatic Molding Machine.......... 12 6 Scale of Automatic Batcher............. 12 7 High-Pressure Steam Curing Kiln with 'Cap" Removed. 13 8 Low-Pressure Steam Curing Chamber......... 13 9 Saw Used for Cutting Specimens........... 30 10 Samples of Aggregates for Test.......... 35 11 Concrete Specimens After Staining and Popouts Test. 61 12 Set-Up for the Determination of Net Volume of the Unit........................ 65 13 Electrical Automatic Oven Used for Drying Masonry Units................... 69 14 Compressive Strength Test............ 78 15 Schematic Diagram of-Apparatus for Flexural Test. 89 16 Flexural Strength Test (General View)....... 90 17 Flexural Strength Test (Close-up)......... 90 18 Modulus of Rupture vs. Unit Dry Weight..... 96 19 Dynamic (Sonic) Test Apparatus for Determining the Elastic Constants.............. 0 0113 20 Schematic Diagram of the Sonic Apparatus...... 114 21 Curve for the Coefficient C1, Calculated by Pickett for j.= 1/6......11........ 117 xi

LIST OF ILLUSTRATIONS (Cont.) Figure Page 22 Static Test Apparatus for Determining the Elastic Constants (Close-up)................ 124 23 Static Test Apparatus (General View)........ 124 24 Schematic Diagram of Simplified Wheatstone Bridge Circuit............. o..... 125 25 Static Test Specimen with SR-4 Gages........ 130 26 Stress-Strain Curves for 1C52............ 131 27 Young's Modulus (Static) vs. Unit Dry Weight.... 137 28 Young's Modulus (Static) vs. Modulus of Rupture... 137 29 Schematic Diagram of Freezing and Thawing Apparatus. 143 30 Specimen Chamber of the Freezing and Thawing Apparatus..................... 143 31 Freezing and Thawing Durability of Test Specimens Before and After Copper Jacket is Added....... 144 32 Control Specimen for the Freezing and Thawing Test. 144 33 Drilling Hole in One End of Specimen for Cementing Steel Plug................... 166 34 Extensometer with Dial Micrometer Used for Length Measurements,................. 166 35 Moisture Loss - Drying Time Relation........ 171 36 Shrinkage - Drying Time Relation.......... 172 37 Shrinkage - Moisture Content Relation....... 177 38 Copper Bar with Compensating and Active SR-4 Gages for the Thermal Expansion Test........... 188 39 Apparatus for Thermal Expansion Test Using the SR-4 Gage Method (Close-up with Oven Open)........ 192 xii

LIST OF ILLUSTRATIONS (Cont.) Figure Page 40 Apparatus for Thermal Expansion Test Using the SR-4 Gage Method (General View with Oven Closed)..... 192 41 General View of the Hot Plate Apparatus...... 210 42 Close View of Measuring Instruments and Switch Box for the Hot Plate Apparatus......... 210 43 Control Panel for the Hot Plate Apparatus...... 211 44 Source of Electrical Current for the Hot Plate, Motor-Generator Set... o o 211 45 Looking Inside the Cabinet with Hot Plate, Cold Plate and Specimen in Between, Clamped Together... 214 46 Overall Average Results of Tests on Concrete Masonry Products...................... 227 47 Identification of Flexural and Torsional Modes of Vibration..................... 251 48 Stress-Strain Curves for 2CES5 - Preliminary Investigations................. 261 49 Stress-Strain Curves for 3SG5 - Preliminary Investigations....... o......... 263 50 Stress-Strain Curves for 9S5 - Preliminary Investigations.............. o... 265 51 Stress-Strain Curves for 1C51............ 267 52 Stress-Strain Curves for 2CES51........ 269 53 Stress-Strain Curve for 2CES52.......... 270 54 Stress-Strain Curves for 3C51............. 272 55 Stress-Strain Curves for 3C52............ 273 56 Stress-Strain Curve for 3SG51........... 275 57 Stress-Strain Curve for 3SG52......... 277 58 Stress-Strain Curves for 4SG51........... 279 xiii

LIST OF ILLUSTRATIONS (Cont.) Figure 59 6o 61 62 63 64 65 66 67 68 69 70 71 72 73 Stress-Strain Curve for 4SG52. Stress-Strain Curves for 5CS51 Stress-Strain Curve for 5CS52. Stress-Strain Curves for 6ES51 Stress-Strain Curve for 6ES52. Stress-Strain Curves for 7W51. Stress-Strain Curve for 7W52. Stress-Strain Curve for 8CEL51 Stress-Strain Curve for 8CEL52. Stress-Strain Curves for 9S51. Stress-Strain Curve for 9S52. Stress-Strain Curves for 10CA51o Stress-Strain Curve for 10CA52 Cut-Away Plan for Hot Plate. Page 281 283 284 286 287 289. 290 ~ 292 293 295 297 299. 300 * 307 0 * 0 0 0 a 0 0 4 * Location of Thermocouples for Two Companion Specimens.................. 309 xiv

LIST OF TABLES Table Page 1 Rate of Growth of Production of Concrete Masonry Units in the Years 1945-1947........... 7 2 Rate of Growth of Production of Concrete Masonry Units in the Last Decade.. o o.,..... 3 Summary of Field Reports.......... 0 31 4 Grading and Unit Weight of Aggregates....... 46 5 Unit Weight Requirements of Lightweight Aggregate for Concrete Masonry Units......... 47 6 Physical Properties of Aggregates........ 55 7 Dimensional Properties of the Modular Units (Average Values)..................... 66 8 Weights of the Masonry Units and the Unit Weights. 70 9 Moisture Content and Absorption..... 73 10 Compressive Strength of the Masonry Units.... 81 11 Physical Requirements.............. 85 12 Modulus of Rupture (R) Calculations........ 93 13 Flexural Strength of Concrete - Analysis of the Result......... 95 14 Calculations for Young's Modulus of Elasticity (Dynamic).....,........... 120 15 Calculations for Modulus of Elasticity in Shear, G, and Poisson's Ratio, p, (Dynamic),..... 121 16 Stress-Strain Data (Static) for 1C52 (Cinders).. 129 17 Elastic Constants (Static Test).......... 133 18 Average Results of Elastic Constant from the Static and Dynamic Tests............. 134 19 Comparative Study of Young's Modulus (Static) with Other Concrete Properties......,. 136 xv

LIST OF TABLES (Cont.) Table Page 20 Freezing and Thawing Durability - Initial Young's Modulus and Weight Measurement........ 147 21 Freezing and Thawing Durability - Effect of Test on Young's Modulus of Rupture......... 148 22 Results of Staining and Popouts of Test of Concrete Masonry Units......... o......... 158 23 Moisture Loss (% of Saturation) - Drying Time Relation.............. O.... 168 24 Shrinkage (% of Lengths) - Drying Time Relation.. 170 25 Shrinkage (% of Length) = Moisture Content (% of Saturation) Relations.............. 176 26 Total Shrinkage and Moisture Content at the End of Drying. Wet and Dry Weights of Specimen..... 178 27 Coefficient of Thermal Expansion (inch per inch per 1 F) x 106.................... 194 28 Calculation of the Coefficient of Thermal Conductivity (k) Btu per (hr)(sq. ft.)(deg. F./inch)... 217 29 Coefficient of Thermal Conductivity (k), Average Mean Temperature, and Unit Dry Weight...... 220 30 Summary of the Overall Average Results of Test on Concrete Products.... o....,.. 228 31 Calculation of Specific Gravity and Absorption of Coarse Aggregate.............. 31 32 Calculation of Specific Gravity and Absorption of Fine Aggregate................... 232 33 Calculation of Absorption (% of Volume)...... 233 34 Popouts Test - Laboratory Concrete Mixes,..... 235 35 Calculation of Net Volume of the Modular Units.. 237 36 Calculation of Gross Area and Gross Volume of the Modular Units................... 238 37 Net Area as Percentage of Gross Area........ 239 xvi

LIST OF TABLES (Cont.) Table Page 38 Weights of Masonry Units (Sampled, Wet and Dry Conditions).................... 241 39 Calculation of Weights of Masonry Units (Samnpled, Wet and Dry Conditions) in Pounds per Cubic Foot (Based on Net Volume)............... 242 40 Absorption of the Masonry Unit.......... 243 41 Calculation of Moisture Content of the Masonry Units as Sampled............... 245 42 Calculation of Gross Cross-Sectional Areas of Masonry Units Used for Compressive Strength Test...247 43 Calculation of Compressive Strength of Concrete Masonry Units Based on Gross Cross-Sectional Area. 248 44 Dimensional Analysis of Specimens for the Sonic Test-....................... 254 45 Weights and Radius of Gyration - Length of Specimens for the Sonic Test................ 255 46 Calculation of the Coefficient (C) for the Flexural Vibration.................... 256 47 Resonant Frequencies for Flexural and Torsional Vibration.............. 257 48 Calculation of the Shape Factor (R) for Torsional Vibration..................... 258 49 Calculation of the Coefficient (B) for Torsional Vibration............ 259 50 Preliminary Investigation - Stress-Strain Data (Static) for 2CES5 (Cinders and Expanded Slag).. 260 51 - Preliminary Investigation - Stress-Strain Data (Static) for 3SG5 (Sand and Gravel)........ 262 52 Preliminary Investigation - Stress-Strain Data (Static) for 9S5 (Air-Cooled Slag)......... 264 53 Stress-Strain Data (Static)for 1C51 (Cinders).. 266 xvii

LIST OF TABLES (Cont.) Table 54 55 56 57 Page Stress-Strain Data (Static) for 2CES51 (Cinders and Expanded Slag). *....... Stress-Strain Data (Static) for 2CES52 (Cinders and Expanded Slag).............. Stress-Strain Data (Static) for 3C51 (Cinders). Stress-Strain Data (Static) for 3C52 (Cinders).. 268.. 268.. 271.. 271 58 59 60 Stress-Strain Data (Static) for 3SG51 Gravel)............... Stress-Strain Data (Static) for 3SG52 Gravel)........ Stress-Strain Data (Static) for 4SG51 Gravel)............... Stress-Strain Data (Static) for 4SG52 Gravel)............... Stress-Strain Data (Static) for 5CS51 and Sand)......... Stress-Strain Data (Static) for 5CS52 and Sand)......... Stress-Strain Data (Static) for 6ES51 Slag)................ Stress-Strain Data (Static) for 6ES52 Slag)........ (Sand and (Sand and (Sand and (Sand and (Sand and (Cinders (Cinders (Expanded (Expanded * a * * * * 62 63 274 276 278 280 282 282 285 285 288 288 291 291 294 296 65 66 67 68 69 70 71 Stress-Strain Data (Static) for 7W51 (Waylite).. Stress-Strain Data (Static) for 7W52 (Waylite)... Stress^Strain Data (Static) for 8CEL51 (Celocrete). Stress-Strain Data (Static) for 8CEL52 (Celocrete). Stress-Strain Data (Static) for 9S51 (Air-Cooled Slag).................. Stress-Strain Data (Static) for 9S52 (Air-Cooled Slag)....................... xviii

LIST OF TABLES (cont, ) Table Page 72 Stress-Strain Data (Static) for 10CA51 (Cinders plus Fly Ash)............................. 298 73 Stress-Strain Data (Static) for 10CA52 (Cinders plus Fly Ash) O e... e..... o 0 o 298 74 Calculation of the Coefficient of Thermal Expansion of Water Saturated Specimen (inch/inch/deg. F) 106.......... 303 75 Calculation of the Coefficient of thermal Expansion of Dry Specimen —Test Method No. 1 (inch/inch/deg. F) 106..... 304 76 Calculation of the Coefficient of Thermal Expansion of Dry Specimen —Test Method No. 2 (inch/inch/deg. F) 106..,.. o. 305 77 Final Current and Temperature Determinations for Thermal Conductivity Tests (Two Runs per Product)......... 310 xix

INTRODUCTION This is a laboratory performance study of commercially manufactured precast concrete masonry units made with lightweight aggregate. Evaluation of the performance of these units is made by comparison with that of dense units (sand and gravel and air-cooled slag aggregate), all under the same conditions of testo Comparative studies were also made of the aggregates themselves sampled from the current production of the plants involved. This study will also offer comprehensive and useful data on the physical properties of these materials. The study is designed to provide a cross-section view of the products in the State of Michigan. For this purpose ten plants were chosen both for reason of their geographical location and the nature of their production. Four of these are located in Detroit and the rest are located in different representative areas in the state. Materials were sampled from the regular production of these plants, including concrete units manufactured with ten different lightweight aggregates, and three dense aggregates. The lightweight aggregates are as follows: two cinders, one cinder mixed with expanded slag, one cinder mixed with sand, one cinder mixed with fly ash, three expanded slags including the commercial names Waylite and Celocrete, one expanded slag mixed with expanded clay (Waylite and Beslite) and one expanded shale (Haydite)o The dense aggregates are two sand and gravel and one air-cooled blast-furnace slago Tests were performed on the most common size hollow blocks of the modular dimensions 7-5/8 x 7-5/8 x 15-5/8 inches with three 1

-2 - rectangular core openings, and on specimens cut from solid slabs of the modular dimensions 3-5/8 x 7-5/8 x 15-5/8 inch. The study concerns in detail the following physical properties of aggregates and concrete masonry units. A. Aggregates. 1. Grading. 2. Unit weight. 3~ Specific gravity (bulk oven-dry basis, bulk saturated surfacedry basic, and apparent), and absorption. 4. Deleterious inclusions. a) Materials finer than #200 sieve. b) Organic impurities. c) Materials causing popouts of concrete. B. Concrete Units. 1. Dimensional properties of the Modular Unit. 2. Unit weight, moisture content, and absorptiono 3. Compressive strength. 4. Flexural strength. 5. Elastic moduli (static and dynamic). 6. Resistance to accelerated cycles of freezing and thawing. 7. Staining and popouts of concrete masonry unitso 8. Drying shrinkage. 9. Thermal expansion. 10. Thermal conductivity. Results of tests, presented and analyzed in the following chapters, indicate that concrete masonry units having the advantage of light weightgenerally exhibited good performance when compared with denser units.

-3 -Each plant selected (except one) uses one or more lightweight aggregates in manufacturing their products. This is true for a great number of plants in the state. Indeed, the use of lightweight aggregate, because of its many advantages, has advanced so rapidly in recent years that it is now dominating the industry. This, coupled with the very large increase in the use of precast concrete units in construction, makes lightweight aggregate units highly important building materials.

CHAPTER I HISTORY The use of lightweight building material is known early in history. The Romans used natural lightweight stones, such as volcanic cinders, to build their great aqueducts to carry water to their cities. The use of hay to reinforce and reduce the weight of mud bricks was also known at that time. Until recently the use of lightweight building material was on a small scale. In this country, following the introduction of portland cement concrete, the emphasis was, for a long time, mainly on strength which dense aggregate concrete produced. It was not until recently by experience, research and modern technology that lightweight aggregate was fully recognized as an important building material.. A. Lightweight Aggregate. Lightweight aggregate for concrete was first used not strictly because of its physical characteristics. In the western states, because of economy, natural stones such as volcanic cinders were processed and used. In the eastern states, factory by-products such as industrial cinders and blast furnace slag came into use. None of the concrete so produced was accepted as of structural quality. The use of cinder concrete as a filler for roofs was common as early as the year 1896 in the city of New Yorkl* but was viewed with suspicion. This suspicion was greatly lessened by the research of Columbia University and the Underwriter's Laboratory's full scale * Reference 1. See list of references. 4

-5 - fire test. Blast-furnace slag was considered a waste product at the beginning of this century, accumulating in huge quantities which became a major problem for the steel industry. Trying to find some use for this material, the city of Detroit, for example, used large quantities for surfacing unpaved streets or as aggregate for bituminous concrete2. Not until the year 1930 was it fully recognized as a concrete aggregate. Stephen J. Hyde3, a chemist working in his laboratory in Kansas City in the year 1917, perfected a method of making lightweight aggregate by heating a certain shale, slate and clay to incipient fusion, and it was found that this method resulted in lightweight aggregate of good structural strength and insulating value. Two years later a huge source of lightweight natural stone commercially known as Zonolite was discovered in California and used, after processing, as concrete aggregate. Soon the possibility of good concrete from lightweight aggregate became apparent to the engineers and builders who started developing methods and techniques for manufacturing and using these new aggregates. Expanded clay and shale, processed diatomaceous earth, expanded slag, expanded slate, sintered, and natural lightweight aggregate came into use. Today it is estimated that more than 27 million cubic yards of this aggregate is manufactured annually5. This has a variety of uses. However, the bulk of the product is used for lightweight concrete and especially for precast units; - the latter is estimated to have used about 80 per cent of the total in the year 1954.

-6 - B. Precast Building Units. The precast concrete industry was the main consumer and promoter of lightweight aggregate. The use of this aggregate resulted in lighter units with varied surface textures, better thermal insulation and acoustical absorption and less thermal expansion. It also resulted in a decrease of compressive and flexural strength and higher drying shrinkage. Yet the strength of properly made precast units is usually adequate for the type of structure for which it is used. Drying shrinkage is still a problem, but methods of reducing it are currently being investigated. Precast concrete masonry units were originally frequently manufactured by the aggregate producers to dispose of an accumulation of certain sizes of aggregate. Concrete bricks were first manufactured followed by the two core block. Later the three core unit was adopted and became popular because of the ease of cutting the block symmetrically on the job. Concrete precast units were at first made with sand and gravel. Later, industrial cinders and blast-furnace slag were used. The light weight and the availability of these aggregates resulted in the growth of their use. By 1930, a number of specifications were written around these products. This gradually forced out the small manufacturers because of their inadequacy to meet the requirements of the specifications. This eliminated very low quality products from the market, permitting the growth of fewer but better established plants. This growth was continuous through the thirties and part of the forties but was greatly increased in the years following World War II.

-7 - This is apparent from Table 1. The unit price of the product did not change considerably during that time. This fast rate of increase in production following the year 1946 continued through the fifties as will be seen from Table 21. Another fact that the data does not reveal is the increase in the use of lightweight aggregate during that period of time. Available data estimated a consumption of about 20 million cubic yards of all aggregate in 1947, against 27 million cubic yards of lightweight aggregate alone in 1954. Approximately 80 per cent of the 1954 production of lightweight aggregate was used in manufacturing lightweight units. These units constitute about 54 per cent of all the precast units produced with the following percentage distribution: 43 per cent cinder 27 per cent expanded clay and shale 19 per cent expanded slag 7 per cent pumice 4 per cent scoria and other volcanic cinder. TABLE I. RATE OF GROWTH OF PRODUCTION OF C9NCRETE MASONRY UNITS IN THE YEARS 1934-1947.U Gross Selling Value in Year Number of Plants Millions of Dollars 1934 88 1.0 1936 97 1.7 1938 119 3.2 1940 128 5.3 1942 141 7.3 1944 149 6.6 1946 --- 140.0

-8 - TABLE 2:. RATE OF GROWTH OF PRODUCTION OF CONCRETE MASONRY UNITS IN THE LAST DECADE (Data obtained from different sources) Number of 8-inch Equivalent Units Year (in millions) 1944 340 1947 1000 1952 1.780 1953 1800 1954 1900 C. Advancement in Technology. Currently, the precast concrete masonry industry manufactures about 2 billion 8-inch equivalent units annually of different sizes and shapes, more than half of which are made with lightweight aggregate of many varieties. This growth was a result of better products, of advanced technology, and new methods and materials. The modern plant of today is far from the plant of two decades ago. It is larger, better equipped, highly efficient and easier to control by automation. The improvement, extending to all plant operations7, will be summarized as follows: (1) Automation of modern plants results in an economy and. facilitate quality control. Automatic screening of aggregate, batching, mixing, molding and curing (heat and humidity control) are becoming standard. (2) Today, manufacturers, especially those using cinder or slag, frequently have their own aggregate processing equipment. They also stockpile large quantities of aggregate in storage yards

-9 - Cir culat ion!I: -:;'-,-' Figure 1. Cinder Aggregate Processing Equipment. Figure 2. Magnet Pulley for Separating Metallic Particles from Aggregate.

-10 - Figure 3. Pipe Conveyor Transporting Aggregates to Over-Head Bins for Distribution to Batchers. Figure 4. Open Yard Storage of Products. Also notice two fork-lift trucks at left and metal racks at right.

-11 - and over-head binso Bulk cement is also stored in such bins and are all connected to the batcher. (3) Quality control and the use of lightweight aggregate have made the once standard continuous mixer obsolete. In spite of its high capacity and its occupying a relatively small area, this open long drum horizontal mixer has been discarded in favor of the modern fifty cubic foot mixer which is usually equipped with an automatic batcher. (4) Molding machines have undergone a great changeo Before 1940, the tamping machine was the standard and is now displaced by the vibrating machine. The latter is fully automatic. Receiving the concrete from the mixer (usually located above the molding machine) it molds by both vibration and pressure. These molding machines are also superior to the old tamper type by reason of speed, economy atnd capability of handling drier mixes without crushing the softer lightweight aggregates. In addition to its ability to manaufacture many sizes and shapes of units, only one operator is required. Eis dulty is primarily to remove the units, automatically ejected, from the front of the machine to a rack by means of an air hoist. (5) Moist curing, which was standard in earlier pla.nts, has been supplanted by two more efficient curing methods, hightemperature (normal pressure) steam and high-pressure steam curing. The tendency of the modern plant is to have many relatively small curing chambers which are more efficient and easier to co:tl..trol. High-pressure steam curing, although many consider it superior to other methods and certainly less time consuming, is not as common as expected, possibly because of the high cost of equipment.

-12 - Figure 5. Automatic Molding Machine. Figure 6. Scale of Automatic Batcher.

-13 - Figure 7. High-Pressure Steam Curing Kiln with "Cap" Removed. Figure 8. Low-Pressure Steam Curing Chamber.

Some modern plants are mechanically drying their products following the curing cycle. (6) Product handling equipment advancement was essential to greater plant efficiency. Prior to 1938, products were moved by hand or by rail cars. About that time the spring supported platform truck was developed which is capable of handling green product racks and finished products. In the year 1944 the fork-lift truck was introduced, which is more efficient in handling finished products and often moves and loads large cubes of the product without necessity of a pallet. (7) Better quality control has been established. Many plants have their own small laboratory for daily routine tests. D. Research on Lightweight Concrete_ Precast Masonry and Aggregates. The foregoing advancement in materials and technique of manufacture was made possible by research of the industry and public organizations. The history of the development of research in any industry is, usually, timed with the growth of that industry. Twenty-five years ago there was hardly any important research on lightweight aggregate and concrete. An early recommendation on precast masonry by the American Concrete Institute (A.oCoo), dated 1925, did. not recognize lightweight aggregate for 'concrete; precast concrete masonry (blocks) were classified with so called concrete brick and with tile. However, a few years later it started to gain recognition. The report of the Ao C. Io committee, reviewing the above mentioned recommendations, published in 1930 is interesting8.

Precast concrete masonry units were reported gaining importance as building materials with a growing industry of its own (not merely a device employed by the aggregate producers to dispose of an accumulation of a certain size of aggregate). Cinders and Haydite (burned shale) were recognized as concrete aggregate. Advancement of technique of manufacture of precast units, new methods in proportioning, mixing, curing, and molding were also reported and recommended. In the following years, as the use of such products was increasing, research began developing on both old and new materials. This gained momentum during the last decade when the huge expansion in the use of lightweight concrete occurred. Wide study of previous research on this subject would be out of place here; however, a brief presentation of the more important work is made. It would not be significant to study these researches in sequence of their time of conduct. They may better be divided into two classes: one dealing with machine molded concrete, the second dealing with plastic concrete usually poured in forms. 1. Precast Concrete Masonry. Although the use of industrial cinder as an aggregate for plastic concrete or concrete masonry was very common by 1930, there were very few scientific studies on the subject. The literature available at that time was not extensive. A. C. I. committee 203 was set up to bring the information on the subject up to date. Their comprehensive report, published in 1 1931, started by defining industrial cinders (anthracite and bituminous). Chemical composition and their effect on concrete

-16 - was also studied. Results of tests for physical properties of aggregate were presented and discussed. The making of masonry units was also presented. Specifications for aggregate, concrete and its products were offered along with the procedure of the necessary formal testing. This work helped a great deal in removing doubts about these materials. One of the earliest researches on shrinkage of concrete masonry was published by W. D. Allen9, in 1930, aimed at determining the rate and extent of shrinkage occuring in concrete walls as affected by the moisture content of the unit at the time the wall was layed. Twenty-four wall panels, each one block wide and eight blocks high, were constructed of units made with sand and gravel, Haydite or cinder aggregate and were two, seven or ninety days old when tested. Walls were saturated for fortyeight hours, then dried at a temperature varying from 69 to 80 F and relative humidity of 32 to 77 per cent. It was concluded that the use of air dry masonry units will reduce shrinkage 20 to 35 per cent. Recommendations were also offered to reduce drying ahrinkage to a minimum. An interesting early research on the structural performance of concrete masonry walls as related to the strength of the unit was published by the University of Illinois1 in 1931. This relation is important since almost all the building codes put provision on the required strength of the unit. In this work, the walls were divided into two general classes, load-bearing and non-bearing. They were of hollow blocks of four sizes, with two or three, oval or rectangular, cores.

-17 - Seventy walls, 6 feet long and 9 feet 6 inches high and fortyfour wallettes, two feet long and four feet high, all one block thick were constructed. These walls were built with units made of different aggregates, consisting of sand and gravel, Haydite, crushed limestone, cinders, and crushed air cooled slag. All units were commercially manufactured and moist cured for twentyone days plus seven days storage in air. Walls built with two kinds of mortar, 1:1:41/2 cement lime mortar and 1:3 cement mortar containing ten per cent lime, were then stored at 70-75 F and 50-65 per cent relative humidity (RoH4) for 32 days before testing. They were tested to failure in compression and crossbending. Loading was uniform or concentrated, axial or eccentric. Lateral and vertical deformations were measured by dial strain gages. Among the findings were: (1) The compressive strength of a wall panel depends on the compressive strength of the unit with an average ratio of approximately 0.53. But the flexural strength depends on adhesion of mortar to the units with failure following along a horizontal mortar joint. (2) Walls, eccentrically loaded at the edge of the middle third deformed uniformly and developed a strength equal to approximately 3/4 that of axially loaded walls. (3) As expected, the deformations and working stresses of concrete masonry walls were less than that of reinforced concrete walls. (4) There was a constant ratio between results obtained for walls and wallettes of the same materials and,

-18 - therefore, wallettes could be used for testing instead of the larger and more expensive assemblies. An important research that promoted concrete masonry units as building material was the full scale fire endurance test conducted at the laboratory of the Portland Cement Association and published in the year 1932. The load-carrying capacity of concrete masonry walls, made of units of widely varying characteristics, both during and after exposure to standard fire test conditions were investigated11 Comparative data on similar walls not exposed to fire were included. These investigations consisted of tests on more than 200 walls 5-1/2 ft. wide, 6 ft. high and 4, 8, and 12 in. thick. All the blocks were manufactured in the laboratory, using a block machine (tamper) of standard make, with nine different aggregates including Haydite, air-cooled slag, cinder and coke breeze. Among the principle findings were' (1) The comparative strength of walls both before and after exposure to fire was directly proportional to the strength of the unit. (2) The strength of mortar, type of joints and characteristics of mortar bedding had a marked effect on strength both before and after the test. (3) Substantial load-carrying capacity and safety were exhibited by walls during and after fire exposure. The University of Wisconsin test (sponsored by the National Concrete Masonry Association) published in 1939, is held by many as being responsible for speeding the decision of

-19 - the industry to shift from tamping to molding by vibration. It was a comparative study of the physical properties of masonry units which were manufactured by the two methods from the same mixes12. A single grading of each of the seven following aggregates were used: cinders, Haydite, limestone, Pottsco (expanded slag), sand and gravel, Superrock (expanded slag), and Waylite. The cement was a blend of four brands. Six different batches were used for each aggregate. From each batch, 30 half units and 30 full units 8 x 8 x 16 inches were made. These units were tested for compressive strength, absorption, capillarity, specific weight, durability, volume change and thermal expansion coefficient. The findings were numerous and mostly in favor of molding by vibration. Staining and popouts of concrete, especially in cinder masonry units, are common evils of bad quality aggregate. It has received much attention because of the increase in the practice of painting exterior walls. The Portland Cement Association, with the cooperation of a group of cinder block producers, conducted a comprehensive study of the cause and prevention of staining and popouts in cinder concrete. This report, published in 194813, divided popouts into two classes according to whether or not it is accompanied by staining. It also found that staining, whether or not accompanied by popouts, is usually caused by one or more compounds capable of decomposition in the presence of air and moisture, while free lime (or partially hydrated lime) is the main cause of popouts that are not accompanied by staining. Free magnesia and calcium sulphate are considered responsible

-20 - for very few cases of this type of popouts. The results are based on the chemical and microscopical analysis of several hundred cases of stains and popouts and analysis of aggregates collected from forty different sources. Procedure of determining the amount of this deleterious material is based on physical and chemical tests. Methods of treatment were offered for such impurities, ranging from simple magnetic separation (which is usually used for all kinds of cinder aggregate) to a steam or lime treatment for cinders containing many such impurities. Attention was always focused on the important problem of drying shrinkage of concrete masonry units. Two important researches dealing with the subject were published recently. The first is a progress report of A. C, Io committee 716, working on the physical properties of high-pressure steam 14 -cured concrete blocks, focusing the study on drying shrinkage of blocks. The research was conducted by four laboratories using three different procedures, called the "Reference Method", "British Method" and "Rapid Method". Shrinkage was measured by either the Whittemore strain gage, the "Ames" dial comparator or both. In the Reference Method, the blocks were immersed in water for 96 hours at 73 - 3 F; then dried in cabinets at 73 - 3 F over a saturated solution of sodium hydroxide (equivalent to 10 per cent R. H.). in the British Method the blocks were immersed in water for 96 hours at 73 - 3 F; were then dried in cabinets over a saturated solution of calcium chloride (equivalent to 17 per cent R. H.) at 122 - 3 F. In the Rapid Method the blocks were immersed in water for 24 hours and then dried at 220 to 230 F in the oven.

-21 - Tests were conducted on blocks made in four commercia] plants with sand and gravel, cinder and expanded clay. Silica dust was also used with some of the mixes. Some of the main findings were: (1) Twenty-four hours of immersion is adequate for total saturation. (2) The quantity of the moisture lost was independent of the method of drying. (3) The British Method and the Reference Method gave approximately the same total shrinkage; the Rapid Method gave approximately twice the value obtained by either of the other methods. L t It was also suggested that if the foregoing provision (3, above) is made, the Rapid Method might be convenient to adopt. The second research at the University of Toledo was aimed at the relation of shrinkage to moisture content in concrete masonryl5. This was a broad program, ranging in scope from studies in the chemical make-up of the cementitious phase to dealing with the shrinkage cracking of restrained wall panels under controlled condition of moisture and humidity. The blocks, all 8 x 8 x 16 inches were manufactured under commercial conditions with sand and gravel, cinders, expanded shale, expanded slag, sintered shale and pumice. Curing was done by five methods: "moist-air curing", "low-temperature steam" (120 F), "hightemperature steam" (170 F) plus drying at 150 F and "high-pressure steam." The finished blocks were saturated under water for forty-eight hours; then stored in air at 25 ~ 3 per cent R.H.

and 73 ~ 3 F temperature with air circulation at 5-10 mph. Shrinkage measurements were taken until equilibrium was reached. The specimens were then resaturated for forty-eight hours and dried in air at 70 per cent R. H, and 73 ~ 3 F until their shrinkage was again stabilized; then drying was continued at 25 per cent R. H. and 73 + 3 F until the final shrinkage was also stabilized. Among the main findings were: (1) High-pressure steam curing, except for pumice aggregate, reduces the drying shrinkage of the product by approximately one-half that which resulted when other methods of curing were used. (2) A forty per cent moisture content ( the usual specification requirement) is not accompanied by a marked reduction in shrinkage. Considering all factors, only about one-fifth to one-half the potential shrinkage has developed at this moisture content; the balance of shrinkage taking place when the shrinkage reached equilibrium at 25 per cent R. H. 2. Plastic Concrete (Made with Lightweight Aggregate). Since cinders and Haydite were among the first used as lightweight aggregate, early studies of their properties are found. One of these, on cinders, was previously presented on page 9. Another study conducted at the University of Illinois in 1930, was on the physical properties of Haydite aggregate and concrete (plain and reinforced)16. Two types of Haydite concretes were considered, one with sand as fine aggregate, and the second with "all-Haydite" concrete. The construction and design features were also studied for the two types of concrete,

-23 - providing data and recommendations to facilitate the design and construction of this type. Waylite was the subject of the same type of study at the University of Wisconsin7, sponsored by the Waylite Company, Chicago, Illinois. The physical properties of Waylite aggregates and concrete (plain and reinforced) were investigated. Mix design and methods of mixing were also developedo Data on weight, strength, absorption, stiffness, and damage by freezing and thawing were also presented. Limited study was also done on reinforced Waylite concrete. Burned shale and expanded slag aggregate in concrete were studied at the National Bureau of Standards. The physical properties of three grades of concrete from each aggregate were tested8. It was found that if the cement content was low (3 sacks per cubic yard) workability was very poor. This was corrected by entrained-air, up to 25 per cent. For reinforced concrete, bond and flexural stresses were also studied. A comprehensive research on lightweight plastic concrete was sponsored by the Housing and Home Finance Agency and conducted, separately, by both the National Bureau of Standards and the Bureau of Reclamation. The first mentioned, investigated aggregates available in the eastern part of the United States, and the second, aggregates available in the west. The Bureau of Standards worked on 11 lightweight aggregate samples including exfoliated vermiculite, expanded perlite, three samples of expanded blast-furnace slag, sintered fly ash, pumice, 19, 20 expanded slate and clay. The Bureau of Reclamation

worked on 19 samples obtained from 17 different producers, including two samples of expanded shale, one scoria, three expanded slag, five pumice, five expanded perlite, two exfoliated vermiculite, and one diatomaceous earth19' Both laboratories included sand and gravel aggregate in their tests for comparison with the lightweight aggregate. Both studied the physical properties of aggregate and concrete. In the making of concrete, the Bureau of Standards used the aggregate in the same grading as furnished by the producers, adding Vinsol resin, an air-entraining agent, to the mix to lend workability to the concrete while the Bureau of Reclamation tried to make workable concrete without the use of an air-entraining agent, when aggregates were of suitable grading, or by altering the grading, such as crushing or regrading; however, the addition of air-entraining agents was a necessary step for a number of aggregates. Four grades of concrete per aggregate, approximating a cement content of three, five, seven, and nine sacks per cubic yard of concrete, were tested by both laboratories. Concrete was mixed by a drum mixer, molded by hand and cured by fog. Both researches resulted in useful data on the following physical properties of aggregate and concrete: (1) On aggregate! sieve analysis, unit weight, bulk specific gravity, absorption, and crushing strength. (2) On concrete~ unit weight, absorption, compressive strength, modulus of elasticity, modulus of rupture, thermal conductivity, resistance to freezing and thawing, drying shrinkage, sawability, and nailability.

-25 - Both laboratories analyzed their data, drawing conclusions and offering recommendations: (1) The use of air-entraining improves workability to a large degree, especially in leaner mixes. (2) Drying shrinkage is the most unfavorable physical property of lightweight aggregate concrete, because of higher absorption of aggregate and because higher cement conten-t for a given strength are required. (3) Thermal conductivity of lightweight aggregate concrete is much lower than that of dense concrete. (4) Certain lightweight aggregate concretes, containing suitable amounts of entrained air, have a high resistance to freezing and thawing, despite the fact that their densities and strength are relatively low. Much early work was conducted on plastic concrete made with lightweight aggregate. In these early studies, concrete mixes were designed to suit the purpose of the study, i.e., concrete manufactured by laboratory methods and controls and poured in forms. The result of such investigations will not necessarily represent commercially produced concrete masonry. Much research was also done on concrete masonry, some described in the foregoing pages, conducted on specimens which are either manufactured in the laboratory, using block making machines, or are obtained from commercial plants, which, in some cases manufactured the units especially for the tests. Moreover, the majority of this research is aimed at studying

-26 - one property of closely connected properties, such as drying shrinkage and absorption, or a comparative study of a certain method or technique of manufacture. Therefore, a comprehensive, comparative study of the physical properties of commercially produced concrete masonry units made with lightweight aggregates and sampled from the regular production is desirable. After all, these are the products actually being used in building today. In a study of this nature, it will be practical to conduct research on products selected to represent a certain locality. The present work, as said before, was designed to give a cross-section study of the products in the State of Michigan.

CHAPTER II TEST SPECIMENS Discussion will now be made of procuring the test specimens. Each of the ten participating plants was visited to obtain samples and to prepare a field report on materials and methods of manufacture. These field reports are summarized in Table 3, which will be discussed later. A. Sampling. Concrete units were sampled from the lots of the regular products available at the plant at the time. The procedure of sampling was according to the requirements of ASTM Designation C140-52 (Methods of Sampling and Testing Concrete Masonry Units)22. Aggregate samples were also obtained from the processed materials at the plant. The procedure followed in sampling the aggregate was according to the requirements of ASTM Designation D75-48 (Methods of Sampling Stone, 22 Slag, Gravel, Sand, and Stone Block for Use as Highway Materials). A number was assigned to each producer participating in this study. Starting at number one, each number represents a single producer. After each number a letter(s) has been used to represent the aggregate(s). This procedure has been followed throughout this work (see Table 3,: column 1). Plants #2, #3, and #7 each submitted two types of products (made with two different aggregates) while the remaining plants submitted one type each, making a total of thirteen different products submitted by the participating plants. 27

-28 - (1) Twenty-five hollow three-core masonry units (blocks) of the modular dimension 7-5/8 x 7-5/8 x 15-5/8 inch were obtained. (2) Also from each product except 2H and 7BW (see Table 3, column 1 ), ten solid masonry units (slabs), of the modular dimension 3-5/8 x 7-5/8 x 15-5/8 inch, were obtained. Slabs were not available from 2H and 7BWo (3) From all the plants, except #2 and #9, aggregate samples were obtained. Aggregates were not available from these two plants at the time of sampling. Upon arrival at the laboratory, the samples were marked for identification. Five specimens were selected from the hollow masonry units for immediate testing (ASTM Designation C140-52; see also Chapter V); the remainder of the specimens and the concrete aggregate samples were stored in the laboratory under normal temperatures and humidity conditions. Full size hollow blocks were used for testing whenever possible, otherwise specimens of desirable sizes were cut from the slabs. Since products 2H and 7BW were not available in slabs, it was not possible to conduct tests on these in which such specimens were needed. Details of each specimen used for each test will be given when discussing that test. B. Plant Operation. Plant operation is meant to include materials, methods and equipment used in making the products, mix data, methods of mixing, molding, curing, and storage. Plant operations were studied and data obtained from the producers to prepare the field report. These operations, as expected,

-29 - differed from one plant to another, but in general, all were found to have certain features of the modern plant described in Chapter I. The field reports are summarized in Table 113.. Column 1 of this table gives the identification number to aid interpretation of the letter symbols. The fine and the coarse aggregate from which the products were made are listed in columns 2 and 3. Column 4 indicates types of cement. All plants except #1 and #3 use high early cement (plant #3 uses high pressure steam curing). Column 5 covers admixtures and indicates that only plants #1, #5, and #7 use added air-entraining agents (plant #1 also uses a wetting agent); plant #10 adds fly ash to its mix. Mix data are given in columns 6 to 11. Methods of measuring the mix are given in column 6 from which it is seen that all but plants #1 and #4 measure by weight. Plants #1 and #4 measure by volume but, as will be seen later, they also use a different method of mixing. The mixing water reported in column 7, variable in most cases, does not permit precise calculation of the water cement ratio. This is because the moisture content of the aggregate, which stands in large piles in open yards, is variable, and the consistency of the mix is actually measured by that required for molding by pressure and vibration and permitting subsequent stripping of the unit without physical damage. Columns 8, 9 and 10 give the cement, fine and coarse aggregates used in making one batch. It is to be noticed that the maximum size of the coarse aggregate is usually 1/2 inch; some plants use combined fine and coarse aggregate due to the small top size. Column 11 gives the amount of admixture, if used, and is normally added to the mixing water. Methods of mixing are summarized in column 13 from which it is observed that except for

-30 - Figure 9. Saw Used for Cutting Specimens.

TABLE 6. PHYSICAL PROPERTIES OF AGGREGATES Material Specilic Gravity Absorption Passing B.S.G B.S.G. of 0 Organic Type (#200 Sieve Oven-Dry Saturated Dry by Volume Impurities Sample of % by Basis Surface A.S.G. Weight (Dry Colorimetric Popouts No. Aggregate Weight) Dry Basis Compacted) Test 1C9 Cinder Combined 3C9 Cinder 'Combined 3SG9 Sand and Gravel Combined 4SG9 Sand and Gravel Combined 5C9 Cinder Combined 6ES19 Expanded Slag Coarse 6ES29 Expanded Slag Fine 7W19 Waylite Coarse 7W29 Waylite Fine 7B9 Beslite Combined 8CELl9 Celocrete Coarse 8CEL29 Celocrete Fine 10CA9 Cinder Combined 7.3 5.6 2.23 3.15 3.67 0.14 0.59 0.14 6.72 5.70 0.93 8.73 1.71 1.85 1.99 8.40 7.87 No Change in Color 1.74 1.90 2.06 9.00 8.26 No Change in Color 2.63 2.66 2.73 1.50 2.50 Slight Change in Color 2.66 2.69 2.75 1.30 2.47 Very Slight Change in Color 1.68 1.82 1.95 8.20 7.73 No Change in Color 1.47 1.57 1.64 6.75 5.16 No Change in Color 1.78 1.87 1.96 5.11 5.33 No Change in Color 1.48 1.59 1.67 7.57 5.81 No Change in Color 1.80 1.89 1.98 5.12 4.66 No Change in Color 1.50 1.71 1.89 14.13 15.2 No Change in Color 0.97 1.21 1.27 24.3 11.05 No Change in Color 1.49 1.71 1.90 14.41 10.68 No Change ih Color 1.71 1.85 1.99 8.41 8.13 No Change in Color None None None None None None None None None None None None 10.7 None

-32 - plants #1 and #4, batch mixers are used, usually of fifty foot capacity and with an automatic control for batching by weight. Mixing by this method is accurate and efficient, the mixer is located under the cement and aggregate bins and over the molding machine. Some are capable of traveling on rails to serve more than one molding machine. Plants #1 and #4 use continuous mixers and thus proportion their mixes by volume; the horizontal mixer has a capacity of about ten cubic feet. All plants, except #1 and #4, use approximately the same procedure of mixing, although they differ as to mixing time; the lightweight aggregate is first mixed with most of the water, then the cement and the rest of the water added and mixing continued. This mixing sequence is considered to add to the workability and strength of lightweight aggregate concrete. By mixing first with water the surface cells of the aggregate are filled, preventing cement paste from occupying these spaces and thus keeping the water available for workability and cement hydration. Plants #1 and #4 use the procedure of mixing cement and aggregate dry, then adding the water; the same method is used by plant #3 for sand and gravel products. Molding methods are summarized in column 14 from which it can be seen that all the plants use the vibration method, which is considered advantageous (see Chapter I). Most of the plants use a rather complex, fully automatic machine which molds by vibration and pressure. Usually located under the mixer, the molding machine receives concrete through a hopper and ejects the molded products to the front automatically. It is capable of molding many shapes and sizes of masonry at a speed of about 900 blocks, 8-inch equivalent, per hour. Curing is summarized in column 15 which indicates that all the

-33 - plants use low pressure steam (plant #3 uses, in addition, highpressure steam for its cinder products). The procedures are similar although the cycles are different, depending on the size of the curing chamber and the equipment used by the plant. The green products on the racks are moved to the curing chamber until it is full, which usually allows the product more than an hour of setting. With the chamber closed, live steam is injected until the temperature reaches 170-180 F. The concrete is "soaked" for several hours before opening the chamber. Soaking means the storage of the product in the steam chamber after the steam is turned off, thus permitting gradual cooling in the atmosphere of residual moisture. The length of the cycles given in the table varies from 12 to 24 hours. Plant #3 uses high-pressure steam curing for cinder products, and employs a cycle of 5 hours. The product is moved from the curing chambers to the storage yards where it normally remains from one to three weeks before it is distributed to the comsumer. From column 16 it is apparent that most of the plants use open storage yards; some cover their product during the winter.

CHAPTER III AGGREGATES Properties of the aggregates used by the various plants for producing precast concrete units are covered in this chapter. The study consists of a general discussion of the characteristics of commercially produced lightweight aggregate and discussion of the physical properties of the specific aggregates in this program. Tests were conducted on samples obtained from the production lines of the plants concernedo The samples obtained and the procedure of sampling and marking were discussed in Chapter II. Upon arrival at the laboratory, these samples were marked and left to dry at room temperature until tested. Ao General Discussion of the Characteristics of Commercially Produced Lightweight Aggregates for Concrete. Lightweight concrete products in this study are made of three general types of aggregate~ cinders, expanded slag, and expanded clay and shale. The general characteristics of these lightweight aggregates and the common methods of their production will be discussed. 1. Cinders. Cinders suitable for use as concrete aggregate are the inorganic residue from the high temperature combustion of coal or coke and are sometimes called "industrial cinders" as distinguished from the product of small domestic furnaces. The latter product is not usually considered desirable as concrete aggregate. In this study, the word "cinder" will be used to mean the so-called "industrial 34

-36 - 7W19 Waylite, Coarse 7W29 Waylite, Fine 8CEL19 Celocrete, Coarse 8CEL29 Celocrete, Fine 7B9 Beslite, Combined 10CA9 Cinders, Combined Figure 10. Samples of Aggregates, Continued

cinder" as defined above. Two types of cinders are produced by burning two coals: anthracite and bituminouso These are somewhat different in their properties but both have proven suitable, under favorable conditions, for use as a concrete aggregate. Physically, cinder particles are porous with an angular shape and uneven faces. Their physical characteristics depend on both the type of coal and the temperature of burning. Chemically, cinders are composed of silica, alumina, lime and iron with small quantities of magnesium, titanium and alkali compounds. Small amounts of sulphur are generally present in the form of sulphides and sulphates. Carbon from unburned coal or coke is practically always present in cinders due to incomplete combusion. When used as concrete aggregates, cinders must be processed to a specified grading by crshiing arnd screening. They must be cleaned of deleterious materials. Deleterious materials in cinders are excessive amounts of carbonaceous materials, sulphur compounds, iron compounds, and free lime. (1) Carbonaceous materials: These are present as a result of incomplete combustion. 3ituminous cinders contain carbonaceous materials in the form of coke, wnhile anth-racite cinders may contain unburned coal. No relationship has been established between the carbon content of cinder aggregate and the strength of concrete, probably because of the difficulty, in comparative tests, of controlling the consistency of the mix when varying the percentage of dense coal and porous cokeo However, the Underwriter's laboratories found by tests that the combustible content of cinder aggregate should not exceed 35 per cent by weight of dry mixed aggregate.

-38 - (2) Sulphur compounds: Unburned coal in cinders may contain pyrites and marcasite (both iron sulphide), or organic sulphur compound; both are deleterious. Sulphur compounds may also be present as sulphide or sulphate in ashes but these are not necessarily harmful if present in small quantities. (3) Iron compounds: The presence of some iron oxide, such as (Fe203) or (Fe304) in cinders is not by itself harmful to concrete~ Other iron compounds, such as iron sulphide (pyrite), may cause staining and popouts in concrete. Free metallic iron can, by oxidation, have the same deleterious action in concrete exposed to moisture, but it is usually separated out during the processing of the aggregate. (4) Free lime: Free lime in cinder aggregate is very deleterious. It is usually formed by the decomposition at high temperature of the calcium carbonate or the mineral gypsum or anhydrite (calcium sulphate) which may be present in coals. Free lime in cinders occurs as small hard-burned particles which hydrate very slowly, accompanied by a large expansion, often causing popouts. 2. Slags. Slags are formed by many metal producing processes including the iron blast furnace, the open kettle furnace, the Bessemer converter, copper or lead smelters and others. Slag used for concrete is the by-product of the iron blast furnace and the term "slag" is so meant in this study. Slag (iron blast furnace) is defined by the American Society of Testing Materials as the "non-metallic product consisting of silicate and aluminosilicate of lime bases which developed simultaneously with iron in blast furnaces." The development of the slag is essential for the separation of the metal from the oreo

-39 - Cooling of the molten slag is done by three methods. The resulting products are called "air-cooled", '"ranulated", or "expanded slag" and are produced under many commercial names. a) Air-cooled Slag. The bulk of the slag is produced by this method, in which molten slag is run to a large pit where it is left to cool slowly in air. Many kinds of pits are used but the most common are the "furnace pit' and the "modified pit". The furnace pit may be 40 ft. wide and 200 ft. long into which the molten slag runs directly from the furnace. Two such pits are usually used, one for receiving the molten material while the hardened slag is being excavated from the other. Modified pits are of several kinds but they are usually long and narrow. Molten slag is dumped on the side of the pit from large ladles moving on rails along the side. Hardened slag may be excavated from one end of the pit while molten slag is dumped in the other end. In both kinds of pits, hardened slag is formed in layers which can be dug out easily with a power shovel. It must be further processed by crushing and screening to the required specification sizes before it is used as concrete aggregate. Air-cooled slag is characterized by sizable internal cavities, broken surfaces and rough texture. Air-cooled slag is normally not recommended as concrete aggregate if it has a unit weight, compacted dry, of less than 70 pounds per cubic ft. It is therefore, considered as dense aggregate even if it produces concrete lighter in unit weight than does sand and gravelo b) Granulated Slago Granulated slag is produced by cooling the molten slag suddenly in water, causing it to expand and form a glassy

-40 - lightweight material different in physical properties from the aircooled slag. Three methods are used in producing granulated slag commercially: (1) The pit method: Molten slag runs directly into a pool of water. (2) The jet method: Molten slag is poured slowly on a strong jet of water, expanding suddenly in air and falling in the pool of water to quench further. (3) The dry method, For this a special machine is used to break the stream of molten slag into small particles before it is introduced into the water. A lesser amount of water is used by this method, and the still hot, granulated slag usually dries rapidly. c) Expanded Slag. This is also produced by cooling molten slag suddenly. It is different from the granulated slag in that only a limited and predetermined amount of water is used. It is commercially produced by two methods: (1) The jet method: A high pressure jet of water impinges on a stream of molten slag which expands in air and falls into a dry pit with slight further expansion. (2) The machine method: Molten slag is broken into particles which expand by coming into contact with a limited amount of water thus generating steam for quenching. Aggregate is produced practically dry by this method and may be processed directly. Expanded slag is a friable foamed material with a cellular structure containing a large number of vesicular internal cavities. It is, therefore, light in weight and has a rough texture. It is produced under many commercial names, among which are Waylite and Celocrete, both of which are included in this study.

- 41 - Mineralogically, iron blast furnace slag is crystalline. The formation of the crystals depends on the rate of cooling. Crystallation usually starts at 2640 F and continues until the mass is solidified. Crystals may range in size from submicroscopic to about 1/4 inch long. Slag is relatively free of deleterious materials. Being formed at a very high temperature, this eliminates organic materials unless it is contaminated later during storage or handling. Under unusual furnace operations, small quantities of metallic iron, coke or fluxing stone might escape from the furnace into the slago But the metallic iron is separated out if properly processed and the coke does not usually exceed one per cent. However, fluxing stone incompletely consumed by the furnace may be very deleterious and may cause popouts and spalling upon long exposure of concrete to moisture. 3. Expanded Clay, Shale and Slate. Lightweight aggregate suitable for use in concrete is produced by heating clay, shale or slate by a special method to a temperature of incipient fusion when "bloating" occurs, forming, upon cooling, hard lightweight cellular particles. The basic mineral structure of the clay does not have a large effect on the bloating process. Common clay or shale (kaolinite and/or hydromica) may bloat successfully, but the montmorillonite types of clay may be equally good bloaters. The essential requirement is that they contain a small percentage of potential sources of expansive gases such as carbonates, sulphides, sulphates or carbonaceous materials~ Commercially, there are two widely used methods of bloating: the rotary kiln method and the sintering method. They are different methods

-42 - but the essential process is the same: The raw material is heated to incipient fusion, between 1000 C and 2000 C, at which time the material should be in a pyrophastic condition with gases evolving throughout the mass. In this state, the gases cannot escape readily; some become entrapped, causing the bloating to occur, a) Rotary Kiln Method. The raw clay is usually crushed to the size of 1-1/2 inch or larger and fed to slowly rotating kilns in which it is heated. The temperature is allowed to rise slowly until a determined high temperature is reached at which time the temperature is raised suddenly to that of incipient fusion where the bloating occurs. The bloated materials will then solidify by cooling. The hardened material must be processed to a suitable grading before it is used for concrete. b) Sintering Method. Raw clay, crushed to the required sizes, between 3/4 inch and #16 mesh is blended with five to fifteen per cent coal, coke or charcoal that has been ground to pass #16 mesh. The blend is then fed to a pelletizing machine. Pelletizing is important because it has a considerable effect on the type of bloating experienced; each clay appears to have individual pelletizing characteristics. The sintering process is commercially done by two methods: (1) The continuous method: The pellets are fed to traveling grates less than one foot deep traveling in a continuous cycle. At the beginning, the grates pass under the ignition hood which is fired by gas or oil to ignite the fuel in the pellets which will then burn by induced draft as the grates pass a series of wind boxes. By this method, the clay pellets are heated rapidly to the softening point

when the bloating occurs. The fuel in the pellets will soon be exhausted, and the bloated material starts to cool to form cakes of hard cellular materials which are ready for processing. (2) The batch method: This process is essentially the same as that for the continuous method except that the sintering is done by the batch in stationary inclined grates. Each hearth is charged, ignited, bloated, cooled and discharged individually. Expanded clay, shale or slate is light friable aggregate with visicular structure. When well graded, these aggregates frequently provide strong concrete. It has been found possible to make concrete having compressive strength as high as 7000 psi with some expanded shales. These aggregates are produced commercially under several names, among them are "Haydite", which is produced by the rotary method, and "Beslite", which is produced by the sintering method. Both aggregates are included in this study. B. Physical Properties of Aggregates. The physical properties of lightweight aggregate are difficult to test using normal techniques because of their cellular structure, low specific gravity, rough surface and high absorption. Modification of the standard methods of test are sometimes required, and the reproducibility of the results is more difficult than for dense aggregateo Probably because of this, the literature is brief and results are not comparable, due to differences in method or in conditions of testo The ASTM methods of tests for dense, and those suitable for lightweight aggregate, were followed in this work whenever possible. The physical properties studied were the following:

1o Grading 2. Unit weight 3. Specific gravity and absorption 4. Deleterious inclusions a) Materials passing #200 sieve b) Organic impurities c) Popouts 1. Gradingo Grading, maximum size and surface texture of aggregate, including those of lightweight, have important effects on concrete. The workability and consistency of plastic concrete, the amount of cement required to produce a certain strength and surface texture of hardened concrete are all influenced by these properties of the aggregate. Sieve analysis is a measurement of the grading of the aggregate which in turn is essential for the designing of concrete mixes Fineness modulus (FoMo) is a factor used for "identifying" the grading. Defined by the ASTM as an "emperical factor obtained by adding the total percentage of a sample of aggregate retained on each of a specific series of sieves and dividing the sum by 100," The fineness modulus alone does not supply the complete information — the grading of aggregate must be known because two aggregates having the same fineness modulus may differ sharply in their grading. a) Procedure of Test. Samples of aggregate of suitable size were obtained from the original by using a sample splitter and were dried to constant weight prior to the test. The procedure used in conducting this test is as that described by the ASTM Designation

-45 - C136-46 (Method of Test for Sieve Analysis for Fine and Coarse Aggregates).2 Standard woven wire sieves, conforming to the requirements of the ASTM Specifications were used. The following sizes were included: 3/8 inch, Nos. 4, 8, 16, 30, 50 and 100 sieve. Sieving was done with a mechanical shaker. Shaking time of approximately five minutes for lightweight aggregate was found to result in thorough sieving without breaking the aggregate particles. b) Discussion of Test Results. Results of the grading test are given in Table 4. Aggregate samples are identified in the first column of the table in accordance with the procedure of marking discussed in the second chapter, while the second column gives the type of aggregate and grade (fine, coarse or combined) as sampled. The ASTM Designation C136-46 (Method of Test for Sieve Analysis for Fine and Coarse Aggregates) divides aggregates into find and coarse on #4 sieve. In the second column, the aggregates are defined as they were used in the plant mix (see also Table 3). Such a division was attempted because it is more useful for a comparative study of the aggregates in the condition used in the plant mixes. (1) Sieve analysis: Results of the analysis are given in column 3 of Table 4 and are reported as both passing and retained percentages on each of the sieves usedo Comparative study of these values is difficult, because the aggregates are of different materials or sources. Suitable grading of a certain aggregate is determined by physical properties and those required for the concrete. Almost all the plants included in this study have their own processing equipment (see Chapter I) and can therefoe choose ther mhost suitable grading for their purposeo

TABLE 4. GRADING AND UNIT WEIGHT OF AGGREGATES 1 2 3 4 5 6 Total Percentage Retained or Passing the Sieve Indicated Maximum Unit Sample Type Size No. of 3/8 #4 #8 #16 #30 #50 #100 F.M. Inch or Aggregate Sieve No. 1C9 Cinder Retained 3.9 23.0 56.0 73.3 83.0 88.2 91.5 4.19 3/8" Combined Passing 96.1 77.0 44.0 26.7 17.0 11.8 8.5 3C9 Cinder Retained 1.2 24.2 49.9 65.2 75.1 82.3 87.6 3.86 3/8" Combined Passing 98.8 75.8 50.1 34.8 24.9 17.7 12.4 3SG9 Sand and Retained 0.0 15.0 33.9 53.3 74.5 92.3 98.2 3.67 3/8" Gravel Passing 100.0 85.0 66.1 46.7 25.5 7.7 1.8 Combined 4SG9 Sand and Retained 0.0 14.0 25.8 49.1 65.1 81.3 96.3 3.32 #4 sieve Gravel Passing 100.0 86.0 74.2 50.9 34.9 18.7 3.7 Combined Weight Pounds per cu.ft. 58.6 57.3 111.8 118.0 5C9 Cinder Retained 12.2 46.1 61.0 74.9 81.8 86.7 90.8 4.54 Combined Passing 87.8 53.9 39-0 25.1 18.2 13.3 9.2 6ES19 Expanded Retained 30.4 97.5 100.0 100.0 100.0 100.0 100.0 6.28 Slag Passing 69.6 2.5 0.0 0.0 0.0 0.0 0.0 Coarse 3/8" 1/2" 58.8 O\ 47.5 6ES29 Expanded Retained Slag Passing Fine 6.9 42.8 56.3 73.1 83.3 91.2 96.5 4.50 93.1 57.2 43.7 26.9 16.7 8.8 3,5 7W19 Waylite Retained 25.0 88.5 96.2 97.1 97.1 97.6 98.1 6.00 Coarse Passing 75.0 11.5 3.8 2.9 2.9 2.4 1.9 7W29 Waylite Retained 0.0 1.5 15.4 35.3 60.5 79.5 88.7 2.81 Fine Passing 100.0 98.5 84.6 64.7 39.5 20.5 11.3 7B9 Beslite Retained 4.2 34.1 67.4 79.9 85.1 88.2 91.7 4.51 Combined Passing 95.8 65.9 32.6 20.1 14.9 11.8 8.3 8CEL19 Celocrete Retained 0.0 83.4 90.0 90.7 90.7 91.4 92.1 5.38 Coarse Passing 100.0 16.6 10.0 9.3 9.3 8.6 7.9 8CEL29 Celocrete Retained 0.0 4.7 22.5 44.5 60.7 73.8 87.5 2.94 Fine Passing 100.0 95.3 77.5 55.5 39.3 26.2 12.5 10CA9 Cinder Retained 1.7 31.2 53.7 66.8 77.2 84.3 89.3 4.04 Combined Passing 98.3 68.8 46.3 33.2 22.8 15.7 10.7 3/8" 1/2" 3/8" 1/2" 3/8" 3/8" 66.1 47.9 56.9 66.9 28-.5 46.3 3/8" 60.3

-47 - Workability of plastic concrete is an important factor influenced by the grading of aggregateo Research has generally indicated that to obtain a given workability, lightweight aggregate requires a larger percentage of the material passing #4 sieve than dense aggregate. This is probably because the lightweight aggregate contains a higher percentage of voids. From column 3, it can be seen that such is not the case for the aggregates under test. Some plants (see Table 3) used air entraining admixtures or add a certain percentage of sand to their mix, It should also be noted that the workability is not such an important problem when molding by vibration and pressure; the addition of fine material to the mix may, therefore, be to improve the surface texture of the concrete rather than to provide workability, TABLE 5.*- UNIT WEIGHT REQUIREMENTS OF LIGHTWEIGHT AGGREGATE FOR CONCR.EITE MASONRY UNITS Dry loose weight, Size Designation Max. lb. per cubic ft. Fine aggregate 70 Coarse aggregate 55 Combined fine and coarse aggregate 65 Limitations on the grading of lightweight aggregate have been tentatively established by the ASTM Designation C331-53T (Specifications for Lightweight Aggregate for Concrete Masonry Units). The grading of the aggregates in Table 4, except for 6ES29 (Expanded * ASTM Desig. (C331-53T - Tentative Specification for Lightweight Aggregates for Concrete Masonry Units). 22

-48 - Slag, fine) and 8CELl9 (Celocrete, coarse) are found to be in agreement with these limits. Fine expanded slag is found to be coarser and coarse Celocrete found to be finer than these limits; but it should be noted that the same designation permits waiving the grading requirements to acquire special characteristics for the concrete, such as texture or weight, etc. (2) Maximum size: The maximum size of the aggregates are given in column 5 of Table 4 which were calculated by permitting a tolerance of 15 per cent; thus if 15 per cent or less of the aggregate is retained on a certain sieve, the sieve opening will be considered as the maximum size for the aggregate, but if more than 15 per cent of the aggregate is retained, the next largest size will be considered as the maximum size of the aggregate. The maximum size of the aggregate as given in column 5 are 3/8 inch for fine and 1/2 inch for coarse aggregate. Coarse sizes would be unsuitable for producing hollow precast concrete units which have thin sections. (3) Fineness Modulus (F.M.): The F.M. of the aggregates differ considerably as can be seen from column 4 of the table. Because of the maximum sizes, the FoM. of the coarse aggregates are comparatively low. Combined aggregates - lightweight - have a higher F.M. than dense aggregates, with sample 5C9 having the highest value, which might explain the addition of the natural sand to the plant mix. 2. Unit Weight. Expressed in pounds per cubic foot, the unit weight of the aggregate provides useful information. It is used in calculating the percentage of voids in the aggregate which enters into the calculations of concrete mixes or in predicting the weight of the concrete. For

-49 - air-cooled slag and lightweight aggregates, it is used to indicate the quality of the aggregate, which is covered by specifications. a) Procedure of Test. The unit weights of dry compacted aggregates were determined by following the procedure described by the ASTM Designation C29-42 (Method of Test for Unit Weight of Aggregate).22 Samples were dried to constant weight and compacted by the jigging method for lightweight aggregate and the rodding method for dense aggregate in a steel container of Ool0 cubic fto capacity. b) Discussion of Unit Weight Test Results. Results of the test are given in Table 4 which represents the average of five determinations. The unit weight of the dry aggregate depends on the bulk specific gravity and the volume of external voids it contains. The volume of external voids depends on the grading of the aggregate and the shape and size of the particles. It is observed from Table 4 that Beslite (7B9, combined) is the heaviest of the lightweight aggregates, having a unit weight of 66.9 pef compared to 118.0 pcf for sand and gravel (4SG9, combined). Celocrete is the lightest, having a unit weight of 28.5 pcf for coarse aggregate (8CELl9) and 46.3 pcf for fine aggregate (8CEL29). Unit weights of all these aggregates are sufficiently low so as to conform to the requirements of ASTM Designation C331-53T (Tentative Specifications for Lightweight Aggregates for Concrete Masonry Units). 3. Specific Gravity and Absorption. Knowledge of the specific gravity of the aggregate is important for the calculation of its solid volume contribution in concrete mixes. It is sometimes used in the comparative study of the weight of lightweight aggregate independently of the gradation. For

-50 - some aggregates it is also used for judging their quality. The "bulk specific gravity" (B.SoG,) and the "apparent specific gravity" (A.S.G.) are most commonly used in these calculations. There is some confusion in the literature as to the meaning of the different terms used for the specific gravity, but in this study they bear the meaning contained in the following definitions of the ASTM Designation E 12-27 (Definition 22 of Terms Relating to Specific Gravity): "Apparent Specific Gravity (of solids). The ratio of the weight in air of a given volume of the impermeable portion of a permeable material (that is, the solid matter including its impermeable pores or voids) at a stated temperature to the weight in air of an equal volume of distilled water at a stated temperature... O" "Bulk Specific Gravity (of solids). The ratio of the weight in air of a given volume of a permeable material (including both permeable and impermeable voids normal to the materials) at a stated temperature to the weight in air of an equal volume of distilled water at a stated temperature...." Absorption of the aggregate and natural moisture content aid determination of the amount of water to be added to the mix to obtain the required water-cement ratio. Certain physical characteristics of the concrete were studied in terms of the absorption of the aggregate. It has been found that the absorption of cohcrete and, therefore, its shrinkage is affected by the absorption of the aggregate. An attempt has also been made to correlate the durability of the concrete with the absorption, porosity characteristic and the specific gravity of the aggregate. Different methods of tests were used for the fine and coarse aggregate in determining the specific gravity and absorption. a) Coarse Aggregate. Four different coarse aggregates, all lightweight, were tested by the standard procedure of the ASTM

-51 - Designation C127-42 (Method of Test for Specific Gravity and Ab22 sorption of Coarse Aggregate), with slight necessary modifications. Because of the relative lightness of the aggregate tested, a smaller sample (by weight) was used, and since the maximum size of the aggregate is 1/2 inch, it was decided to reject the part of the sample passing #4 sieve instead of 3/8 incho Results of the tests are given in Table 6 and will be discussed later (in part c). Detailed data and equations used in the calculations of the results are presented in the appendix. b) Fine Aggregate. Six combined and three different fine aggregates also were tested by this method. Combined aggregates were considered as fine aggregates because they consist of mostly fine materials (passing #4 sieve). In preparing test samples for the dense aggregates, the procedure specified by the ASTM Designation C128-42 (Methods of Test for Specific Gravity and Absorption of Fine Aggregate)22 was followed. However, modification of this procedure was necessary to determine the saturated surface dry condition of the test samples of the lightweight aggregates. This will be discussed in part c. A procedure different than that specified above was used in conducting the test. The Le Chatelier Flask was used for determining specific gravity. It has a bulb in the bottom which has a capacity of approximately 250 cc and is equipped with a long graduated neck of 24 cc capacity. A clean flask was filled with distilled water to a point reading between 0.0 to 1.0 cc, stoppered and left standing in a bath of water at 20 C while preparing the aggregate sample. Upon reaching the saturated surface dry condition, the sample was mixed

-52 - thoroughly and divided into two portions, the small portion (35 to 50 gms.) weighed and immediately introduced inside the flask while the rest of the sample, used for determining the absorption, was weighed and placed inside an oven to dry to constant weight at 210 F. Just before introducing the sample to the flask, the reading of the water was taken and recorded as the "initial reading". Rolling the stoppered flask on its side was found to help remove the air bubbles from the water; after this, the flask was placed again in the water bath which was constantly maintained at 20 C for one hour before taking the "final reading". The volume of the aggregate sample inside the flask is, of course, equal to the difference between the initial reading and the final reading. When the aggregate sample (of the absorption test) inside the oven reached a constant weight, it was taken outside to cool to room temperature and weighed. The difference in weight between the saturated surface dry condition and the oven dry condition of the sample is, of course, the weight of the water of absorption. Results of the test are given in Table 6 and will be discussed later (in part c). Detailed data and the equations used in the calculation of the results are presented in the appendix. c) Discussion of Test Results for Coarse and Fine Aggregates. It is difficult to determine the saturated surface dry condition of lightweight aggregate. At this condition the aggregate must contain no free surface moisture and must absorb no water; but because of the rough surface texture and the large number of open surface voids, the drying of these surfaces without drawing out water of absorption, is

-53 - a careful and time consuming process, especially for fine aggregates. The "cone method", which was specified by the ASTM Designation C128-42 (Method of Test for Specific Gravity and Absorption of Fine Aggregate) for determining the saturated surface dry condition of fine dense aggregate was found to give inconsistent results when applied to lightweight aggregate. This is, probably, because the rough and angular shapes cause them to interlock when compacted inside the cone by this method, providing a mechanical bond causing the cone of the aggregate to retain its shape in absence of surface moisture. Several other methods were therefore tried and it was decided to establish a procedure actually including three methods simultaneously, as a criterion of the surface dry condition: (1) using a modified cone method in which the aggregate was tamped inside the cone with. a wooden rod 15 times only, (2) determining the point at which the aggregate will run loose through the fingers and (3) by noticing the change in color of the aggregate when it loses its surface sheen when dried. Results given in Table 6 for the specific gravity and absorption, which are the average of two determinations, will be discussed separately. Calculations are given in Tables 31, 32 and 33 in the appendix. (1) Specific Gravity: No great difficulty was met in determining the specific gravity and the absorption of saturated surface dry samples of coarse aggregates. The Le Chatelier method proved difficult to apply for some lightweight aggregates of low specific gravity. These aggregates have a few particles that have a specific gravity close to 1.0 which will float. The number of

these particles is so small in comparison with the test sample that it should not affect the results but does make the final reading difficult. Results are given for the bulk specific gravity and the apparent specific gravity of the aggregates; the bulk specific gravity is calculated on both the saturated surface dry basis and the oven dry basiso Any one set of these results, such as those of the bulk specific gravity oven dry basis, should be adequate for comparison. Thus from Table 6 it is observed that the latter gravity for cinders, except for one sample, range from only 1.71 to 1.74, while expanded slag, Waylite and Beslite have close results ranging from 1.47 to 1.5 for coarse aggregate and from 1.78 to 1.80 for fine aggregate. Celocrete has the lowest specific gravity, 0.97 for coarse and 1.49 for fine. Dense aggregate gave relatively high results of 2.63 and 2.66 for combined aggregates. It is also to be noticed that, generally, fine aggregates have a higher specific gravity than coarse aggregates which may be explained by the fact that coarse aggregates contain a higher percentage of internal voids. (2) Absorption: No difficulty was experienced in determining the absorption of lightweight aggregate after establishing a criterion for the saturated surface dry condition of the sample. It should be noted that the results given in Table 6 represent partial saturation due to 24 hours immersion in water at room temperature. The result of such a test is used in concrete mix calculations because only partial saturation is considered to occur inside the mixer. The absorption is reported in Table 6 in both percentages of dry weight and dry compacted volume. Absorption as percentage of

-56 - dry weight is the most commonly used in the calculation but it is not significant in a comparative study of aggregates with widely varied unit weights as in the case of lightweight aggregate. Considering, therefore, values based on dry compacted volume, it is observed in Table 6 that lightweight aggregates, as expected, have a higher absorption than dense aggregates, and the lightweight coarse particles absorb more water than the fine particles. Waylite, fine (7W29), has the lowest absorption for lightweight aggregate of 4.66 per cent, while dense aggregate has the value of 2.50 per cent and 2.47 per cent for (3SG9) and (4SG9) respectively. Cinders have a high absorption value, ranging between 7.73 per cent and 8026 per cent; Waylite and Expanded Slag have lower values while Celocrete and Beslite have a higher value than cinders. In fact, Celocrete has the highest absorption, approximately six times that of dense aggregateso The number of tests conducted are too limited to draw any conclusions as to the relationship between the specific gravity and absorption; however, others found that absorption is not necessarily related to the specific gravity nor to the total pore space; it has been found possible to produce lightweight aggregate of low specific gravity and absorption due either to an impervious shell or to a visicular structure with a discontinuous pore space. 4. Deleterious Inclusions. Although lightweight aggregates are relatively free of deleterious materials such as organic impurities, they may contain other materials that are injurious to concrete, such as potential staining or popout materials. Three types of extraneous inclusions, all potentially deleterious, will be studied separately: (a) materials

finer than #200 sieve, (b) organic impurities, and (c) materials causing popouts of concreteo a) Materials Finer Than #200 Sieve. Excessive amounts of this very fine material are deleterious to concrete. Two types of such material may occur, both present in the form of surface coating to aggregate particles. These are (1) That resuiting from dust accumulation which might be produced by crushing the aggregate. This does not adhere firmly to the surface of the particles and can be removed by washingo Inside the mixer this will increase the amount of fines and require more water for workabilityo An excessive amount of this material is therefore deleterious. (2) Materials that are clayey in natureo These materials are very deleterious because when wet they will. adhere to the surface of the aggregate particles up to the time they are imbedded in concrete, thus preventing proper bond with the cement. The purpose of this test is, therefore, to determine the amount of such materials in the aggregateo The procedure used for this test is the same as that described by the ASBM Desigration Cli7-49 (Method of Test for Amount of Material Finer than No. 200 Sieve in Aggregates).22 Results are reported in Table 6 as percentage (by weight) of the total aggregate; each result is the average of three determinations. Care was employed to prevent the decantation of coarse particles of some lightweight aggregates that have low specific gravity. In general the results indicate that aggregates are free of excessive amounts of such materials. Dense aggregates showed less than the maximum allowed by the ASIMo There are presently no

-58 - limitations specified by the ASTM for the amount of these materials in lightweight aggregates. In this test, the coarse aggregate showed a small percentage of materials finer than #200 sieve, but combined and fine aggregate showed a higher percentage than coarse aggregate alone. Cinders have a widely varying result ranging from 3.7 per cent for (5C9) to 10.7 per cent for (10CA9), with the latter showing the highest value obtained. Of the fine aggregates, Celocrete showed a high result of 8.7 per cent. b) Organic Impurities. This test provides a quick method for determining the presence of deleterious organic materials in aggregates. The amount of such materials and effect on concrete are to be determined by further tests. (See ASTM Designation C87-52 - Method of Test for Measuring Mortar-Making Properties of Fine Aggregate, and Designation C123-44 - Method of Test for Coal and Lignite in Sand). Slag and expanded clay are relatively free of organic impurities because of the very high temperature at which slag is formed or at which expanded clay is bloated. Cinders are usually free of such materials unless they contain coal or lignite due to poor firing of the furnace. But it should also be kept in mind that aggregate originally free of such materials may be contaminated during storage or handling. Therefore, such a test is necessary. The procedure described by the ASTM Designation C40-48 (Method of Test for Organic Impurities in Sands for Concrete)22 was used on aggregate samples weighing less than one pound and which was obtained from the original samples by a sample splitter. Table 6 contains the results of this test. These results indicate that none of the aggregates have important amounts of

organic impurities. Lightweight aggregate caused no change in color in the sodium hydroxide solution after 24 hours of soaking; one dense aggregate sample (4SG9) caused a slight change in color while the secone sample (3CSG9) showed a little darker but much lighter than the reference color which was also prepared to the requirements of the ASTM Designation mentioned above. These results indicate that further testing for organic impurities in the aggregate and the effect in concrete is not necessary. c) Materials Causing Popouts of Concrete. Popouts in concrete masonry units are a very undesirable characteristic which is the result of a disruptive expansion of considerable magnitude usually generated beneath the surface. It is caused by the presence of certain deleterious materials in the concrete. In cinders, popouts are usually caused by iron compounds, freelime or organic sulphur; in slag they are caused by incompletely fused fragments of flux stones which may be discharged with the slag during unusual operation of the furnace; in expanded clay, raw clay lumps might cause popouts of concrete. The test was conducted on concrete specimens made in the laboratory with the aggregates under test and a constant slump and cement-aggregate ratio. This was in accordance with the ASTM Designation C331-53T (Tentative Specifications for Lightweight Aggregate for Concrete Masonry Units).22 Specimens were then autoclaved for three hours using the procedure ASTM Designation C151-52 (Method of Test for Autoclave Expansion of Portland Cement).22 Two concrete specimens, 2 x 2 x 11 inches were cast from each aggregate sample. Data on concrete mixes are shown in Table 34

in the appendix. Proportioning was done by dry volume, but all measurements were done by dry weight. A constant proportion of 1:6 was used for all mixes, that is, one part of cement to six parts of combined aggregates by volume. Aggregates were combined in the laboratory in the same proportions as used by the producers in their plant mix unless they were sampled from materials that had already been combined by the producers. Regular cement was used without an air-entraining agent. Mixing was done by hand which was more convenient for the small concrete mixes. Trial batches were made to determine the amount of water to be used with the mix to produce a constant slump of approximately 2.7 to 2.9 inch. Workability was poor for lightweight aggregate at such consistency and without the use of an air-entraining agent, but it was found possible to improve the workability by mixing the aggregate and approximately one-half of the mixing water first, then adding_ the 1 th; rest of the water. and continuing mixing. Moist curing was employed at normal pressure and a temperature of 80 F for 24 hours after which the specimens were taken outside the chamber, weighed and autoclaved immediately. Table 6 contains the results of this test which indicates that none of the concrete tested showed any popouts (or staining) on the surface. This fact was determined by visual inspection of the specimen and by reweighing to determine any loss of weight which might result from autoclaving. It may, therefore, be concluded that the aggregates under test are free from deleterious material capable of causing popouts in concrete. See Chapter X, however, for results of staining tests on full size commercially manufactured units.

-61 - Figure 11. Concrete Specimens After Staining and Popouts Test.

CHAPTER IV DIMENSIONS OF THE MODULAR UNIT The term "modular unit" is used here to mean a hollow precast concrete masonry unit having the nominal dimensions of 8 x 8 x 16 inches. The actual overall dimensions of this unit are specified as 7-5/8 x 7-5/8 x 15-5/8 inches, thus allowing 3/8 inch for a standard mortar joint. Although a large number of different shapes and sizes of precast concrete units are available, the bulk of the product is presently being made from these modular units. Substantial economy is the reason for this modular coordination; it requires the dimensional standardization of building materials and equipment on the basis of a certain module and the fitting of these materials into a building planned and dimensioned on this modular basis. The "American Standard Module" is equal to 4 inches; nominal dimension of the modular unit is a multiple of this. Flexibility in building dimensions equivalent to this module is accomplished by using supplementary units or cutting the modular unit. In this chapter study will be made of the dimensional properties of the modular unit including its actual overall dimensions, gross cross-sectional area, net volume and their interrelation. A. Definitions. 22 (1) Gross cross-sectional area: ASTM Designation C140-52 defines the gross cross-sectional area of the modular unit as the "total area of a section perpendicular to the direction of the load, including areas within cells and within re-entrant spaces unless these 62

-63 - spaces are to be occupied in the masonry by portions of the adjacent masonry." (2) Gross volume: The gross volume of the modular unit is the total volume included within its overall dimension. (3) Net cross-sectional area: The net cross-sectional area of the modular unit is the average of the areas of concrete present in sections perpendicular to the direction of load. (4) Net volume: The net volume of the modular unit is the total volume of concrete in the unit. (5) Hollow masonry unit: The Ao S. A. defines the hollow masonry unit as "a masonry unit whose net cross-sectional area in any plane parallel to the bearing surface is less than 75 per cent of its gross cross-sectional area measured in the same plane."29 B. Procedure of Test. The overall dimensions of the unit were measured directly by a caliper and a scale which reads to 0.01 inch. Five specimens were tested for each type of product; the length, depth and thickness were separately measured at the two faces of the specimen and averages recorded. A different procedure (using the same specimens) was used for determining the net volume (ASTM Designation C140-52 - Methods of Sampling and Testing Concrete Masonry Units22). Each specimen was saturated with water for 24 hours, then weighed suspended in water, removed from the water and left to drain for one minute over a 3/4 inch wire mesh; visible surface moisture was then removed and the specimen immediately weighed in air. The net volume of the specimen is, in the metric system, numerically equal to the difference between the

-64 - weight of the saturated specimen in air and its weight suspended in water. Further details on this procedure will be presented in Chapter V. C. Discussion of Test Results. Dimensional properties of the modular units are reported in Table 7. Only average results (of the 5 specimens) are given in this table, while detailed data and calculations are presented in Tables 35, 36 and 37 in Appendix II. The industry's adoption of the system of modular coordination of dimension is evident in these tables,, In columns 3, 41 and 5 of Table 7, the average overall dimensions (height, thickness and length) of the modular units are presented. These results correspond to standard dimensions of 7.625 x 7.625 x 15.625 inches. Specifications (see references 28 and 29) permits a maximum deviation in the dimension of the individual unit of + 0.125 inch; only three types of the products tested (see Table 36 in Appendix II) showed units with deviations exceeding this. These products are 1C1, 3C1, and 8CEL1; of these, 8CEL1 has the largest deviation, +0.225 inch. It is also noticed that no single specimen of a given product showed deviation from the average exceeding that permitted. Columns 6 and 8 of Table 7 present the average gross volume and gross area of the units: both are useful in estimating, for example, total requirements for a building. Column 10 presents the average net areas as percentage of the gross areas: these values fall between 50 to 60 per cent which indicates all are within the definition of hollow masonry units (see definition on page 63); data for calculations of these percentages are given in Table 37 in the appendix.

-65 - Figure 12. Setup for the Determination of Net Volume of the Unit.

- 66 - TABLE 7.* DIMENSIONAL PROPERTIES OF THE MODULAR UNITS (AVERAGE VALUES) 1 2 3 4' 5 6 7 8 9 10 Gross Dimensions Net Thick- Gross Gross Area Type ness Length Height Volume Net Area Net as % of Sample of t L h (t-L'h) Volume (t*L) Area Gross No. Aggregate in. in. in. in3 in3 in2 in2 Area 1C1 2CES1 2H1 3C1 3SG1 4SG1 5CS1 6ES1 7W1 7BW1 8CEL1 9S1 lOCAl Cinders Cinders & Expanded Slag Haydite Cinders Sand & Gravel Sand & Gravel Cinders & Sand Expanded Slag Waylite Beslite & Waylite Celocrete Air-Cooled Slag Cinders & Fly Ash 7.70 7.67 15.68 15.63 7.74 7.67 935.6 918.4 7.68 7.69 7.67 7.72 7.70 7.70 7.68 7.63 7.77 7.64 7.71 15.65 15.74 15.64 15.74 15.61 15.65 15.63 15.65 15.72 15.61 15.67 7.67 7.63 7.62 7.71 7.67 7.70 7.67 7.65 7.73 7.62 7.67 922.5 924.1 914.5 937.6 921.7 927.7 921.3 913.9 944.4 908.3 925.7 477.2 517.3 515.2 554.3 502.3 469.4 464.5 459.0 474.4 470.6 507.0 512.6 503.1 120.7 119.8 61.7 67.5 120.3 121.1 120.0 121.6 120.2 120.4 120.1 119.5 122.1 119.3 120.7 67.2 72.6 65.9 60.9 60.4 59.5 61.8 61.5 65.5 67.3 65.6 51 56 55 60 55 50 50 50 51 51 54 56 54 * Notes: 1. Values are average of 5 specimens. 2. Net Volume - see Table 35 in the appendix. 3. Net Area = Net Vol. (See Tables 35 and 37 in Appendix II.) h 4. Net Area as per cent of Gross Area = Net Area x 100 (Table 37). Gross Area

CHAPTER V UNIT WEIGHT, MOISTURE CONTENT AND ABSORPTION The weight of the masonry units, as sampled, and in the wet and dry conditions, moisture contents and absorption are considered in this chapter. Sampled weight of the masonry unit is useful for calculating its moisture content at the time of sampling and for estimating its dead load on structural members. Wet weight is the weight in air of the units after 24 hours of submersion in water, used in calculating the absorption. Dry weight is the weight of the unit after drying in an oven to an essentially constant weight. Unit weight of the concrete (in pounds per cubic foot) was determined under three conditions (1) as sampled, (2) wet, and (3) dry. These unit weights were calculated using the measured weight of the unit under these conditions and the net volumes of the individual units (presented in Chapter IV). Unit weight (dry) is probably the most convenient basis for the correlation of the physical properties of the various concretes. The moisture content of the unit when sampled (sometimes called "natural moisture content") is usually expressed as percentage of 24 hour absorption. Specifications frequently limit the permissible natural moisture content to prevent excessive shrinkage of concrete masonry walls. Drying shrinkage of concrete is also found to be correlated with absorption of the aggregates with which it is made.19 It is more pronounced in those having high absorption. Resistance of concrete 67

-68 - to the action of freezing and thawing and the attack of silage acids is found to be correlated with the rate of absorption.12 A. Procedure for Determining Unit Weight, Moisture Content and Absorption. Five full size units were used for each product. They were selected at random upon bringing the samples to the laboratory, marked and weighed in air. This weight, called here the "sampled weight", was determined for the specimen within less than 12 hours after sampling at the plant. In determining the moisture content and absorption of the units, the procedure of the ASTM Designation C140-52 (Methods of Sam22 pling and Testing of Concrete Masonry Units) was followed: Specimens were submerged in water at room temperature (60 to 80 F) for 24 hours. They were then weighed suspended in water, taken from the water, allowed to drain for one minute and weighed in air (see Chapter IV, part B). This latter weight is called "wet weight". Specimens were then dried in a forced ventilation oven at 210 F-230 F for more than 24 hours. Drying was discontinued when two successive weighings at 2 hour intervals showed a loss in weight less than 2 per cent of the last previously determined weight of the specimen. This last weight is called the "dry weight". B. Discussion of Test Results, 1. Unit Weight. Unit weight of the specimens (as sampled, wet and dry) are given in Table 8. These are calculated from the weights of the units in the three foregoing conditions (also given in Table 8) and the unit volume as first reported in Chapter IV. Each value in Table 8

-69 - Figure 13. Electrical Automatic Oven Used for Drying Masonry Units.

-70 - TABLE 8. WEIGHTS OF THE MASONRY UNITS AND THE UNIT WEIGHTS 1 2 3 4 5 6 7 8 Unit Weight Weight of the Masonry - -, Wgh We iht of the Masonry Net (based on net volume) Unit (Pounds) Volume (pcf) Sample As of the As No. Sampled Wet Dry Unit Sampled Wet Dry cu, ft. 1C1 2CES1 2H1 3C1 3SG1 4SG1 5CS1 6ES1 7W1 7BW1 8CEL1 9S1 10CAl 26.8 29.7 27.8 27.7 39.6 36.4 27.3 28.8 27.3 28.7 24.6 38.3 26.4 29.3 32.1 29.4 32.1 41.1 38.3 29.3 30.9 29.7 31.4 26.8 40.8 28.9 25.1 27.9 25,1 26.7 38.2 35.5 26.1 26.9 26.7 27.8 22.6 37.5 24.8.276.299.298.321.291.272.269.266.275.272.296.297.291 97.1 99.1 93.1 88.5 136.3 134.1 101.5 108.4 99.3 105.3 83.0 129.2 90.7 106.1 106.7 98.7 100.1 141.2 140.8 108.9 116.3 108.3 114.1 90.3 137.7 99.0 90.9 93.1 84.2 83.4 132.8 129.2 97.1 101.3 97.3 102.2 76.2 126.3 85.2

is the average of the results of the five specimens tested from each product. Detailed data and calculations of the results are given in Tables 38 and 39 in Appendix IIIo The unit weight in dry condition is the most convenient form for a comparative study. Returning to column 8 of Table 8, it can be seen that the weights of the products vary, ranging between a minimum of 76.2 pcf (pounds per cubic ft.) for 8CEL1 (Celocrete) to a maximum of 132.8 pcf for 3SG1 (sand and gravel). The other sand and gravel product (4SG1) had a similar weight of 129.3 pcf. Air-cooled slag (9S1) averaged 126.3 pcf. Concrete made with such aggregate is usually classified as dense. The maximum weight of the products made with strictly lightweight aggregate is 97.3 pcf for 7W1 (Waylite). Rough correlation exists between the unit weight of dry aggregate and concrete. This can be seen by comparing the values in Tables 4 (of Chapter III) and 8. Good correlation is not to be expected because of the difference in plant mixes and the fact that some plants add air-entraining agents to their mixes. 2. Moisture Contents and Absorption. Moisture contents and absorption of the masonry units are presented in Table 9. Each value is the average of the results of five specimens tested for each product. Detailed data and calculations are given in Appendix III (Tables 40 and 41). a) Absorption. Absorption is presented in two forms: as pounds of water absorbed per cubic foot of concrete (pcf) and as percentage of dry weight of the unit. The latter form, although it is a common and convenient method of expressing the absorption, is not very useful for a comparative study of aggregates of widely varying

-72 - specific gravity, as in these tests. The prior method of reporting the absorption (as pcf) is based on the net volume of the unit. Concrete made with lightweight aggregate generally absorbs more water than does that made with dense aggregate (see column 4 of Table 9). All-cinder products gave high absorptions with 3C1 producing the maximum of 16.8 pcf followed closely by 1C1, also cinder, averaging 15.1 pcf. Cinders mixed with sand (5CS1) or fly ash (lOCAl) produced lower results of 12.3 and 13.9 pcf, respectively. Other lightweight aggregates showed varied results, ranging between 14.9 pcf for 6ES1 (expanded slag) and 11.0 pcf for 7W1 (Waylite). Dense aggregate gave absorptions of 8.5 pcf for 3SG1 (minimum value) and 10.3 pcf for 4SG1 (both are sand and gravel), while air-cooled slag had a higher absorption of 12.7 pcf. The ASTM Designation C90-52 (Specifications for Hollow Load-Bearing Concrete Masonry Units)22 specifies a maximum permissible absorption of 15 pcf. Only two of the products under test had absorptions exceeding this maximum, both are cinder products, namely 1C1 and 3C1. There is very slight correlation, if any, between the concrete absorption and the unit dry weight or between the absorption of aggregates and concretes. Research has indicated that type of aggregate, mix proportion, use of admixtures and, probably, the method of curing all affect the absorption of the units. b) Moisture content. Moisture contents, as sampled, of the masonry units are presented in Table 9, column 7, as percentages of the absorption. Results varied, ranging between a low 19.4 per cent for 3C1 (cinders) and a high 62.6 for 2H1 (Haydite). The ASTM

-73 - TABLE 9. MOISTURE CONTENT AND ABSORPTION 1 2 3 4 5 6 Absorp- Moisture Water tion Content Type Ab- Absorp- % of As Sample of sorbed tion Dry — Sampled No. Aggregate gmns pcf Weight gmns _, I IJ-. I 1C1 Cinders 2CES1 Cinders & Expanded Slag 2H1 Haydite 3C1 Cinders 3SG1 Sand & Gravel 4SG1 Sand & Gravel 5CS1 Cinders & Sand 6ES1 Expanded Slag 7W1 Waylite 7BW1 Waylite & Beslite 8CEL1 Celocrete 9S1 Air-Cooled Slag 10CAl Cinders & Fly Ash 1896 1900 1972 2441 1110 15.1 14.0 14.1 16.8 8.5 16.6 15.0 17.3 20.2 6.4 7.9 1264 10.3 1503 12.3 12.7 1797 14.9 14.7 795 844 1236 476 503 467 557 874 266 4o8 934 424 727 7 Moisture Content As Sampled Month % of of Absorp- SamplinE tion 41.9 May 43.7 May 62.6 May 19.4 May 45.4 May 37.5 May 37.2 May 48.6 May 19.7 June 25.5 June 49.7 June 27.7 June 39.7 October ) DI D 1369 1602 1895 1529 1833 11.0 13.0 14.0 12.7 11.3 12.7 18.8 9.0 13.9 16.3 Designation C90-52 specifies a maximum moisture content of 40 per cent. Table 9 shows six different products having moisture contents exceeding that value. It will also be noted that in Table 3 (Chapter II) that

almost all the products are normally stored in open yards and, therefore, their moisture content varies considerably from time to time, being higher in damp seasons. Column 8 indicates that most of the sampling was done during the normally moist months, May and June. It may also be repeated here (see Chapter I, page 22) that some investigators actually consider the usual specification requirement of the maximum moisture content (40 per cent) as being too lenient in view of anticipated later shrinkage. But it may also be pointed out that these investigations are based on gradual drying in air rather than the rapid drying in oven which was used in this test. Other research also indicated that rapid drying in the oven at over 210 F results in an abnormally high value for shrinkage or absorption. They concluded that such drying "causes not only the withdrawal of the mechanically suspended water which might be absorbed or withdrawn at ordinary temperatures, but also a certain amount of absorbed or colloidal water more tenaciously held in the cement gel. Hence the absorptions are too large and do not reflect the condition or behavior 25 of the concrete in its usual temperature-humidity environment."

CHAPTER VI COMPRESSIVE STRENGTH Compressive strength is perhaps the most thoroughly studied physical property of concrete. It is variously employed in mix design studies and in testing concrete-making qualities of cement and aggregates. Many of the other physical properties of concrete are studied in terms of, or correlated with, the compressive strength. It was found that the elasticity of concrete, its tensile and flexural strength, wear, permeability and fire resistance are affected (besides the affect of the physical characteristics of the aggregates) by the compressive strength.25 What makes this physical property of concrete even more important is that the usual methods of design are based on allowing concrete to carry compressive stresses only. This makes compressive strength, in this case, the sole criterion of the strength of the concrete. This wide use of the compressive strength of concrete is (among other reasons) because of its economy and simplicity of test. However, the evaluation of the results is, indeed, not as simple as the test itself. It involves many factors affecting the actual strength of concrete and its "indicated strength" (strength indicated by tests which is not necessarily equal to the actual strength). Among factors affecting the actual strength of concrete is the type and quality of the ingredients and their proportions in the mix, water-cement ratio, methods of mixing and curing.... Other factors further affecting the indicated compressive strength are the age of 75

-76 - the specimens, shape and size, capping, speed of testing and even temperature during the test. Some of the factors affecting both types of strengths will be discussed in more detail later. It may also be noted that the strength of concrete structures is not necessarily proportional to either the actual or indicated strength. In plastic concrete the test specimens must be made (mixed, cured...etc.) in the same manner and from the same materials used for the concrete in the structure. In concrete masonry, workmanship is found to affect the strength of the walls. This will also be discussed later. A. Procedure of Test. The procedure of the ASTM Designation C140-52 (Methods of Sampling and Testing Concrete Masonry Units)22 was used for this test. Five full size units were selected at random from samples of each product, capped and tested not later than 60 hours from the time of their arrival at the laboratory. Test load was applied in a direction the same as that for the unit when built in the wall. The two bearing surfaces of the specimen were capped with a paste made of one part (by volume) of cement and one part of calcined gypsum (plaster of paris) mixed with water so that it was plastic and spread evenly without excessive flowing. A cap not thicker than 1/8 inch was placed in the same manner described by the ASTM Designation mentioned above. Allowing the first cap to harden for six hours, the specimen was turned over for capping the opposite surface. Capped specimens were then aged for 24 hours before testing. Imperfect, cracked or damaged caps were completely removed and replaced by new caps.

-77 - The 300,000 pound test machine was equipped with a circular, spherically seated upper bearing block which has the same diameter (10 inches) as the lower bearing block. The blocks did not, therefore, cover the bearing area of the specimen, and two bearing plates having machine polished plane surfaces, were placed at top and bottom, between the bearing blocks and the specimen. These plates were 16 inches long, 12 inches wide and 1-3/4 inches thick, thus covering the bearing area of the specimen. During test they were placed so that a vertical line passed through their centers and the centers of the specimens and the bearing blocks of the machineo Placed thus, the distance between the edge of the bearing blocks and the corner of the plate is 5 inches which is less than 3 times the thickness of the plate as required by ASTM Designation C.14-052. Loading was applied at the rate of 40,000 pounds per minute which conforms with the ASTM requirement. Most of the specimens were broken within 2 to 3 minutes. B. Discussion of Test Results. Results of the compression test are summarized in Table 10, which contains the averages of the results of five specimens tested from each product. Detailed data and the calculation of the results of all the specimens are given in Table 43 in Appendix IV. In Tables 10 and 43 the compressive strength (psi) is reported on the basis of the gross cross-sectional area of the unit (defined in Chapter IV), These areas were found by the actual measurements of the overall dimensions of the units using the procedure discussed in Chapter IV,

-78 - Figure 14. Compressive Strength Test.

-79 - In Table 10, the compressive strengths are also given on the basis of the net cross-sectional area (defined in Chapter IV) of the units (column 7). These areas are also reported in column 5. They are, unlike the gross areas, not found by measurements of the same specific test specimens, but are the average results (also five specimens per product) of measurements on companion mnits. These latter units were also used for the dimensional properties studied in Chapter IV and which are reported in Table 7, column 9. The reason for this is that it was considered that the compression test specimens should be tested in the sampled condition and without being exposed to any preliminary conditioning. Net sectional areas of the unit are calculated from their net volume. Testing for net volume (see Chapter IV, part B) involves submerging the specimen in water for 24 hours. The effect of this saturation on the strength of units made with those various dense and lightweight aggregates is not accurately known, and it was further found detrimental to successful capping. Drying of the specimen is not desirable because drying by heat (after soaking) is considered by some to adversely affect the strength of the concrete. Drying in air at room temperature is not practical because it requires considerable time, while the ASTIM Designation C140-52 (Standard Methods of Sampling and Testing Concrete Masonry Units )L~ requires that compression tests should be performed within 72 hours after sampling. The results reported in Table 10 of the compressive strength (net cross-sectional area basis) should not deviate much from the results that would be obtained by using the actual net areas of the tested specimen. This can be concluded by comparing the gross crosssectional areas of the units used for the compression test and those

-80 - used for the dimensional analysis (of Chapter IV and which are also used in these calculations) which are presented in Tables 42 (Appendix IV) and 36 (Appendix II) respectively. It is obvious, from this comparison, that the gross areas of a certain product in both tables are very similar and the corresponding net areas may, therefore, be expected to be also similar. Referring again to Table 10, study of the compressive strength based on the gross area of the unit reveals that: (1) The highest strength of concrete made with any of the aggregates (dense and lightweight, 1332 psi) is given by 2H2 (Haydite), followed by 9S2 (air-cooled slag) with 1243 psi. (2) Products made with sand and gravel show a lower strength of 1208 psi for 3SG2 and 994 psi for 4SG2. (3) Products made with all-cinder aggregates show a lower strength than those made with cinders mixed with other materials. For example, 1C2 and 3C2, both made with all-cinder aggregates, give strengths of 698 psi and 754 psi respectively, while 2CES2 (cinders mixed with expanded clay) gives 1004 psi. Cinders mixed with sand (5CS2) or fly ash (10CA2) give 877 psi and 906 psi respectively. (4) Products made with expanded slag generally gave lower results of 695 psi by 6ES2, 777 psi by 7W2 (Waylite) and 647 psi by 8CEL2 (Celocrete). Those made with expanded slag mixed with expanded clay (Waylite and Beslite) 7BW2, have higher strength of 989 psi. In the beginning of this chapter it was noted that the actual compressive strength depends on numerous factors, among which are the types of cement and aggregates, proportioning, water-cement ratio, use of admixtures, method of mixing and curing. Further affecting the

-81 - TABLE 10, COMPRESSIVE STRENGTH OF THE MASONRY UNITS _ - ---- ~1... 2.. Type Sample of No. Aggregate 1C2 Cinders 2CES2 Cinders & Expanded Slag 2H2 Haydite 3C2 Cinders 3SG2 Sand & Gravel 4SG2 Sand & Gravel 5CS2 Cinders & Sand 6ES2 Expanded Slag 7W2 Waylite 7BW2 Beslite & Waylite 8CEL2 Celocrete 9S2 Air-Cooled Slag 10CA2 Cinders & Fly Ash 3 Compressive Force lbs. 83,900 120,000 158,620 90,900 145,770 Il 4 Gross Area in2 120.0 119.6 119.1 120.5 120.8 5 6 7 Net Area in2 61.7 67.5 67.2 72.6 65.9 120,700 121.4 60.9 104,400 119.0 6004 83,000 119.9 59.5 Compressive Based on Gross Area psi 698 1004 1332 754 1208 995 877 695 777 989 647 1243 906 I Strength Based on Net Area psi 1357 1778 2361 1252 2212 1983 1728 1395 1506 1923 1199 2195 1663 92,900 118,300 78,600 147,800 119.4 119.6 121.7 119.0 61.9 61.5 65.6 67.3 109,120 120.5 65.6 indicated compressive strength are the size, shape and age of the be discussed in the following. specimens. Some of these factor (1) Type of cement: In Table 3 (Chapter II) it is reported that all but plants 1 and 3 use high-early strength cement. Plant 1

-82 - (product 1C2) and Plant 3 (product 3C2 and 3SG2) use regular cement. Since the age of samples is not reported, the effect of the type of cement on strength cannot be evaluated. (2) Crushing strength of aggregate: The effect of the crushing strength of the aggregates on the strength of the units cannot be evaluated because such tests were not conducted. However, it is believed that it does not have a marked effect at such comparatively low strength as exhibited by the units tested. It has been pointed out previously that some of the lightweight aggregate products have a higher strength than those made with dense aggregates. The lightweight aggregates tested are recognized to have lower crushing strength than dense aggregates. It was found possible by some researchers20 to make concrete from expanded shale having a strength (based on standard test cylinders) of 7000 psi. Compressive strength as high as 4000 psi (much higher than the present test values) is not uncommon with concrete made with lightweight aggregates. (3) Proportioning: Accurate study of the cement content of the plant mixes (Table 3) is not possibleo All mixes (except for products 1C2, 3C2 and 4SG2) are reported by weight and since the specific gravity of the aggregates varies widely, a comprehensive study on weight basis cannot be attained. However, rough analysis was employed by converting some of the mixes to volume using the available unit weights of the aggregates as determined by tests and reported in Table 4 (Chapter III). The conclusion on the basis of this rough analysis is that there is no correlation existing between the cement content and the strength of the product.

-83 - (4) Water-cement ratio: An accurate determination of the water-cement ratio of the plant mix is not possible because in most cases the amounts of water added vary with the moisture contents of the aggregates at the time of mixing. Many plants find it necessary to adjust the amount of water several times during the day. In many cases the consistency of the mix is measured by that required for molding by pressure and vibration and permitting subsequent stripping of the units without physical damage (see Chapter II). It may therefore be expected that the consistency varies slightly from one mix to another in the same plant and that the variations are even larger in the mixes of different plants. The effect of the water-cement ratio on strength cannot be over-stressed and is believed responsible for the observed variations in the strength of some units in the same product (see Table 43, Appendix IV). (5) Age of the specimen: The effect of the age of the specimen on the indicated strength is important. Most of the plants, however, do not specifically classify their products according to age and its effect is, therefore, undetermined in this study. The compressive strength of the units based on their net cross-sectional areas (Table 10, column 7) gives a better view of the strength variation of concrete in the units. However these strengths are not equivalent to those in design codes (or literature in general) which are usually determined by tests on standard 6 x 12 inch cylinders. Research indicates that the size and shape of specimens have an important effect on the indicated compressive strength. Correlation between the compressive strengths and the unit dry weight of the products is attempted but such relations were found

not to be existing. Correlation between the compressive strength and the flexural strength or the modulus of elasticity of concrete will be discussed later. The ASTM Designation C90-52 (Specifications for Hollow Load-Bearing Concrete Masonry Units)22 specifies for the compressive strength at the time of delivery to the site of the work to be as in Table 11. The compressive strengths in Table 11 are based on gross cross-sectional area of the unit. By comparing the average results of tests in Table 10 in terms of the values specified in Table 11, the following may be concluded: (1) Only four products have average strengths higher than 1000 psi (column 6). These are 2CES2, 2H2, 3SG2 and 9S2. The product 3SG2 has one individual unit showing a strength below 800 psi (see Table 43). (2) Three products display a relatively high result; these are 4SG2 (995 psi), 7BW2 (989 psi) and 10CA2 (906 psi). None of these showed a unit to have a strength lower than 700 psi. (3) The remaining products (cinder or expanded slag aggregates) have average results ranging between 877 psi (5CS2) and 647 psi (8CEL2). It may also be noticed that except for the latter, all of the remaining products averaged higher or very close to 700 psi. From Table 43 it may further be seen that of all these products investigated, only two showed individual units with strengths below 600 psi. These are 6ES2 (averaged 695 psi) with only one such unit and 8CEL2 (averaged 647 psi) with 3 such units.

-85 - TABLE 11. PHYSICAL REQUIREMENTS* Compressive Strength, Minimum Face min., psi Shell Thickness (Average Gross Area) in. Average of Individual 5_ Units Units 1.1/4 or over: Grade Aa 1000 800 Grade Bb 700 600 Under 1.1/4 and over 3/4 1000 800. -- _- _ —. —,,- _ --- -. --- —- -,-~. a. For use in exterior walls below grade and for protected exterior walls above grades that may be exposed to frost action. b. For general use above grades in walls not subjected to frost action or where protected from the weather with two coats of portland cement paint or other satisfactory waterproofing treatments approved by the purchaser. * Table 11 and its footnotes are from Table 1 of the ASTM Designation C90-52 (reference 22). It is of significance here to note that the actual bearing strength of a concrete masonry wall is lower than the compressive strength of its units, This was proven by the University of Illinois tests on the structural performance of concrete masonry walls as related to the strength of the unit and which was discussed previously in Chapter I. It was there concluded that the compressive strength of a wall panel depends on the compressive strength of the unit with an average ratio of approximately 0.53. The American Standard Building Code Requirements for Masonry (1954)29 specifies the allowable stresses of concrete masonry of hollow units to be 85, 75 or 70 psi (based on gross area) according to the type of mortar used. This permits a safety factor (considering also the University of Illinois

-86 -tests) of approximately 4.5 to 9.0. It is also indicated (see page 28 of reference 29) that the "compressive strength of masonry of block or tile depends importantly upon the completeness with which the units are bedded. For end-construction masonry, the bedding has ranged from covering the two face shells only, to complete coverage of the ends of all shells and webs.,...

CHAPTER VII FLEXURAL STRENGTH Concrete is much weaker in tension than it is in compression. The flexural strength of plain regular concrete beams is found to be approximately equal to 0.10 of its compressive strength. However, concrete made with some lightweight aggregates has produced flexural 21 strength as high as 0.25 of its compressive strength. Yet the flexural strength of concrete is always much lower than its compressive strength. Consequently, failure of plain concrete beams in cross-bending inevitably occurs in tension at the outer fibers. This is why, as was mentioned in Chapter VI, most methods of design for concrete members (slabs, beams, columns....) depend on concrete to carry only compressive stresses. However, in some concrete structures such as plain footing and especially roads and air field slabs, high flexural strength is considered imperative. Tensile stresses can develop also in masonry concrete walls by bending under transverse loads such as from wind or earth pressures. Masonry units in these walls will, therefore, be under such stresses. As will be discussed later, the allowable stresses for tension in the extreme fibers of masonry units has not yet been proposed by the common codes and specifications dealing with such materials. The resistance of unreinforced masonry walls to such stresses is usually neglected. Research has also proved that the strength of mortar is a very important factor in this resistance; this will also be discussed later. The flexural strength test of the masonry unit is important because it does give an indication of the tensile strength of concrete 87

-88 - and thus its resistance to cracking. It is also used as an additional measurement of the effect of repeated freezing and thawing on concrete. As in the compression test, the indicated flexural strength depends among other things on the type and size of the specimen and method of test; these will be discussed later. A. Test Specimen. Test beams of the approximate dimensions 2-1/2 x 3-5/8 x 15-5/8 inches were sawed off solid masonry units (slabs) of the modular dimensions 3-5/8 x 8-5/8 x 15-5/8 inches (see Chapter II). Three beams were obtained for this test from each product except 2H and 7BW. The products 2H (Haydite) and 7BW (Beslite and Waylite) were not available in slabs. B. Apparatus. The test apparatus is shown in Figures 15 and 17. It was constructed in the University of Michigan shops. It consists of a base and two support blocks. The base is made of an American standard channel, size 8 x 2-1/4 inches, 40 inches long. Mounted on the face of the channel are the two supporting blocks, each consisting of a base plate and edge support. The base plate has the dimensions 1 x 4 x 8 inches and is bolted to the face of the base channel. Additional bolt holes are provided in the face of the channel so that the span between the two supports is adjustable. Mounted on the base plate are the two edge supports which have the shape shown in Figures 15 and 17, thus providing a line support to the specimen. One of these edge supports is seated on a steel rod while the second is seated on a steel ball. This arrangement permits a free bending

BREAKING SECTION NO.1 AND t OF BEARING BLOCK HEAD OF COMPRESSION MACHINE LOAD PIECE ALONG WIDTH BEARING BLOCK SPHERICALLY SEATED Co \l i CONCRETE SPECIMEN I II e I 11 2x 3- x15 5 2 8 8 FIG. 15 - SCHEMATIC DIAGRAM OF APPARATUS FOR FLEXURAL TEST

-91 - of the specimen while keeping the supports parallel to the direction of application of the load. The loading block at the head of the machine is spherically seated. The load is applied through a steel half cylinder along the entire width of the specimen, thus providing a line bearing. C. Procedure of Test. Using the apparatus described above, the specimens were tested as simply supported beams. The mid-point loading method is used because it is more convenient for specimens of the size used for this test. The procedure of the ASTM Designation C293-54T (Tentative Method of Test for Flexural Strength of Concrete - Using Simple Beam with Center-Point Loading)2 was used. By using a span length of 8-1/2 inches, it was possible to break the specimen at two sections thus obtaining two test values for each specimen. This span was held constant throughout the program. The specimen was centered in the supporting blocks. The steel half cylinder was placed at midspan along the width of the specimen. The load applying block was then carefully brought to contact. The load was applied at the rate of 200 pounds per minute. This speed corresponds to an increase in the extreme fiber stresses of less than 150 psi per minute. The breaking force was recorded, then the second marked section of the specimen was centered and broken using the same procedure as above. The dimensions of the two breaking sections of the specimen were then measured and the average recorded. D. Calculations. Modulus of rupture is used as a convenient means of expressing the flexural strength of beams. It is the calculated tensile

-92 - stresses at the extreme fibers of the beam caused by the breaking load indicated by the machine. For simply supported beams broken by concentrated load at the center, the modulus of rupture can be approximated by equation (2) which is derived from the pure bending formula of equation (1). s= Mc (1) I 3 P L R= 2 (2) 2 b d Where, R = modulus of rupture (psi) P = the breaking load (pounds) L = span length in inch (equal to 8-1/2 inches, constant) b = width of the specimen at the breaking section (inch) d = depth of the specimen at the breaking section (inch). Equation (2) is derived by ignoring the weight of the specimen which is negligible. The assumption used and the derivation of equation (2) is presented in Appendix V. E. Discussion of Test Results. Calculations of the results are given in Table 12. The modulus of rupture of the two breaking sections of each specimen are calculated separately and reported in columns 8 and 14. Their average (the average result of each specimen) is given in column 15 of the table. While the average of the results of the three specimens tested from each product is also reported in column 16 of the table. The modulus of rupture, like the compressive strength, varies widely from one product to another and sometimes substantial

TABLE 12. MODULUS OF RUPTURE (R) CALCULATIONS 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Breaking Section No. 1 Breaking Section No. 2 Average Dimensions Dimensions Average Modulus Breaking of the Breaking of the R of Type Force Section 3 P L Force Section 3 P L 2 for the Rupture Sample of P Width Depth (L=8.5 inch 2bd R1 P Width Depth (L=8.5 inch 2bd RN2 Specimen for the No. Aggregate lbs. b d constant) psi lbs. b d constant) psi psi Product in. in. in. in. psi 1C41 Cinders 1C42 1C43 2CES41 Cinders & 2CES42 Expanded 2CES43 Slag 3C41 Cinders 3C42 3C43 3SG41 Sand & 3SG42 Gravel 3SG43 4SG41 Sand & 4SG42 Gravel 4SG43 5CS41 Cinders 5CS42 & Sand 5CS43 6ES41 Expanded 6ES42 Slag 6ES43 7W41 Waylite 7W42 7W43 8CEL41 Celocrete 8CEL42 8CEL43 9S41 Air-Cooled 9S42 Slag 9S43 10CA41 Cinders & 10CA42 Fly Ash 10CA43 460 515 485 610 410 610 720 580 565 1070 1075 1065 1155 1405 1010 620 655 660 430 705 670 530 540 670 310 470 280 1140 890 940 460 285 550 3.66 2.45 11,730 3.62 2.52 13,132 3.65 2.60 12,368 3.63 2.50 15,555 3.64 2.30 10,455 3.65 2.55 15,555 3.67 2.56 18,360 3.67 2.40 14,790 3.62 2.45 14,407 3.67 2.50 27,285 3.65 2.35 27,412 3.62 2.62 27,158 3.75 2.58 29,452 3.75 2.51 35,828 3.78 2.38 25,755 3.65 2.41 15,810 3.65 2.53 16,702 3.65 2.51 16,830 3.58 2.40 10,965 3.65 2.65 17,977 3.67 2.45 17,085 3.65 2.50 13,515 3.66 2.38 13,770 3.65 2.60 17,085 3.78 2.45 7,905 3.76 2.48 11,985 3.75 2.35 7,140 3.65 2.62 29,070 3.68 2.48 22,695 3.56 2.48 23,970 3.65 2.41 11,730 3.65 2.25 9,818 3.63 2.61 14,025 43.9 267 46.o 285 49.3 251 45.4 343 38.5 271 47.5 328 48.1 382 42.3 350 43.4 332 45.9 594 40.1 685 49.7 546 49.9 590 47.3 757 42.8 602 42.3 374 46.7 357 46.0 366 41.2 266 51.3 350 44.0 388 45.6 296 41.4 333 49.4 346 45.4 174 46.2 259 41.3 173 50.1 580 45.3 500 43.8 547 42.3 277 36.9 266 49.4 284 480 500 520 250 180 245 775 605 505 780 730 705 1195 1270 955 500 590 600 470 795 580 630 425 595 350 320 325 980 800 980 470 430 600 3.68 2.45 12,240 3.65 2.52 12,750 3.66 2.56 13,260 3.63 2.50 6,375 3.63 2.25 4,590 3.61 2.55 6,248 3.65 2.60 19,762 3.65 2.40 15,427 3.63 2.45 12,877 3.68 2.50 19,890 3.62 2.30 18,615 3.62 2.60 17,978 3.76 2.62 30,472 3.75 2.46 32,385 3.78 2.41 24,353 3.68 2.40 12,750 3.65 2.50 15,045 3.65 2.50 15,300 3.57 2.40 11,985 3.62 2.62 20,272 3.65 2.50 14,790 3.65 2.52 16,065 3.65 2.32 10,837 3.65 2.60 15,172 3.78 2.45 8,925 3.75 2.56 8,160 3.75 2.40 8,287 3.68 2.59 24,990 3.68 2;48 20,400 3.65 2.45 24,990 3.65 2.45 11,985 3.65 2.26 9,650 3.65 2.65 15,300 44.2 277 272 46.4 275 280 47.9 279 265 45.4 140 242 36.7 125 198 46.9 133 231 49.4 400 391 42.0 367 358 43.6 295 314 46.0 432 513 38.3 434 56 48.9 368 457 51.6 591 591 45.4 713 735 43.9 555 579 42.4 300 337 45.6 330 344 45.6 336 351 41.1 292 279 49.7 408 379 45.6 324 356 46.4 346 321 39.1 377 355 49.3 307 327 45.4 197 186 49.1 166 212 43.2 192 183 49.3 507 544 45.3 451 476 43.8 571 559 43.8 274 276 37.2 259 263 51.3 298 291 272 354 510 635 344 338 334 193 526 277

-94 - variations exist between specimens of the same product. This latter variation can probably be explained as in the compressive strength. However, flexural strength is recognized as being more sensitive to the character of the aggregates than is compressive strength. Further study of Table 12 reveals that: (1) Some variations exist in the results of the two breaking sections of the specimen. However, the products 2CES4 (cinders and expanded slag) and 3SG4 (sand and gravel) show larger variations. This is believed to be caused by variation in concrete quality and not by method or procedure of test, otherwise it should appear in a greater number of specimens. (2) Products made with dense aggregates have higher moduli of rupture than those made with lightweight aggregates. Average results for the products made with sand and gravel are 510 psi and 635 psi for 3SG4 and 4SG4 respectively, while those made with air-cooled slag (9S4) averaged 526 psi. The maximum result for lightweight aggregate products is 354 psi by 3C4 (cinders). (3) The modulus of rupture of the products made with lightweight aggregates averaged between 354 psi for 3C4 (cinders) and 193 psi for 8CEL4 (Celocrete). However, four of these products have higher results than 330 psi. These are 3C4, 5CS4, 6ES4 and 7W4. The rest are lower than 300 psi. Very poor correlation exists between the flexural strength and the compressive strength of the products, as can be seen in Table 13. The compressive strengths are the average results of tests on hollow masonry units based on net areas and were originally reported in Table 10. Moduli of rupture are taken from column 16 of Table 12.

- 59> This lack of correlation can be seen also from column 5 of Table 13 where the ratio of the modulus of rupture and the compressive strength for each product is given. TABLE 13. FLEXURAL STRENGTH OF CONCRETE - ANALYSIS OF THE RESULT 1 2 Sample No. 1C4 2CES4 3C4 3SG4 4SG4 5cs4 6Es4 7W4 8CEL4 9S4 10CA4 Dry Weight p.c*f. 90.9 93.1 83.4 132.8 129.2 97.1 101.3 97.3 76.2 126.3 85.2 3 Compressive Strength psi 1357 1778 1252 2212 1983 1728 1395 1506 1199 2195 1663 4 Modulus of. - Rupture psi 272 224 354 510 635 344 338 334 193 526 277 Mod. of Rupture -Comp. Strength Ratio.20.13.38.23.32.20.24.22.16.24.17 However, rough correlation exists between the moduli of rupture and the unit dry weights of the products. This can be seen from Table 13 or Figure 18. The unit dry weights are the average results of tests conducted on hollow masonry units and originally reported in Table 8. Comparison of the results with those reported by other investigators is difficult because of the wide variation in the results

-96 - 800 LU) (0 a. -J a. 0 LL 0::):3 0 5 600 400 200 --- __- =- _ ---. I0 I 81 - -i^ -- -— T T2 ** - ~v - - - - - - - - - - - - - - - - - - 0 70 90 110 130 PCF 150 UNIT DRY WEIGHT LEGEND I 2 30 3S 4 5 IC4 20E ES4 304 3SG4 4SG4 5CS4 6 = 6ES4 7 = 7W 8 = 8 EL4 9 = 9S4 10 = 10CA4 OF RUPTURE FIG. 18-MODULUS VS. UNIT DRY WEIGHT

-97 - and their significance. This is probably because the results are easily affected by the method and rate of application of load, dimensions of the specimen, its temperature and moisture condition. Many theories were advanced to explain and account for the effect of these and other factors on the indicated flexural strength of concrete; but there is no general agreement as to which of these theories is the most valid. The most important of these are the works of Weibull and later of Tucker on the "weakest link" theory, the "strength summation" theory of Tucker30 and the works of Reagel and Willis33, Wright and Garwood31 and most recently, Nielson.32 The weakest link theory assumes that the concrete beam is formed of intersecting links (sections) along the length of the beam and are probably of various strengths. The failure does not therefore necessarily occur in the section of highest stress but might occur in the weakest section, although the stress might be lower in this section. The strength summation theory assumes that the strength of a specimen is equal to the sum of the strengths contributed by the component parts or elements. This theory has a different application as we shall see later. Some of the factors affecting the results of test will be discussed in the following: (1) Method and rate of application of the load: Research proved that mid-point loading gives higher results than third-point loading (ASTM Designation C78-49: Method of Test for Flexural Strength of Concrete - Using Simple Beam with Third-Point Loading), but the result of the latter method is more uniform. This was explained statistically by the weakest link theory as being due to the fact that in the third-point loading larger portions of the beam (the entire

-98 - middle third) is under maximum stress while in center-point loading, only the middle section is under maximum stress. The modulus of rupture has been found to vary directly with the rate of applying the load. In conducting the present tests, the maximum rate of application of the load was about 200 pounds per minute corresponding to an increase in stress at the extreme fiber of about 150 psi per minute which is in accordance with the ASTM method. 31 Wright suggested 100 psi per minute for standardization in Britain. (2) Size of the specimen: It was found by Tucker, Wright, Garwood and others that the modulus of rupture decreases with the increasing length or depth of the specimen. This was also explained by the weakest link theory by assuming that the beam is formed of intersecting layers or fibers and that the failure occurs in the weakest fiber and not necessarily in the outer fiber as it is usually assumed. Tucker also showed that the increase in the width of the specimen does not affect the modulus of rupture numerically. However, the dispersion of the results decreases with increasing width. This was explained by Tucker's strength summation theory. (3) Moisture condition of the specimen: The moisture condition of the specimen affects the result of test. When the outer layers of the specimen dry out, tensile stresses set up in these layers reduce the indicated modulus of rupture. This effect varies with the 32 depth of the specimen. Nielson concludes that this might explain the decrease in modulus of rupture with increasing depth and not the weakest link theory discussed before. Concrete design codes, and specifications in general, contain no proposal for the allowable tensile stresses in extreme fibers for

concrete made with lightweight aggregates. However, most codes propose values for regular concrete, usually equal to three per cent (0.03) of the compressive strength. Table 13 shows this ratio to be between.13 and.38 with most of the products showing over 0.2. Notwithstanding the many factors affecting both types of tests, the results imply a factor of safety of from 4 to 9 on the usual 0.03 ratio. Concrete masonry codes, as previously stated, do not propose values for allowable stresses for tension in the extreme fibers of the masonry unit because agreement has not yet been reached as to the appropriate values under various conditions of loading. 29 It is also repeated that the flexural strength of unreinforced masonry walls depends largely on the strength of the mortar. This was shown by the University of Illinois testl0 (see also Chapter I) on full size wall panels which were supported laterally at top and bottom, providing a 9 foot span and loaded laterally above a horizontal center line. Almost all of these walls failed in a mortar joint near mid-height of the wall, usually along a horizontal joint with most of the failure in the adhesion of the mortar to the unit. Attention is invited to the fact that many specifications now require the use of reinforcement in the mortar joint to alleviate this deficiency.

CHAPTER VIII ELASTIC PROPERTIES Concrete is considered elastic within certain limits; that is, when subjected to the action of external forces it will suffer a deformation which will disappear upon the removal of these forces. It is also considered, by the designer, as elastically homogeneous and isotropic. Actually, the above are only approximations. Since concrete is a coalescence of materials, its elastic properties depend on the elastic properties of the individual components and their proportions. The elastic properties also depend on the age of concrete, its moisture and temperature conditions and type of loading. Nevertheless, tests have justified these approximate assumptions thus permitting the application of the theory of elasticity in concrete designo Elastic properties of concrete are measured by its Young's modulus of elasticity (E), its modulus of rigidity in shear (G) and its Poisson's ratio (M). Young's modulus of elasticity is based on Hooke's law which states that stress is proportional to strain within the elastic limit. The modulus E is this proportionality factor and therefore is defined as the ratio of unit stress to corresponding unit strain. The modulus of elasticity in shear, G (sometimes called the "modulus of rigidity"), is defined as the ratio of unit shear stress to corresponding unit shear strain of a body subjected to pure shear stresses. 100

Poisson's ratio (,) of isotropic materials is the ratio of lateral strain to the corresponding axial strain. For concrete these elastic constants can be determined experimentally. Two general methods are used: (1) the static method, and (2) the dynamic method. The results of the two differ somewhat for reasons that will be discussed later. In the static method, Young's modulus of elasticity (E) is usually determined on prismatical or cylindrical specimens which are subjected to known compressive, tensile or flexural stresses; the corresponding strains are then measured by suitable mechanical or electrical strain gages. The modulus of elasticity in shear (G) can be determined statically by applying a known torque to a specimen of suitable dimensions (usually cylindrical) thus subjecting the specimen to "pure" shearing stresses; the corresponding shear strain is then measured as the angle of twist developed over a certain gage length. Poisson's ratio can be determined statically by measuring the axial and lateral strains caused by a certain compressive or tensile stress. Sensitive gages are needed for measuring the lateral strain because of its small order of magnitude. A suitable mechanical-optical system or electrical variable resistance strain gage usually possesses sufficient accuracy. In the last decade a special nondestructive dynamic method of determining the elastic constant was developed and used extensively. In this so-called "sonic" method (sometimes called the "resonant" method) the specimen (cylinder or prism of suitable dimensions) is vibrated in free-free fundamental mode; transverse (flexural) vibration is used for determining E, while torsional vibration is used for

-1G2 - determining G. Poisson's ratio (p) can then be calculated from the interrelation of E and G. Accurate determination of the elastic constants for concrete, although they are affected by many factors as has been previously mentioned, is desirable: Young's modulus of elasticity (E) is necessary for designing plain and reinforced concrete and for predicting the deformations of concrete members under certain loadings, or for the calculation of stresses caused by volume change of concrete by variation in temperature and moisture. The modulus of elasticity in shear (G) is important for the design of concrete members subjected to torsion such as those of some statically indeterminate structures. Poisson's ratio (,u) also enters into the design of some structures (domes, skewarches....) or for interpreting strain data of concrete. As will be seen later, dynamic methods might not be suitable for the determination of absolute values of the elastic constants used for calculation of static cases such as discussed above. However, since the dynamic test is not destructive, it serves as a check on the values of the static tests and is excellent for detecting the rate of change of concrete quality as affected by certain other treatments (freezing and thawing, wetting and drying....). This chapter will present the elastic constants E, G and X determined by both the static and dynamic methods. A. Test Specimen. Since the dynamic (sonic) test is not destructive and to facilitate the analysis of the results, it was decided to use the same specimen for both static and dynamic tests. However, in order to obtain specimens of suitable dimensions for each method, the following procedure was used.

-103 - The dynamic test was conducted first on specimens of the approximate dimensions of 2-1/2 x 3-5/8 x 15-5/8 inches which were cut from solid masonry units (slabs) of the modular dimensions of 3-5/8 x 7-5/8 x 15-5/8 inches. (See Chapter II.) Two specimens were tested from each product. Upon the completion of the dynamic test, the specimens were cut in half and one-half of each specimen (approximate dimensions of 2-1/2 x 3-5/8 x 7-1/4 inches) was selected for the static test. Some of the remaining halves were also used as inactive specimens (dummies) in this latter test as will be discussed later. B. Dynamic (Sonic) Method. As has been said previously, E and G can be determined dynamically by determining the resonant frequencies of, respectively, transverse (flexural) and torsional vibrations of a specimen vibrating in free-free fundamental mode. Theory leading to this method will be discussed briefly in the following: 1. Theoretical Discussion. a) Young's Modulus of Elasticity. (E) When a prismatic bar is vibrating in free-free transverse (flexural) mode of vibration, the elementary form of the differential equation of vibration (references 35, 36) is as follows: Er2 a4v a2v + + o ~ (3) p x at2 Where: x is a member of mutally perpendicular coordinates (x, y, z) at the center of the cross section of the bar; y is in the direction of the vibration while x is in the direction of the length; v is the deflection (displacement) of the bar in the

direction of y axis; r is the radius of gyration of the crosssection with respect to z axis; t = time; p = mass per unit volume. Equation (3) does not include the effect of rotary or lateral inertias or of shear. More accurate equations have been presented by Rayleigh, Love and Mason in which corrections were made for the effect of rotary or lateral inertia. However, Timoshenko's equation accounts for the effect of rotary moment of inertia and shear (see reference 35). This equation is as follows:36 Er2 a4v a2v 2 E a 4v rp p + at x2r ( 1t2 G) KG 0 (4) For which, in addition to the notations given before: K = constant introduced by Timoshenko to correct for the effect of shear on the elastic line, defined as the ratio of average unit shear across a section to the unit shear at the neutral axis. Equation (4) does not correct for the lateral inertia which was proven to be small. The general solution of this differential equation for the problem under consideration, substituting the boundary conditions (both ends are free), can be reduced to the following form:36 M N, - + 0O. (5) tan h a tan P 2 2 Equation (5) called frequency equation, applies to the first, third, fifth, etc., modes of vibration. Goens presented

-10l5 - solutions in which a and p are determined by satisfying the differential equation (4) above. The values a, P, M, and N are given in r terms of, r, K and k. L Where: L is the length of the specimen; i is Poisson's ratio; K is the shear constant defined above; k is as follows: 2 -4 2 1/4 k (4 P L n (6) r2 E or 4 2 p L n E = 2 4 (6a) rk or 4 rL2 L3 W n2 E =k4 (6b) I g or E = C W n2 (7) in which 4,2 L3 c 4 (8) I k g Where in addition to the notations previously given: W = weight of the specimen n = resonant frequency, fundamental mode of transverse vibration g = acceleration of gravity (386.4 inches per second squared) I = moment of inertia of the cross-section with respect to the z axis.

-106 - Equation (7) will be used for calculating E after determining (n) experimentally. The coefficient C is a dimensional factor r which can be calculated, for any specimen of any L value, from equation (8) by finding the value of k that satisfies equation (5). Approximate values of the Poisson's ratio (At) and the shear constant (K) can be 1 assumed for this calculation. For concrete, p = - is usually assumed 6 and, as it will be shown later, changes in the value of ti do not affect C materially. The shear constant of a rectangular section is assumed equal to 2by Timoshenko and 5 by Goens. Pickett36, who solved the 3 6 frequency equation (5) by satisfying the boundary conditions in the mathematical equations of elasticity, found accurate values for K by selecting the empirical values that give results most nearly in accord when solved by Goens' equation or the mathematical theory of 8 elasticity. He concluded that K = is the best value to use for a 1 prism when K = Goens' solution for C is as follows: Suppose at the limit r as L approaches zero, k approaches m, and a = P = k, and if both ends are free, M = No For this case, equations (5) and (8) approach equations (5a) and (8a): m m tan 2 + tan h - = (5a) 2 2 4 2 L3 C = (8a) g I m Where C' is the limiting value of C and m is the limiting value of k.

-107 - Now setting C = C' T (9) Where C is defined by equation (8) then k T = m (10) T is called the correction factor and was introduced by Goens. Equation (5a) is the solution (frequency equation) of the elementary equation (3). The value of m, computed from equation (5a) by satisfying equation (3) is equal to 4.730 for the first mode. Substituting this value of m in equation (8a), the value of C can be r found to equal the following f6r any value of -: L3 T C = 0o0002044 -i (for a prism). However, the coefficient C should be corrected by multiplying by the factor T (equation 9) to satisfy the (adopted) differential equation (4)~ The values of the correction factor T should, therefore, r be obtained for each value of L. Goens arrived at an approximate r equation for T in terms of L, and K by expanding the frequency equation obtained by his solution in a Taylor series and retaining 8 1 only the first few terms. Using K = - and p = 6 Goens' equation for the first mode (reference 36))becomes~ 1314 (L4 (r 2r2 T = 1 + 81.79 () + 81.09(r (11) Equation (11) is used largely for the calculation of the correction factor because it involves less computation than other

-108 - methods. Pickett36 found that values of T calculated by equation (11) agree closely with those obtained from the original frequency equation from which the Taylor's expansion was obtained. This agreement was r almost exact for values of - not higher than 0.3. Since, as is apparent from Table 45 (Appendix VI), the specimens used for this test have - ratios much lower than 0.3, it was decided to use Goens' equaL tion. However, the actual calculations of T, C' and C were not necessary. Pickett has calculated the coefficient C using Goens' equation for prismatic and cylindrical specimens of different sizes. He also plotted the results versus the ratio - of the specimen (y here being L the depth of the specimen in the driven direction). Pickett's curve was, therefore, used for the calculation of C. This will also be discussed with details later. b) Modulus of Elasticity in Shear (G). Equations for the calculation of (G) from the resonant frequency of the free-free funda36, 22 mental mode of torsional vibration are as follows: G = B W (n')2 (12) Where: G and W are as defined before n' = torsional frequency 4 LR B= gA (13) Where: L and g are as defined before A = gross sectional area of the specimen

-109 - R = -ratio of the polar moment of inertia to the shape factor of the specimen. This ratio is unity for cylinders and equal to the following for a specimen with rectangular cross section. a b R = b a (14) 4 - 2.52 () + 0.21 (a) Where: a and b are the dimensions of the cross section, with a < b. Equations (12) and (13) for shear modulus of elasticity are analogous to equations (7) and (8) for Youngs's modulus of elasticity. However, the nature of the factor (R) will be indicated in the following: When a prismatical bar is twisted, its section does not remain plane but warps and radial lines through the center do not remain straight. The distribution of the shear stress is therefore not linear. The ratio (R) was introduced to correct for the effect of warping. c) Poisson's Ratio (u). Poisson's ratio of homogeneous and isotropic materials can be calculated from the following equation: c = E - 1 (15) 2G where E and G are determined by the dynamic method as discussed previously. However, it should be kept in mind that concrete is never perfectly elastic, homogeneous or isotropic, and equation (15) will only give approximate results when applied to concrete. The significance of this will be discussed later.

-110 - 2. Test Apparatus. The apparatus used for determining the resonant frequency of the specimen in transverse and torsional vibrations is shown in Figures 19 and 20. It consists mainly of three parts: the driver circuit, the pickup circuit and the specimen support. These components, except for the oscillograph, were assembled by "Electro Products Laboratories". The oscillograph is manufactured by Allen B. Du Mont Laboratorie s. a) The Driver Circuit. The driver circuit consists mainly of (1) variable frequency audio oscillator, (2) amplifier, and (3) driver -unit. The audio oscillator, power and pickup amplifier and the indicator are housed in the "'sonometer" cabinet (Figure 19 at left). The frequency of the oscillator can be varied between 20 cycles to 20 kilocycles per second with approximately 2 per cent accuracy of operation. The oscillator is coupled directly to the power amplifier which in turn is coupled to the driving unit. The amplifier can deliver up to 18 watts of controlled power (controlled by variable resistor) to the driving unit. This power is used to actuate the driver hammer which vibrates the specimen. The hammer is small in mass when compared to the specimen so that its resonant frequency is different than that of the specimen. b) The Pickup Circuit; The pickup circuit consists of (1) pickup unit, (2) amplifier, (3) resonance indicator, and (4) oscillograph. Motion generates voltage in the pickup unit which is carried to the pickup amplifier and then to the resonance indicator

-111 - (both housed in the sonometer cabinet). Similarly, the pickup needle has much smaller mass in comparison with that of the specimen. Output of the amplifier is also controlled by a variable resistor before it is delivered to the resonance indicator and the oscillograph. The resonance indicator is a milliameter with a range from 0.0 to 1.0. The cathode-ray oscillograph37 is used as an auxiliary instrument to facilitate the identification of the modes of vibration and resonant frequency. This instrument indicates visually the essential characteristics of the signal received from the amplifier. This is done in Lissajou's figures that are traced on a screen by feeding the known sinusoidal signals of the driver into the horizontal channel of the instrument and the unknown signals of the pickup into the vertical channel. The ratio of the unknown frequency to the known frequency is the number of loops (of the Lissajou's figure) cutting the vertical axis to those cutting the horizontal axis of the screen. This ratio also indicates the mode of vibration. The ratio 1:1, appearing on the screen as a circle or ellipse, indicates a fundamental vibration which is used in this test. The type of vibration (flexural, tor" sional....) can also be checked by determining the phase difference of the driver and the pickup signal and the nodal points of vibration. This will be discussed later. c) Specimen Support. The specimen support consists of a heavy concrete block base and thick rubber pads. The concrete base supports, beside the specimen, the driving and pickup units. Its mass is much larger than any of the specimes tested. The three rubber pads are put between the specimen and the concrete base. This

-112 - thick rubber support permitted the specimen to vibrate in a free-free condition without restrictions. 3. Procedure of Test. Before test, dimensions of the specimens were determined at several sections and their averages were recorded. Also the apparatus was turned on to warm up for 30 minutes before tests were started. While the apparatus warmed, the specimen was weighed and placed on the rubber pads so that it was perpendicular to and almost touching the driver at one end. When the apparatus was ready the right end of the vertical face of the specimen was brought into contact with the driver at a point located at about one inch from the end and thus away from the nodal points. The pickup needle was then brought into contact with the left end of the center line of the upper face of the specimen. The specimen was then forced to vibrate transversely at a slowly increasing frequency delivered to the driver which was at first pulsating at low frequency. Meanwhile the motion was being studied by observing the resonance indicator and the oscillograph screen which were both brought to a varying "operating level" by the signal received continually from the oscillator and pickup units. When the (fundamental) resonant frequency was reached the resonance indicator responded to the highest reading and the Lissajou's figure on the oscillograph screen formed an ellipse with a well defined peak. The maximum indications occur at resonance frequency because at this frequency the signal received from the pickup unit is maximum; this is because the electrical output of the pickup is proportional to its mechanical input which is determined by the motion of the specimen; at the resonance frequency the amplitude of vibration is,

-113 - Z i: ,! : I I Figure 19. Dynamic (Sonic) Test Apparatus for Determining the Elastic Constants.

OSCILLATOR POWER AMPLIFIE I I I i,, DRIVER SONOMETERG CABINET I 'R = _......... -....___ ______HAMM - I I I X-AXIS 1 L 7 AIOSCILLOGRAPH I Y-AX 1 I SPICKUP CATORI I ____ __ __ __ IER -"orcvltN M 1 I (CRVY~STAL) 1 — (CRYSTAL) I RESONANCE INDIA PICK-UP AMPLIFIER FIG. 20 SCHEMATIC DIAGRAM OF THE SONIC APPARATUS

-115 - of course, maximum. The driving force was held constant after determining the resonant frequency, while the phase difference of the driver and the pickup frequencies and the nodal points were being checked to insure that the motion was a fundamental flexural vibration. The nodal points and the phase difference were determined by bringing the pickup to contact with the key points along the entire length of the specimen (detailed procedure is given in Appendix VI). The resonant frequency was then recorded directly from the frequency dial of the apparatus which is marked in cycles per second. Returning the pickup unit to its first position (at the left end of the specimen) the frequency of the driver was again increased slowly, this time for determining the resonant frequency of the torsional vibration of the specimen. The procedure is generally the same as described above except this time the specimen was forced to vibrate in torsion. When the resonant frequency was determined the motion was also checked in the same manner used for flexural vibration, by bringing the pickup to contact with the key points of the fundamental torsional vibration along the entire length of the specimen. More discussion of the operation is in Appendix VI. The specimen was first tested by driving in the direction of the larger side of its section. After several determinations were made for flexural and torsional frequencies with the speciman in this position, it was turned 900 so that the driving was this time in the direction of the shorter side of the section and tests were repeated. By this procedure it was possible to check results of the test and arrive at a more accurate average. For flexural vibration the two resonant frequencies of the specimen obtained by

the two tests described above were different because the flexural stiffness of a bar is proportional with the depth. Young's modulus of elasticity calculated from the two frequencies, however, are theoretically equal and the experimental values must be close (see Table 14). For the torsional vibration a quick check was possible because the resonant frequencies obtained from the two tests are theoretically equal as were the experimental values, as may be seen in Table 47 in Appendix VI. 4. Calculations. Calculations of Young's moduli of elasticity and the moduli of elasticity in shear for the specimens are given in Tables 14 and 15 (text) and Tables 44 to 49 (Appendix VI). These calculations will be discussed in the following: a) Calculation of Young's Modulus of Elasticity. Young's moduli of elasticity of the specimens were calculated from the previously given equation (7) which is again as follows: E = C W (n)2 (psi) Where the dimensional factor C (sec2/in2) has been defined by equation (8), W is the weight of the specimen in pounds and n is the resonant frequency of the flexural vibration (cycles per second). As was previously discussed, values of C for the specimen in this test were taken from Pickett's curve. The portion of the curve that was used in these calculations was replotted in Figure 21. In this figure, values of C x z are plotted versus the corresponding values of - where z is the width of the specimen and y is the depth of the specimen during vibration (i.e., y is in the direction of the

1.0 0.9 0.8 0.7 I 0.6 0 Z 0.5 o 0.4 bJ V~) cn s 0.3 cn 0.2 0 IL _;~~~~~~~~~~~~.......... r:.- — = ---- -- -- -- - ---.. -—. --- — ~-~c --- —-~~~~~.~ ---t ~~ _ -- - _.: _.-.-;.~..... _ —..._........._ _ _ _ _L_ __..__ ___...................... _ __.._... _........ _ _,._..._, _.......... _. '.. I -- - '.. ---- --- ' -' ---... --- - F ---- - —:.-.. I __ --- _ —,. _._ l -- I ___ -..... -..,_._ - -. ) -- I -.. I-.. ".. -. - -- - - - - L ---T-C1 -.-.- ---i.-. —I~ — - -~I ---- ~ -, —, ---... --- _......... -- _ i r.. _, _. _....... - I:- -- ~t-" - _._. _ __ __ _. _. _ --- -1 1=c777. -: 1 - --- ~ ~ -.. —.:- _- _ _ _.:..... - -.._.. - ' — t-I I H _n N X i 0., F - I, i ~t1i-. I I I I _, _ I 1 — - i i |I til l i I I I I I I I I I 1 1 1 I I I I I I i lI I I I i, t! w * I 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 RATIO OF LENGTH TO DEPTH (-) y = DEPTH OF BEAM IN DIRECTION OF DRIVING z= WIDTH OF BEAM FIG. 21- CURVE FOR THE COEFFICIENT C, CALCULATED BY PICKETT FOR U = 6

-118 - drive). It must be noted that Pickett's curve is based on Poisson's ratio (l) equal to - which is used commonly for concrete. This value 6 of p differs from those obtained for some of the specimens: however, actual calculations showed that these variations in the values of p. have little effect in the values of C; it was therefore decided to use Pickett's curve without correcting for the variation in.. Dimensions of the specimens (a, b, and L) and the ratio L a and b are given in Table 44. Values of C x z corresponding to these ratios are given in Table 46 (Appendix VI). Two sets of E calculations were made for each specimen corresponding to vibrations in b and a directions respectively. Resonant frequencies (n) of the specimen and their squared values for both directions of the vibration are given in Table 47 while weights (W) of the specimens are given in Table 45 (both tables are in Appendix VI). Final calculations for equation (7) were made in Table 14. Young's modulus of elasticity, when the specimen is driven in b direction, is called b, and Ea when the specimen is driven in the direction of a. Their average E (column 9 of Table 14) is considered as representing Young's modulus (dynamic) for the specimen tested. b) Calculation for Modulus of Elasticity in Shear. Equations for the calculation of the dynamic modulus of elasticity in shear were previously discussed on pages 106 and 107. These are equations (12), (13), and (14) which will be repeated here for convenience: G = B W (n' ) (12) 4LR B (13) g A

-119L and a b + - b * a R a a2 a6 (14) 4 - 2.52 (!) + 0.21 (b) b b and as previously given: W = weight in pounds, n' is the torsional resonant frequency in cycles per second, g is acceleration of gravity (386.4 in. per sec2), a and b are sectional dimensions in inches (b larger than a), A = a x b, in2 and L is the length of the specimen in inches. Calculations of R are started in Table 44 where a, b, and the a b ratios and a- are given and are completed in Table 48. In these b a calculations the last term of the denominator in equation (14) is omitted because its effect is negligible. Coefficients B for all the specimens were calculated, by equation (13), in Table 49. Resonant frequencies (n') are given in Table 47, columns 6 and 7 for the two directions of drive, a and b. It may be noticed that the values in columns 6 and 7 of each specimen are equal (see page 255)Final calculations of (G) by equation (12) are given in Table 15. c) Calculations of Poisson's Ratio. Poisson's ratios are calculated from previously given equation (15) which is as follows: E 1 = - 1 2G These calculations are given in Table 15, Values of the modulus E for these calculations are taken from Table 14, column 9.

-120 - TABLE 14. CALCULATIONS FOR YOUNG'S MODULUS OF ELASTICITY (DYNAMIC) 1.2 3 5 6 7 9 Drive in b direction Drive in a direction Weight n2 C n 2 C Average Specimen W (Cycle 2 Sec2 Eb (Cycle2 Sec2 Ea E No. Pounds \ Sec in2 ksi Sec i 7.- ksi ksi 1C51 1C52 6.84 6.70 1,404,000 1,440,000.106.107 1014 1033 714,000 810,000.187.193 912 lo46 963 1040 2CES51 7.39 1,626,000 2CES52 6.92 1,690,000.105 1263.113 1323 865,000 828,000.186 1191.225 1288 1227 1305 3C51 3C52 6.24 6.31 1,346,000 1,428,000.115.113 966 1016 706,000 706,000.226.215 995 957 981 986 3SG51 3SG52 10.63 2,592,000 9.76 2,723,000.108 2962 1,440,000.114 3034 1,276,000.190 2910.223 2778 2936 2906 4SG51 4SG52 5CS51 5CS52 11.03 3,610,000 10.85 3,276,000 7.83 1,769,000 7.98 2,074,000.102 4077.100 3568.107 1479.104 1726 1,742,000 1,513,000 960,000 1,188,000.203 3896.192 3152.199 1498.187 1771 3987 3360 1489 1748 6ES51 6ES52 8.38 2,045,000 8.21 1,960,000.107 1830 1,103,000.107 1727 1,082,000.198 1827.201 1780 1829 1754 7W51 7W52 8.64 7.74 2,624,000 2,592,000.105 2374.115 2315 1,513,000 1,416,oo000.182.236 2376 2581 2375 2448 8CEL51 8CEL52 9S51 9S52 6.95 6.37 1,960,000 1,823,000.102 1389.108 1255.109 2558.110 2711 1,000,000 846, ooo 1,254,000 1,346,000.197 1371.229 1237.201 2404.207 2574 1380 1246 2481 2643 9.53 2,465,000 9.26 2,657,000 10CA51 10CA52 6.-92 1,513,000 6.62 1,346,ooo.112 1173.114 1015 764,000 723,000.218.227 1150 1087 1161 1051 - -; - --- - - - -- - -- - - - --- - --

-121 - TABLE 15. CALCULATIONS FOR MODULUS OF ELASTICITY AND POISSON'S RATIO,, (DYNAMIC) IN SHEAR, G. Weight Specimen W No. Pounds B2."in/ G Eaer E E ksi ksi 2G 2G 1 = 1C51 1C52 2CES51 2CES52 3C51 3C52 6.84 2,560,000 6.70 2,756,000 7.39 2,890,000 6.92 2,89o,000 6.24 2,372,000 6.31 2,496,000.0236 413.0242 447.0237 506..0267 534.0266 394.0257 405 963 1.16 104o 1.16 1227 1.21 1305 1,22 981 1.25 986 1.22.16.16.21.22.25.22 3SG51 3SG52 4SG51 4SG52 5CS51 5CS52 6ES51 6ES52 10.63 9.76 4,928,000 4,840,000.0238.0263 1247 1273 11.03 5,808,000 10.85 5,290,000.0249 1595.0241 1383 2936 2906 3987 3360 1489 1748 1.18 1.14 1.25 1.22 1.17 1.17.18.14.25.22.17.17.16.16 7.83 7.98 3,312,000 3,960,000.0246.0237 638 749 8.38 3,842,000 8.21 3,725,000.0246 792 1829 1.16.0248 758 1754 1.16 7W51 7W52 8.64 7.74 4,973,000 4,601,000.0233 1001 2375 1.19.0273 972 2449 1.26.0246 604 1380 1.14.0272 513 1246 1.21 8CEL51 6.95 3,534,000 8CEL52 6.37 2,958,000 9S51 9.53 4,494,000 9S52 9.26 4,731,000.19.26.14.21.17.19.15.16.0247.0253.0262.0267 1058 1108 2481 2643 1.17 1.19 1.15 1.16 10CA51 6.92 2,789,000 10CA52 6.62 2,560,000 506 1161 452 1051

-122 - C. Static Method. Elastic constants were also determined statically by subjecting the specimen to known compressive stresses and measuring the corresponding axial (and for some specimens also lateral) strains by resistance wire strain gages. The elastic constants were calculated from these measurements. 1. Test Apparatus* In addition to the compression machine, the test apparatus consisted of the gages, strain indicator and a switch box. The test machine can be operated to hold the load constant while the strain is determined. a) Strain Gages. The variable resistance wire strain gage known as Baldwin SR-4 Strain Gage was used for this test. These gages are made of different patterns, sizes and materials, but type A-1 was used because of suitable size and adequate sensitivity. It is flat grid of thin wire mounted on paper base and covered with felt. This gage has a nominal gage length (length of grid) equal to 1 inch and grid width of - inch. The gage length was adequate since (see 64 Table 4, Chapter III) maximum sizes of aggregate were usually 3 inch. 8 The grid wire is constantan (copper-nickel alloy) which has a gage factor (ratio of unit resistance change, ohms per ohm, to unit strain change inch per inch in the wire) of 2.07 and a resistance of 120 ohms. The operation of the SR-4 gage was as follows: The gage was cemented firmly to the specimen (methods will be discussed later), direct electric current was passed through the gage, known compressive force was then applied, causing the specimen and the gage to

-123 - undergo a certain strain. Change of the electrical resistance of the wire is proportional to the strain and was measured by a speciallydesigned Wheatstone bridge. During the test a specimen may suffer an additional strain caused by temperature or moisture changes. To compensate for these unknown additional strains a similar gage attached to a duplicate specimen was added to the circuit. This so called "dummy" specimen was then placed as close as possible to the tested "active" specimen so it would be under the same atmospheric conditions but not subjected to statical stresses. b) Strain Indicator. A specially designed Wheatstone bridge (Baldwin portable strain indicator, type MB, Figure 22 at middle) was used for this test. The instrument is calibrated to read directly in micro-inches per inch, with a measuring range of 2000 mic in per in. A ten-step switch extends the range another 1000 mic in, per in. for each step. A simplified Wheatstone bridge circuit is shown in Figure 24. c) Switch Box. A switch box was used as auxiliary equipment to facilitate the selection of gages during the test. The available instrument was Baldwin's SR-4 twenty point switching unit which was used although no more than two active gages were recorded during the test. This instrument (Figure 22, at left) has 20 poles for active gages and the same number of compensating gages. It is switched manually and provides no means of initially balancing the individual gages.

-124 - Figure 22. Static Test Apparatus for Determining the Elastic Constants (Close-up). Figure 23. Static Test Apparatus (General View).

-125 - A l.] emf Ill ~o -R2 0 - A= ACTIVE GAGE C= COMPENSATING GAGE I= STRAIN INDICATOR R 8 R2 =RESISTORS emf = BATTERY (D.C.) FIG. 24 SCHEMATIC DIAGRAM OF SIMPLIFIED WHEATSTONE BRIDGE CIRCUIT

-126 - 2. Preliminary Investigations. Because of the cellular characteristic of lightweight aggregate products, further preparation of the surfaces of the specimen before cementing the gages was necessary. Preliminary investigations were conducted on companion specimens to those used for the final tests to develop a suitable method, Specimens were taken from three different products: sand and gravel (3SG5), air-cooled slag (9S5), and cinders plus expanded slag (2CES5). Three methods of surface treatment were investigated, filling the surface cavities with (1) Duco cement, (2) plaster of Paris and (3) neat portland cement paste or sand-cement mortar. In the first method surface cavities were filled by applying several layers of thinned Duco cement. Each layer consists of applying a liberal amount of this cement and leaving-i-t-to dry before the succeeding layer is added. For specimen 2CES5 as many as eight layers were used with this method. In the second method a small quantity of slightly thinned paste of plaster of Paris was pressed into the cavities and the excess paste removed. The specimen was aged for 24 hours before grinding the surface carefully so that all the plaster was removed except that which was in the cavities. In the third method the cavities were filled with neat Portland cement paste on one face of the specimen and with sand-cement mortar on the second face. This was done by the same procedure used in the second method, including the curing for 24 hours before grinding the surface.

Early in this investigation, the third method was proven unsatisfactory because of excessive shrinkage and difficulty of grinding away the excess material. The first method was found time consuming. The second method was adopted after a comparative study was made of the performance of gages cemented to faces of specimen that had been prepared by this method and the first method. This latter study was conducted on specimens taken from the product 3SG5, 9S5 and 2CES5o Two opposite faces were prepared by the first method while the two remaining faces were prepared by the second method. Gages of the two opposite faces were then connected in a series and the specimen was tested in compression. The results (presented in detail in Appendix VI) of the two methods of preparing the surface were very similar. 3. Procedure of Test. The procedure that was finally used for preparing the surfaces of lightweight aggregate test specimen was similar to the second method described above, except that the surface of the specimen to which the gages were to be cemented had first been smoothed by a commercial grinding machine, Plaster of Paris was then applied to the surface as in the second method, this time the pre-smoothed surface was easily cleaned of the excess paste. The specimen was then aged for 24 hours before grinding with a grinding belt, This second grinding was a necessary step to roughen the surface for better gage binding and to further clean plaster from the surface outside the cavities. Surface preparation by the method above was found unnecessary for specimens made with sand and gravel (products 3SG5 and 4SG5) since their surfaces are free of large cavities.

-128 - The following procedure was used for cementing the gages to the test specimen: the surface was cleaned thoroughly with acetone, a thin layer of Baldwin SR-4 precoat was applied and left to dry for about 1/2 hour, finally Duco cement was applied and the gages placed on their predetermined location (vertical gage at center of the face of the specimen, horizontal gage directly under and perpendicular to the vertical gage, see Figure 25) and pressed evenly for 15 minutes after which a weight of approximately one pound was applied through a rubber pad and held constant for 12 hours. Gages on the opposite face of the specimen were applied by the same procedure. About half of the specimens had vertical and horizontal gages (in both faces) while the remaining specimens had only vertical gages. Each gage was connected with its companion gage on the opposite face of the specimen in series, so that errors due to possible buckling of the specimen under load were eliminated~ During the test an active specimen was centered under the load applying block at the head of the machine (see Figure 22) while the dummy was placed on the bed of the machine close to the active specimen. Wire leads of the gages were then connected directly to the strain indicator when the specimen was bearing vertical gages only, otherwise connections were made through the switch box where the vertical and horizontal gages were connected to the active points 1 and 2 respectively while dummy gages were connected to both points 1 and 2 at dummy poles of the switch. Initial balances of the bridge were then accomplished for the gage. Load was applied at uniform speed and held constant at 1000 pound intervals until the indicator was balanced again for the gage (s). Tests were stopped at 8000

-129 - pounds for specimers made with lightweight aggregates and 14,000 pounds for specimensmade with dense aggregates. 4. Calculations. Detailed data and calculations of the results are given in Tables 53 to 73 and Figures 51 to 71 (in Appendix VI) along with the stress for vertical and horizontal (if present) gages for all specimer tested except 1C52 which were presented here (Table 16 and Figure 26) for purpose of discussion. TABLE 16. STRESS-STRAIN DATA (STATIC) FOR 1C52 (CINDERS) 1 2 3 4 5 6 Load Strs Vertical Gage Horizontal Gage Load Stress --—.. ----. ---- ---- Pounds psi Reading Strain mic-in mic-in mic-in mic-in 5 6 0 0 1305 0 1370 0 1000 123 1180 125 1380 -10 2000 246 1050 255 1400 -30 3000 369 920 385 1420 -50 4000 491 41790 515 1440 -70 5000 614 1650 655 1460 -90 6000 737 1515 790 1485 -115 7000 860 1380 925 1510 -140 8000 983 1245 1060 1536 -166 In Table 16 compressive stresses are given in column 2 while the corresponding strains are listed in columns 4 and 6 for vertical and horizontal gages respectively. From these, information stress-strain curves for the vertical and horizontal gages were

-130-:-:: -- - -- Figure 25. Static Test Specimen with SR-4 Gages.

-131 - 1200 - - -- 1050 1050- --------— C --- —-.- - - 900 ------- a. I I I I UT iH fTTT111i 0600 - ---- - 300 cfl450 -id A X o I 300- -- 0 0 150 300 450 600 750 900 0 50 100 150 200 FOR HORIZONTAL GAGE STRAIN MICRO-INCH FIG. 26 - STRESS-STRAIN CURVES FOR IC 52

-132 - plotted in Figure 26. Young's modulus was then calculated as the slope of the initial tangent (straight portion) of the stress-strain curve of the vertical gage. Poisson's ratio ( ) was calculated as the average ratio of the horizontal and vertical unit strains corresponding to the straight portions of the curves. Values near zero stress are nedt reliable and were excluded from the calculations. (See also discussion in Appendix VI). Modulus of elasticity in shear was then calculated from the previously given equation (15)o Results of all specimerB were calculated in the same manner and summarized in Table 17. D. Discussion of the Results of Elastic Constants as Determined by Dynamic and Static Methods. Average results of the static and dynamic tests are summarized in Table 18. Each value is the average result of two specimens except values of p and G of the static test which are mostly the results of one specimen (see Appendix VI). The most apparent feature in studying these results is that the values of the static tests are lower (some are appreciably so) than those of the dynamic tests. This has been observed by almost all other investigators (see reference 19, 39, 40) and it was concluded (reference 40) that the same properties of concrete are not measured by the two tests. It is also true that in a material like concrete, with less homogeneity and isotropy, that the theoretical equations used in the calculation of the results of the dynamic test and some of the static test are only approximately applicable. This, coupled with some possible creep in the static test, makes a difference in the results of both tests expected. This is why it was suggested by 4o some investigators, as was discussed at the beginning of this

TABLE 17. ELASTIC CONSTANTS (STATIC TEST) 1 2 3 4 5 Modulus of Young's Elasticity Modulus Poisson's in Shear Specimen E Ratio 2 (1 + ) E No. ksi 2( + p) ksi i. I I. 1C51 1C52 2CES51 2CES52 3C51 3C52 3SG51 3SG52 4SG51 4SG52 5CS51 5CS52 6ES51 6ES52 7W51 7W52 8CEL51 8CEL52 9S51 9S52 10CA51 10CA52 908 920 1013 1020 784 880.161.150.188.167.150 2.32 2.30 2.38 2.334 2.300 391 400 426 336 383 2333 2133 3357 2765 1056 1593 1705 1687 2200 2061.240.171.133.154 2.48 1354 2.34 2.266 2.308 752 953 1388 1109 2100 2416 797 984.130.126 2.26 2.252 929 354 chapter, that the dynamic test may not be suitable for determination of constants to be used in the design of static structures. Any of the two sets of the elastic constants are suitable for a comparative study of the products tested; the results of the

-134 - TABLE 18. AVERAGE RESULTS OF ELASTIC CONSTANT FROM THE STATIC AND DYNAMIC TESTS 1 2 3 5 6 7 8 Dynamic Test Static Test Modulus of Young's Elasticity Poisson's Specimen Modulus in Shear Ratio E G No. Aggregate E G ksi ksi ksi ksi -. g...~~~~~~~~~1. 10S.'_L_*____ 1C5 2CES5 3C5 3SG5 4SG5 5CS5 6ES5 7W5 8CEL5 9S5 10CA5 Cinders Cinders & Expanded Slag Cinders Sand & Gravel Sand & Gravel Cinders & Sand Expanded Slag Waylite Celocrete Air-Cooled Slag Cinders & Fly Ash 1002 1266 984 2921 3674 1619 1792 2412 1313 2562 1106 430 520 400 1260.16 022.24.16 914 1017 832 2233 396.16 426.19 360.16 1489 3061 1354.24 694 775 987 559 1083.17.16.23.18.18 1325 1696 2131 1249 2258 451.17 752.13 953 929.15.13 479 891 354.13 - -: ----......__ __ ___ dynamic test will be used, however, because they are the average of two specimenso Returning to Table 18, it may be observed that products made with dense aggregates have higher results than those made with lightweight aggregates. Of the former, 4SG5 (sand and gravel) has the maximum results of E = 3674 ksi (kips per sq. in.) and G = 1489 ksi.

-135 - Other dense aggregates, 3SG5 (sand and gravel) and 9S5 (air-cooled slag), have the results of E = 2921 ksi and G = 1260 ksi and E = 2562 ksi, G = 1083 ksi respectively. Lightweight aggregates 7W5 (Waylite) has the highest results of E = 2412 ksi while 3C5 (all cinders) has the lowest value of E = 984 ksi. Other studies include correlation of Young's modulus with dry weights, compressive and flexural strengths. For these, values determined by the static tests were used because the specimen in these tests were also subjected to high stresso Data presented in Table 19, columns 3, 4, 5, and 6 were taken from Tables 18, 8, 10, and 12 respectivelyo It can be observed in Table 19 that (1) rough correlation exists between compressive strength and Young's modulus. This correlation is better between products of similar aggregates (cinders, expanded slags....), (2) correlation also exists between Young's moduli and unit dry weights of the products. This is also obvious in Figure 27, (3) rough correlation also exists between modulus of rupture (R) and Young's modulus (E) as may also be seen in Figure 28o Another observation pertains to Poisson's ratio as related with the intensity of the stresses. As may be seen from the stressstrain curves of the horizontal gages, Poisson's ratio is higher at relatively high compressive stresses. This was also observed by other investigators.41

-136 - TABLE 19. COMPARATIVE STUDY OF YOUNGTS MODULUS (STATIC) WITH OTHER CONCRETE PROPERTIES 1 2 3 4 5 6 Young s Modulus Compressive Modulus Specimen (Static) Dry Wt. Strength of No. Aggregate E pcfo psi Rupture 'Ii, -'.,. i,... ksi psi - - I 1C5 Cinders 914 90.9 1357 272 2CES5 Cinders & 1017 93.1 1778 224 Expanded Slag 3C5 Cinders 832 83.4 1252 354 3SG5 Sand & 2233 132 8 2212 510 Gravel 4SG5 Sand & 3061 129.2 1983 635 Gravel 5CS5 Cinders & Sand 1325 97.1 1728 344 6ES5 Expanded 1696 101o3 1395 338 Slag 7W5 Waylite 2131 97.3 1506 334 8CEL5 Celocrete 1249 76.2 1199 193 9S5 Air-Cooled 2258 126.3 2195 526 Slag 10CA5 Cinders & 891 85.2 1663 277 Fly Ash - -- - ----

-137 - 3500 - 4 2500..J 93S _ - -- -- ---- -- ----- 0 >- IO 500-o z ------ ----- - --- 0 70 90 110 130 150 DRY WEIGHT PCF FIG. 27 - YOUNG S MODULUS (STATIC) VS. UNIT DRY WEIGHT 3500 LEGEND FOR FIG. 27 E 28 CO 3) D 250 I:3 0 0 CD z 0 50' e 4 I 0 --- }_ — - -- -— * --- 8 5 _5 2 - = - t0, --- - t0 - _ --— _ _ _ _ _ _ _ _ o - - - - - - - - - - - - - 2 30C 3S= 4= 5 6 7 8 = 9 = I10 IC5 2CES5 305 3SG5 4SG5 5CS5 6ES5 7W5 8CEL5 9S5 10 A5 100 300 MODULUS 500 700 OF RUPTURE PSI 900 FIG. 28 - YOUNG'S MODULUS (STATIC) VS. MODULUS OF RUPTURE

CHAPTER IX FREEZING AND THAWING DURABILITY Repeated freezing and thawing is one of the most destructive agencies to which concrete is exposed. In some climates, concrete masonry may be exposed to many cycles of freezing and thawing annually. The severity of the exposure depends on the seasonal temperature changes and the location of masonry units in the structure. Units, for instance, near moisture at the ground level or in parapet walls may suffer a severe exposure. Investigators, realizing the extent of damage caused by this agency, have attempted to solve the problem of predicting the freeze-thaw durability of concrete. Most of their progress was accomplished in the last decade by extensive research, generally aimed at either investigating the mechanism of the forces of deteri43 oration and conditions of their action such as the work of Powers..., or at developing various methods and apparatus for laboratory accelerated freezing and thawing and conducting tests to determine the com44,45 parative resistance of various concretes..., or at determining the resistance by exposing concrete to long-time weathering at different climates and in different conditions. Attempts were also made to correlate the results of accelerated tests with those of natural 46,47 weathering. These investigations have indicated that damage, originating in the paste or the aggregate, or both, is mainly from stresses developed by rise of the internal pressure during freezing of the retained water. It may also be caused by thermal stresses due to sharp rise 138

-139 - and fall in temperature; or by expansive action of deleterious aggregate particles. The mechanism of frost action in the paste is adequately 43 treated by Powers43 in his theory of hydraulic pressure and water diffusion. Cement paste is capable of holding large quantities of freezable water in its capillary structure. This water produces hydraulic pressure upon freezing, and is caused by the resistance of this fine capillary structure to the flow of water displaced by the advancing ice front. Pressure may also be caused by the growth of the ice body in the capillary structure by diffusion of gel water at lower temperatures than at which the ice bodies were originally formed. Both kinds of pressure can produce destructive expansion of concrete although the latter acts more slowly. Presence of air voids relieves these pressures provided they are closely and uniformly distributed so that the displaced water can readily move to them. This explains the role of air-entrainment in improving the resistance of concrete to freezing and thawing. For concrete masonry units made with lightweight aggregates another factor may affect their resistance decisively. It was noticed that all the products tested in this study contain a large percentage of open air voids aside from those contained by the aggregates. This is apparently because the mixes are designed to result in highly cellular concrete, thus promoting light weight. This condition however, permits the concrete to retain a large quantity of freezable water in some cases (as may be seen from Tables 6 and 9, Chapters III and V respectively), thus making it very vulnerable to severe cycles of freezing and thawing such as those employed by the accelerated test.

-140 -Coarse aggregate was also shown to play an important part in the overall freeze-thaw durability of concrete.8 Concrete made with aggregate that is structurally weak, possessing high rate of absorption or containing deleterious particles may exhibit poor performance. Rapid drop and rise of temperature in accelerated freeze-thaw test might develop destructive thermal stresses, especially when the thermal incompatibility of the paste and the aggregate is large. A short time before the present tests were conducted, the Michigan State Highway Department Laboratory at Ann Arbor finished building a new accelerated freeze-thaw apparatus and have since been carrying on a rather large program of tests to evaluate the durability of concrete made with the different aggregates used in Michigan. The author was fortunate enough to secure space in the chamber for the specimens of his tests. A brief description of this apparatus will follow later. The cycle employed consists of 3 hours of freezing in air at 0 F, followed by 1 hour of thawing in water at 40 F. This procedure was designed to expose the aggregates in concrete to the more severe conditions of the test52 and proved unsuitable for the lightweight concretes. Concrete specimens made with most of these deteriorated after a few cycles. This is believed to be caused mainly by the absorption of the highly cellular concrete containing the lightweight aggregates and their capacity of retaining large quantities of freezable water as was discussed previously. Since it has been intended to study only the comparative performance of the products made with dense and lightweight aggregates, it was decided to modify the preceding conditions by keeping the moisture content of the specimen constant during the test. This was accomplished by sealing the specimens in tight copper jackets. Question arose as to the proper moisture content to be employed. The actual moisture content of concrete masonry in normal

-141 -service is presently indeterminate, and, of course, variable. However, the average of the natural moisture contents of all the products, as sampled from lots stored in open yards, was about 37 per cent of the 24 hour saturation (see Table 9, Chapter V). Sampling was done in the relatively wet season. Thus it was decided to choose a moisture content equal to 40 per cent of the 24 hour saturation of the particular specimen. The chosen moisture content is also equal to the maximum allowed by the ASTM specifications for units delivered to the construction site. A. Specimen. Test specimens were beams of approximate dimensions of 2.2 x 3.2 x 15.6 inches, cut from solid masonry units of the modular dimensions of 3-5/8 x 7-5/8 x 15-5/8 inches (see Chapter II). Absorption of each specimen after 24 hours submersion was determined by the same procedure that has been described in Chapter V. The dried specimens were wrapped tightly in a thin (No. 30 gage) copper foil which was then sealed by solder except at one end where a quantity of water, equal to 40 per cent of the 24 hour saturation absorption, was injected just before this end was also sealed. The jacket was tested by submerging the specimen in water for 24 hours. Increase in weight indicated leakage in which case the jacket was removed, the specimen dried and equipped with a new jacket. B. Apparatus.52 Test apparatus consists of freezing equipment, thawing equipment and the specimen chamber. As is shown in the simplified diagram of the apparatus (Figure 29) the specimen chamber is located inside the cold room. The well insulated room is about 12 x 10 feet in size and continuously kept at a temperature of 0 F to -5 F, The specimen

chamber is open at the top and is placed under the cooling fans where it is subjected directly to the cold air circulation. Thawing water is stored in a tank, located just outside the cold room at a lower level, at a temperature of approximately 42 F. During the thawing cycle the water is circulated to the specimen chamber by a centrifugal pump. The loss in temperature of the thawing water is small because the capacity of the tank is much larger than that of the specimen chamber and is corrected in the thawing tank by five immersion heaters. The apparatus is completely automatic. The freezing equipment circuit is separate from that for thawing. The apparatus performs within the requirements of ASTM Designation C291-52T (Tentative Method of Test for Resistance of Concrete Specimen to Rapid Freezing in Air and Thawing in Water). C. Procedure of Test. The specimen was first cooled to 40 F and then weighed; its fundamental resonant frequency of transverse vibration was determined using the sonic apparatus and method described in Chapter VIII. However, preliminary investigation had indicated that accurate determinationoof the sonic modulus is difficult when the drive and pickup were made through the copper jacket. This difficulty was overcome by equipping the specimen with brass plugs at these points. Plugs were cemented in holes drilled in the specimen. The specimen was then moist cured for 24 hours and air cured for one week (in normal room conditions) before drying in the oven prior to wrapping in copper foil. The plugs extended about 1/8 inch beyond the surface of the specimen, through the copper jacket (see Figure 31) and were suitable for transmitting the drive and pickup. Test showed that the presence

Figure 29. Schematic Diagram of Freezing and Thawing Apparatus. Courtesy of the Michigan State Highway Department. Figure 30. Specimens Chamber of the Freezing and Thawing Apparatus.

-144 - Figure 31. Freezing and Thawing Durability Test Specimen Before and After Copper Jacket is Added. Figure 32. Control Specimen for the Freezing and Thawing Test.

-145 - of the copper jacket and the brass plugs affected the frequency reading but slightly because their masses are small compared with that of the specimen. The resonant frequencies of the specimen thus determined at 40 F was considered as a measure of its initial Young's modulus of elasticity. After the initial resonant frequencies of the test specimen were determined it was placed in the specimen chamber at the beginning of the thawing cycle. Each specimen was placed vertically in a grid compartment so that it was surrounded by uniform air space (see Figure 30). As was previously indicated, the apparatus is automatic and the freeze-thaw cycles were continued without interruption (except when sonic readings were made), until the termination of the test at the end of 500 cycles. Resonant frequencies and weight of the specimen were measured at the end of 6, 12, 18, 30 and every additional 24 cycles thereafter. Specimens showing signs of deterioration were measured more frequently. In all measurements, the specimens were taken and returned to the chamber during a thawing cycle. D. Discussion of Test Results. Results of the test are given in Table 21; additional data in Table 20. In Table 21 the durability of the specimen is measured by the effect of the freezing test on Young's modulus and the modulus of rupture of the specimen. These results are based on the alternate freezing to 0 F and thawing at 40 F maintaining a constant moisture content (40 per cent of 24 hours saturation) in the specimen. Two control specimens, one of which was made with sand and gravel and the other with expanded slag, were used to measure the change in

temperature at their centers during the cycle. The temperature of the lightweight aggregate specimen dropped from 40 F at the start of the freezing cycle to 32 F in 20 minutes and 0 F in 140 minutes, It then remained at 0 F to -2 F for the last 40 minutes of the freezing cycle. The temperature climbed rather rapidly at the beginning of the thawing cycle, reaching 32 F in 35 minutes and to 40 F in about 50 minutes from the start of the thawing cycle. The time-temperature relation of the dense aggregate specimen followed the same pattern except it reached the indicated temperatures in slightly shorter time. Detailed data of the sonic and weight measurements which were made regularly during the test, as described in part C, need not be presented here, Weight measurements were made to insure against the leakage of the copper jacket; initial weights, given in column 8 of Table 20, remained constant during the testo Effect of a number of cycles of test on the specimen is evaluated by its Young's modulus at the end of these cycles (Ec) which was determined as a percentage of the initial modulus (Eo). Such values ( of EC x 100) are given for the test specimens in Eo columns 3 and 4 of Table 21 for the number of cycles indicated in column 2. Actual calculations of the values of Young's modulus are not necessary for these calculations since it is directly proportional to the n value (n being the resonant frequency of transverse vibration. See equation (7) in Chapter VIII). The initial values of n (copper jacketed specimen) are given in columns 2 and 3 of Table 20 for drive in the directions of a and b respectively (a and b are dimensions of the section of the specimen, in the same manner used in

TABLE 20. FREEZING AND THAWING DURABILITY - INITIAL YOUNG'S- MODULUS AND WEIGHT MEASUREMENTS 1 2 3 4 5 6 7 ' Initial Resonant Frequency- Initial Modulus of Copper Jacketed Elasticity (E ) Specimen of Concrete Weight of Specimen Specimen na nb Individual Average Dry Wet Test No. Cycles Cycles ksi ksi Weight Weight Weight per sec. per sec,.Ens.. g.s.. g ms.s. 1C81 1C82 775 1100 790 1100 2CES81 2CES82 3C81 3C82 3SG81 3SG82 4SG81 4SG82 5CS81 5CS82 6ES81 6ES82 790 810 840 840 1100 1050 1100 1180 825 840 985 1010 1100 1110 1100 1110 1550 1550 1550 1700 1150 1200 1330 1340 1271 1344 1438 1467 1109 1123 3772 3539 3875 4695 1603 1791 2517 2504 15 2618 2944 1453 2647 2974 2378 2726 2688 1308 2256 2634 2790 2982 2964 2940 3528 3498 1116 2950 3555' 3510 2633 3875 3656 2664 3963 3989 4056 3697 3948 4101 4285 3853 4057 4223 1697 2676 2954 1697 2802 3088 3047 3207 1 3088 3412 3461 2511 2969 3280 3330 7W81 7W82 8CEL81 8CEL82 9S81 9S82 10CA81 10CA82 1050 1010 800 900 995 1030 810 775 1450 1410 1180 1250 1390 1400 1120 1050 2918 2603 1448 1672 2824 3175 1462 1249 2761 2925 1560 2925 _ 2426 1560 2339 3170 3455 3190 3297 2733 2767 2642 2682 3000 3312 2586 3684 3445 3736 3969 2531 2889 1356 2367 2737 2907 2764

TABLE 21. FREEZING AND THAWING DURABILITY - EFFECT OF TEST ON YOUNG'S MODULUS AND MODULUS OF RUPTURE 1 2 3 4 5 6 7 Per Cent of Initial Modulus of Rupture "R" Modulus of Elasti- "Original" city = Tests Number Ec After 500 Freeze- (No FreezSpecimen of x 1 Thaw Cycles ing and No. Cycles Individual Average Individual Average Thawing). per cent per cent _ psi.psi psi 1C81 1C82 2CES81 2CES82 3C81 3C82 3SG81 3SG82 4SG81 4SG82 5CS81 5CS82 6ES81 6ES82 7w81 7W82 8CEL81 8CEL82 9S81 9S82 lOCA81 10CA82 500 tf It 11 11 It if l! If ft T! 102 13 100 96 100 100 100 98 100 100 100 99 94 i00 98 96 100 99 83 92 70 53 93 93 98 100 99 100 100 97 97 100 88 389 456 529 517 570 564 698 689 897 774 419 431 558 542 574 594 305 239 443 271 506 378 423 523 567 694 836 272 224 354 510 635 425 550 584 272 357 338 334 193 526 277 500 Tf 93 442

-149 - Chapter VIII)o Sonic measurements were always made for both directions and the average results presented, Examination of the res-i.ts of Table 21 reveals that, generally, all specimens except those made with air-cooled slag (9S8) withstood well the exposure to the test conditionso Considering for the moment the reduction in Yowig gs modulus as the measure of durability, it can be seen from col.um 4 of the table that Younggs moduli at the end of 500 freeze-thaw cycles (E,) are equal to their initial values (Eo) for products made with said and gravel (3SG8 and 4SG8) and two of the lightweight aggregatess, namely cinders plus expanded slag (2CES8) and Waylite (7W8)o The rest of the lightweight aggregate products, except for Celocretleg did almost as well, producing values of E between 99 and. 93 pe:. -:t of original E. Celocrete (8CEL8) averaged a lower ratio, 88 per cent. Of t:he dense aggregates, air-cooled slag (9S8) producjed a mufk.h lower result. Young's modulus of one specimen cf this product was lowered to 70 per cent of its initial value after 1'2 Tcces -whi.e the ratio of the second specimen was 53 per cent after only 13 cycleso The sig.ificance of these results will be disaoussed l. ter Modulus of r-pture (R) is another measure of freeze-thaw durability of concrete. All specimens were broken in flexure at the end of the test, Values of R are g-xven in Table 21o All products except air-cooled slag, showed larger values of R than those obtained earlier ("original" test in Ta-le 21) from similar specimens, as may be seen from columns 6 and 7 of Table 21. Values of column 7 were taken from Table 12, Chapter VIZo Th>is increase is believed to be due to gain in strength d-:ring the period separating the two tests

-150 -(one year) in which the specimenswere stored in normal laboratory conditions. Nevertheless, the results conform with the indications offered by Young's modulus. Air-cooled slag showed a substantial reduction (instead of increase) in its R value. The increase in strength during the storing period is also apparent in the values of Young's modulus. This can be seen by comparing the values reported in Tables 20 (column 5) and 18 (column 3, Chapter VIII). Values reported in Table 20 are taken directly after the specimens were taken from storage (they were not computed from na and nb of columns 2 and 3 of the table which are the initial values for copper jacketed specimens), those in Table 18 are for similar specimensthat had been tested earlier. The causes of the relatively poor performances of the aircooled slag and Celocrete products are not apparent. The performance of the Celocrete may partially be assigned to its relatively high absorption and inherent structural weakness. But this is not true for air-cooled slag which exhibited the poorest durability performance. Neither can the amount of absorption be the reason at constant moisture content of 40 per cent. Another reason considered is the thermal incompatibility of aggregate and mortar which is in some cases held responsible for some durability failures.49 As will be shown later in Chapter XII, the thermal coefficient of expansion of this product is not far from that of sand and gravel productso It can also be seen in Chapter XII that the thermal coefficient of expansion is affected by the moisture condition of the specimen which complicates matters. It may also be noticed that cinder products in general have lower thermal coefficients which may indicate greater incompatibility; this, however, showed no effect on their freeze-thaw durability. The

-151 - probable reason for the poor performance of the air-cooled slag is deleterious particles in the coarse aggregate which can inflict considerable damage by expansion during freezing and thawing. Fine cracks generating radially from a point, discovered upon removal of the copper jacket of specimen 9S82, support this conclusion. Poor durability of concrete cannot normally be assigned to a single cause. More accurate determination of the probable reasons of the performances of these products requires a larger program of tests than was possible. It was also not possible to isolate the effect of one element on the results, such as coarse aggregate, mortar, air entrainment...obecause, as was discussed in the opening of this chapter, the overall durability of concrete depends on the rest of its physical properties such as the amount of absorption, presence of effective air voids, elastic properties, and presence of deleterious& materials.....The preceding properties depend, besides the individual properties of the constituents, on the water-cement ratio and other proportioning of mix, mixing, curing....These as have'been shown before (Table 3, Chapter II) and varied from one product to another. Consideration was also given to the proper point of termination of test. It was decided, after reviewing the literature, that because of the nature of the study no important information could be obtained that warranted extending the test beyond the 500 cycles employed. It should also be noted that the results obtained by this test do not necessarily indicate the actual performance of the products under natural weathering. It is desirable that correlation with natural weathering include at least one aggregate of known

-152 - history of field performance in the accelerated freeze-thaw test. However, the comparative study of the durability of dense and lightweight aggregate products as affected by accelerated freezing and thawing permitted drawing certain broad conclusions. These conclusions are summarized in the following: (1) These lightweight aggregate and air-cooled slag con, cretes of a highly cellular nature, unlike sand and gravel products, are vulnerable to a very limited number of cycles of accelerated freezing and thawing when fully saturated. This is probably because of their high absorption and capability of retaining a large amount of freezable water in their highly cellular structure as was discussed at the beginning of this chapter. When the moisture content is lower, about 40 per cent of the 24 hour saturation which is the maximum allowed by ASTM specifications for units going into masonry, the resistance of lightweight aggregate products is generally equivalent to sand and gravel. (2) Indications are that thermal incompatibility of coarse aggregate and mortar did not affect the freezing and thawing durability materially. This conclusion is offered with reservation. It is based on the observation that products which were believed to have larger thermal incompatibility such as those made with cinders (see Chapter XII) did not exhibit poor performance. Researchers differ on the role of thermal incompatibility in durability. For instance, the Bureau of Reclamation studies5 of concrete with large thermal incompatibility reached a conclusion comparable to that offered above, while the Waterway Experiments Station studies on concrete with 4similar characteristics reached an opposite conclusion similar characteristics reached an opposite conclusion. It is

-153 -believed that the picture is clouded with the fact that the moisture content of the specimen affects the thermal coefficient of expansion of the mortar. A more detailed discussion on thermal incompatibility is presented in Chapter XII. (3) Poor freeze-thaw durability performance of air-cooled slag products is probably caused by deleterious particles in the coarse aggregate. Additional tests are necessary before reaching a positive conclusion, however.

CHAPTER X STAINING AND POPOUTS OF CONCRETE MASONRY UNITS Staining and popouts are unwanted characteristics of concrete masonry, producing a disagreeable appearance when prevalent in a wall. Furthermore, popouts, especially when accompanied by cracking, weakens concrete and makes it more vulnerable to the destructive action of weathering. Lightweight aggregates, as has been discussed in Chapter III, may contain certain deleterious materials that are capable of producing staining and popouts in concrete o3'23 Stains may be produced by metallic (tramp) iron or iron compounds such as pyrite (iron sulphide) which may exist in partially burned pieces of coal in slag and cinders. Both are capable of causing popouts. However, pieces of free lime which are formed by the high temperature decomposition of calcium carbonate are the main source of popouts that are not accompanied by staining. The presence of air and moisture is necessary for the process 'of oxidation of these deleterious materials which result in the staining and popouts. By oxidation iron sulphide (pyrite) is turned into soluble iron sulphate, which may hydrolyze to form iron hydroxide (the insoluble stain) and sulphuric acid may react with lime in the cement yielding the iron hydroxide and calcium sulphate. The hydration of a particle of free lime located at a point below the surface may result in an expansion of considerable force, causing the popout of concrete at the surface. 154

-155 - In this test favorable conditions were created for the process of staining and popout by storing the products in the presence of moisture and relatively high temperature for a period of 26 months. Performances of these products were then determined by visual examination and weight measurements. The application of paint to the outside surface of concrete masonry has become a common practice. It offers some protection from water penetration by sealing the surface cavities, especially for products made with lightweight aggregateso A variety of such paints are commercially available. One of these, made with a cement base, was used for painting one face of the specimen before exposing to test conditions. A. Procedure of Test. Test specimens were hollow masonry units (blocks) of the modular dimensions 7-5/8 x 7-5/8 x 15-5/8 inches (see Chapter II). Two specimens were used per product. Units were cleaned (of dust, etc.) and one face painted with a commercial cement base paint. Two layers of paint were applied with a stiff brush. In the first application, the paint was carefully worked inside the surface cavities of the specimen, which was then moist cured for 24 hours before the second layer was applied. Curing after the second application was 7 days in the moist room followed by 7 days in air at normal room conditions. At the end of this period the specimens had lost moisture during the air curing and were further dried in the oven at 210 F for 24 hours. Near the end of this drying, two successive weighings at 2 hour intervals showed a loss of weight not exceeding 2 per cent of the last previously determined weight of

-156 - the specimen which was the criterion used before for discontinuing the drying (see Chapter IV). Specimens were taken then from the oven and allowed to cool to room temperature before they were taken back to the moist room for the long time (26 months) storage. Inside the moist room, temperature is approximately 75 F and relative humidity is 100 per cent. This condition is maintained by fine water jets and an automatic cooler. At the end of the storage period, the specimens were taken outside the moist room and immediately weighed. They were then individually examined for possible stains and popouts. After examination they were taken to the oven for drying. This second drying followed the same procedure as the first drying although required more than 48 hours. B. Discussion of Test Results. It was previously stated that performance of each specimen was determined by visual examination and by dry weight measurements before and after the test. Visual examination indicated the following: (1) None of the products tested showed evidence of staining or popouts except those made with cinders. (2) Of the cinder products, 1C31 showed significant popouts. The resulting hole was conically shaped, about 3/4 inch in diameter at the surface and more than 1/2 inch deep. It was not accompanied by staining and examination revealed that it was caused by hydration of free lime. (3) Stains were seen on the surfaces of the specimens 2CES31 (cinders and expanded slag), 3C31 (cinders) and 5CS32 (cinders

-157 - and sand). Each was a single spot of small area (except for 5CS32). All staining was caused by tramp iron observable by the naked eye. (4) Popouts and stains appeared only on the unpainted surface of the specimens. Dry weights before and after test are given in Table 22, columns 3 and 5 respectively. These did not indicate loss in weight during the test which might be expected. In fact, all of the specimens showed slight increase in dry weight after test. The increase is less than 3 per cent and believed to be caused by water still held in the capillary structure. As was previously noted (part A) both weights are based on minimum drying time of 24 hours after which drying was discontinued when the weight of the specimen reached a certain stability. Weight stability was reached for all specimens in the first drying time of 24 hours but not before 48 hours of the second drying (after storage). After this latter drying, although stability of weight was reached, the specimens, as said before, still held a certain amount of water in their capillary structure. Another interesting observation is that in previous tests of companion specimens for determining their capacity for water absorption by 24 hour submersion (see Chapter V), such a degree of..weight stability was reached within 24 hours of drying. This suggests that a higher degree of saturation was obtained in the long time storing of this test than was possible by the 24 hour submersion. The former processes provide,. ample time for the rather slow operation of filling the finer capillary structure with water, a certain portion of which remained after 48 hours of dryingo Further experimenting with some

-158 - TABLE 22. RESULTS OF STAINING AND POPOUTS OF TEST OF CONCRETE MASONRY UNITS 1 2 3 4 5 6 Dry Weight Dry Weight Type Before Wet After Staining Specimen of Storage Weight Storage and No. Aggregate gnMs gms ggms Popouts 1C31 1C32 2CES31 2CES32 2H31 2H32 3C31 3C32 3SG31 3SG32 4SG31 4SG32 5CS31 5CS32 6ES31 6ES32 7W31 7W32 Cinders Cinders & Expanded Slag Haydite Cinders Sand & Gravel Sand & Gravel Cinders & Sand Expanded Slag Waylite 11,820 11,770 13,075 13,030 11,560 11,470 12,600 12,240 17,260 17,345 16,340 16,710 11,750 12,310 13,135 12,360 12,400 12 250 13,830 13,810 15,460 15,500 14,410 14,310 15,610 14,650 8, 540 18,610 17,690 17,850 13,600 13,970 i4,88o 4, 190 13, 900 13,770 12,020 11,980 13,300 13,330 Popout Stain 11,700 11,600 12,830 12,430 Stain 17,460 17,590 16,560 16,860 i1,930 12,510 Stain 13, 400 12,640 12,600 12,470 7BW31 7BW32 8CEL31 8CEL32 9S31 9S32 10CA31 10CA32 Beslite & Waylite Celocrete Air-Cooled Slag Cinders + Fly Ash 11,735 13,100 8,810 11,430 16,960 16,750 11,535 11,460 13,650 ]5, 010 11,720 13,930 18,530 18,100 13,840 13,700 11,96o 13,360 9,110 11,620 17,160 16,950 11,730 11,650

-15'9 -of the lightweight aggregate products determined that it was not possible to return the specimen to its original dry weight by as much as 60 hours of oven drying.

CHAPTER XI DRYING SHRINKAGE Drying shrinkage and thermal expansion are main causes of volume changes in concrete masonry. The latter will be presented in Chapter XII. Concrete shrinks when losing moisture. The magnitude of this shrinkage depends on the physical and chemical properties of the constituents (cement and aggregate), mix proportions, mix consistency and method of manufacture, and especially on the method of curing (see references 53,54155; see also Chapter I). Shrinkage may subject concrete to stresses exceeding its ultimate strength, and being relatively weaker in tension may result in cracking. Pickett56 advanced a theory formulating shrinkage stresses on concrete by employing the laws of heat diffusion, i.e., by assuming "that the laws governing the development of shrinkage stresses during drying are analogous to those governing the development of thermal stresses in an ideal body during cooling", and then accounting for the effect of inelastic deformation and change in the moisture content. He showed that when a saturated specimen is dried at 50 per cent relative humidity (R.Ho) the stress developed is high. In fact, he showed that it would be much greater than the strength of the concrete if it were not for the effect of the plastic flow. Plastic flow, occuring when concrete is exposed to sustained stress, such as that caused by shrinkage, serves to relieve this stress. The "relative immunity" to cracking of concrete masonry

was therefore found5 to depend upon its modulus of elasticity under sustained stresses, its tensile strength, its tendency to shrink and its degree of restraint. As will be seen later, saturated concrete may lose a great portion of its moisture before shrinkage stresses start to develop --- this is the water contained in the relatively large voids existing in both the aggregate and paste. However, the loss of moisture from fine pores like the capillaries, or the "desorption" of the surface adsorbed water of the cement gel, will cause rapid development of these stresses. Knowledge of the mechanism of water adsorption is therefore essential. 57 Powers and Brownyard5 studied this mechanism in hardened cement paste which influences the shrinkage of concrete. They proposed a theory based on combining the Brunauer, Emnent and Teller Theory (which is also known as the multimolecular-adsorption theory, or B-E-T theory for brevity) and the capillary condensation theory. Briefly, the BlE-T theory implies that a solid surface which is exposed to continuous bombardment of gas molecules attract physically and holds some of the gas molecules at least momentarily. Moreover, when the gas is a vapor, such as water, the molecules caught on the surface will be in a condensed state which may be considered as a separate phase. This condensation is accompanied by evolutions of heat energy. Some of the adsorbed molecules may escape the solid surface by acquiring sufficient kinetic energy; this causes a continuous interchange between the interior of the vapor phase and the solid surface but the average of molecular concentration at the latter surface remains

=162 - higher by the virtue of surface attraction0 It was also recognized in the derivation of the mathematical expressions of this theory that a molecule may be condensed on a bare solid surface or on a layer of previously consensed molecules. The capillary condensation theory57 is based on the fact that the surface of the liquid in the capillaries is "the seat of the available energy". When the capillary is partially emptied the water surface assumes a curvature and becomes subject to tension, consequently the vapor pressure of the water in the capillaries will be less than normal for the existing temperature. This pressure is inversely proportional with the curvature of the surface. Evaporation of the moisture from the surface or its condensation depends on whether the vapor pressure in contact with the surface is, respectivelyj lower or higher than that of the liquido Since hardened cement paste is a porous solid with its capillary structure usually containing water, Powers and 3rovwnyard proposed that condensation of moisture in its surface is influenced by the combined action of the forces considered in the B-E-T and the capillary condensation theorieso Another requirement of the B-E-T theory is that at any given pressure the amount of liquid ads:orlbed is directly proportional to the surface area of the solid, Powers and Erownyard were therefore able to measure the apparent surface area of the solid in the hardened cement paste by measuring the quantity of water adsorbed. Other investigators measured this surface by the amount of nitrogen adsorbed, with a large difference in the results of the two methodso This was observed by, among others, Kalousek53, whoa in his more recent work, also found that the amounlt of waer adsorption of concrete masonry units (autoclaved and normally cured) was approximately twice that of

-16 5 - the nitrogen adsorption. On the other hand he showed that the amounts of water adsorbed by xonotlite (hydrous monocalcium silicate with large surface area) is equal to that of nitrogen. These observations are of importance because water and nitrogen molecules do not differ greatly in size and nitrogen adsorption is considered to provide correct mea — surement of the surface area. Moreover, xonotlite undergoes a small drying shrinkage. Kalousek concluded that the excess water adsorbed is accommodated in the structure of the paste which he called the "interlayer water". He further assumed that the movement of this water into and out of the crystal lattice is one of the causes of volume changes in concrete products. In view of the preceding and other findings, he then proposed a tentative theory for the mechanism of stresses causing shrinkage dhring the various stages of dryingo With this theory he explained the moisture-shrinkage relations in a rather large program of tests on concrete masonry units cured with various methods including high-pressure steam and normal-pressure steam.15 Briefly, the theory states that shrinkage is caused by two types of stress developments: firs;, by surface tension of the partially filled capillary water, and, second, by "'spontaneous desorption" of the interlayer water. According to the theory, the first stress occurs during the first stages of drying and is released when the capillaries are dried in which case the specimen should re-expand. But re-expansion did not occur beca-se simultaneously with the elastic reexpansion, another stress, due to desorption of interlayer water, is developed and shrinkage continued later. These stresses being of opposite sign may also balance each other at certain humidities and thus permit no change in length during this periodo

-164 - A large number of investigations have been conducted in the past decade in an effort to contribute to the solution of the problem of drying shrinkage of concrete masonry. These researches are of two types: (1) basic,which are aimed at studying the mechanism of loss of moisture and shrinkage such as the works discussed previously, or at developing methods of tests for determining the shrinkage of concrete (see references 14, 55), and (2) comparative studies of the shrinkage and loss of moisture properties of concrete made with different mixes, with different aggregates or by various methods of curing (see references 54, 55, 57)0 Some of the findings in the latter researches will also be referred to later. The present work is a comparative study of loss of moisture and shrinkage by drying of concrete masonry units made with the various dense and lightweight aggregates under study. It involves one cycle of drying of saturated concrete, to a constant moisture content in air at 73 F and 50 R.Ho Tests were conducted on bars sawed from solid concrete masonry (slabs). Volume changes were measured, as the decrease in the linear dimension of the test specimen, by a comparator at certain time intervals. A. Procedure of Test. Test specimens were bars having a 2 x 2 inch cross section and approximately 11 inches longo They were cut from solid concrete masonry units of the modular dimensions 3-5/8 x 7-5/8 x 15-5/8 inches. (See Chapter II.) Two specimens were used per product. Reference points were provided at both ends of the specimen by standard gage plugs, which were installed as follows: Two holes (3/8 x 1/2 inch) were drilled at both ends along the longitudinal center line of the

-165 - specimen as shown in Figure 33, cleaned, moistened and partially filled with thick cement paste with 2 per cent (by weight) calcium chloride 22 (ASTM Designation C341-54). Plugs were then inserted to points providing a gage length (distance between the innermost ends) of 10 inches. Care was taken to keep the outside ends clean and free of cement paste. Specimens were then moist cured for 3 days, after which the reference points were tested and cleaned before their storage in water. Test procedure consists of storing the specimens 14 days under water at 73 ~ 2 F, followed by drying in air at the same temperature and 50 per cent R.H. At the end of the water storage each specimen was drained for one minute, the surface water was wiped off, and the specimen weighed prior to length measurements. These measurements were considered the initial values to which changes in weight and length of the specimen were referred. Weight measurements were made with a balance that reads to 1 gram, while length measurements were made by an extensometer with a dial micrometer reading to o0001 inch as shown in Figure 34. After determining the initial readings for all specimens they were stored in air, having the previously stated conditions, on steel racks with a space of not less than one inch around each specimen to permit the slow circulation of airo This storage continued for 133 days during which weights and lengths of practically all the specimens were stabilized. Weight and length measurements were made at the end of 1, 4, and 7 days and several 1 week, 2 week, and 3 week periods successively as may be seen from Table 23.

-166 - Figure 33. Drilling Hole in One End of Specimen for Cementing Steel Plug. Figure 34. Extensometer with Dial Micrometer Used for Length Measurements.

-16.7 - B. Discussion of Test Results. Results of the test are presented in terms of the accumulated loss of moisture and shrinkage at stated drying times (days) from the saturated condition. They are referred to the initial values of length and weight measurements at the end of the water storage as described in part A, Moisture loss is expressed in percentages of the total water of saturation. (Table 23.) The latter was determined by oven drying at 210 F to constant weight as is given in Table 26. Drying shrinkage is expressed in percentage of the gage length which is similar for all specimens (Table 24)o Results of Tables 23 and 24 are plotted in Figures 35 and 36 respectively. Each curve in these figures represents the average results of the two specimens tested for the indicated product. Figure 35, therefore, shows the rate of moisture loss during the drying, while Figure 36 shows the rate of shrinkage. Examination of Figure 35 reveals that all moisture loss curves followed a general pattern consisting of large loss of moisture (20-51 per cent) after the first day of drying and continued at a relatively fast rate during the following three days to reach values of 59-83 per cent at the end of 4 dayso They then continued to increase at successively lesser rates in the following periods reaching a comparative stability at the end of 21 to 35 days for most products. This stability continued for a long time, then was interrupted for some products which resumed slight moisture loss, at slow rate, to near the end of dryingo Likewise, examination of Figure 36 also indicates a similarity of pattern for shrinkage. The products underwent no shrinkage

-168 - TABLE 23. MOISTURE LOSS (% OF SATURATION) - DRYING TIME RELATION Drying Time (Days) Specimen 1 4 7 14 21 5 49 63 84 98 112 133 1C71 41 81 85 89 89 89 89 89 89 89 91 91 1C72 33 80 85 90 91 91 91 91 91 91 92 93 2CES71 24 75 80 88 90 91 91 91 93 93 93 93 2CES72 18 69 75 84 87 87 87 89 89 89 89 89 3C71 25 75 81 89 89 89 89 89 89 91 92 92 3C72 31 75 82 91 91 91 91 91 92 92 93 94 3SG71 43 75 78 83 84 84 84 84 85 87 88 89 3SG72 46 75 79 84 85 85 85 85 86 88 88 89 4SG71 49 69 75 80 82 82 82 82 82 83 85 85 4SG72 52 75 80 85 85 85 85 85 85 85 87 88 5CS71 52 76 79 85 86 87 87 88 89 89 91 91 5CS72 39 79 83 88 89 90 90 90 91 91 92 93 6ES71 30 62 69 79 82 83 83 83 85 86 87 88 6ES72 34 67 73 80 83 83 83 83 85 87 88 89 7W71 38 66 73 82 84 84 84 84 84 85 86 88 7W72 36 64 70 79 81 82 82 82 83 84 85 87 8CEL71 23 60 66 74 77 78 78 78 78 80 82 84 8CEL72 17 58 64 72 76 77 78 78 78 80 81 83 9S71 45 86 89 92 92 92 92 92 92 92 92 92 9S72 37 79 83 89 90 90 90 90 90 90 92 94 10CA71 27 81 84 90 91 91 91 91 93 93 94 94 10CA72 26 80 84 90 91 91 91 91 93 93 94 95 in the first day of drying, and only a small shrinkage occurred at the end of 4 days. However, shrinkage progressed at rapid rate afterward to the end of 21 days when, for most of the products, it stopped temporarily or continued at low rate in the period between 21 and 35 days of drying. The first 21 days will therefore be called the first stage

-16'9 - of shrinkage. During the second stage, after the 35th day of drying, shrinking resumed at a much slower rate than in the first stage but continued at approximately uniform rate to near the end of drying time when the lengths seemed stabilized for most of the productso By considering the two sets of curves in Figures 35 and 36, it can be seen that large loss of moisture in the first four days of drying produced only small shrinkage (o001 -.007 per cent), which indicates that these losses were by evaporation from the comparatively large voids of the aggregates and paste. Afterwards, when drying progresses to the smaller pores and to the capillaries, moisture loss continued at successively decreasing rates while shrinkage was accumulating at a fast rate, which may indicate full development of the capillary stresses through the mechanism discussed at the beginning of this chapter. It can also be seen that moisture content of most of the products were nearly stabilized after 21 days of drying while rates of shrinkage decreased drastically during the 21 - 35 day period, which later increased and continued approximately uniformly to near the end of the drying time. This may indicate a partial release of the stresses in the 21 - 35 day period by complete drying of some of the larger capillaries. However, stresses, and consequently shrinkage, resumed and continued afterward, without any significant loss of moisture, possibly by partial drying of the smaller capillaries and desorption of the interlayer water. This is the mechanism of shrinkage proposed by Kalousek and which was discussed at the beginning of the chapter. Close examination of Figure 35 will reveal the comparative losses of moisture from the various products during the various periods

TABLE 24. SHRINKAGE ( % OF LENGTHS) - DRYING TIME RELATION Type Specimen of Drying Time (Days) No. Aggregate 1 4 7 14 21 35 49 63 84 98 112 133 1C71 Cinders.000.006.014.022.025.026.030.032.034.039.041.043 1C72.000.004.014.025.029.031.037.039.041.047.050.052 2CES71 Cinders &.000.005.011.019.026.028.031.036.041.045.048.049 2CES72 Expanded.000.005.012.020.027.029.034.038.044.049.052.054 Slag 3C71 Cinders.000.003.008.013.016.017.018.022.026.029.030.030 3C72.000.002.007.015.017.017.021.023.027.029.030.031 3SG71 Sand &.000.004.012.019.021.021.023.026.029.029.030.030 3SG72 Gravel.000.002.009.020.021.022.026.028.032.033.033.034 4SG71 Sand &.000.005 13.021.025.027.033.036.041.044.044.046 4SG72 Gravel.000.005.014.024.029.029.032.036.040.044.045.046 5CS71 Cinders.000.006.013.022.026.027.030.034.040.044.046.046 5CS72 & Sand.000.006.015.028.031.033.035.037.042.044.046.047 6ES71 Expanded.000.003.009,021.027.029.032.036.041.045.047.049 6ES72 Slag.000.001.007.016.023.026.029.034.039.043.045.046 7W71 Waylite.000.001.007.015.020.022.027.031.037.042.045.045 7W72.000.001.005.013.018.020.022.025.029.035.037.037 8CEL71 Celocrete.000.007.018.036.046.050.057.062.068.073.076.079 8CEL72.000.002.012.029.042.048.052.057.064.069.072.074 9S71 Air-Cooled.000.002.007.011.012.013.015.018.022.023.024.024 9S72 Slag.000.001.005.010.011.012.013.016.020.022.023.023 10CA71 Cinders +.000.013.018.026.026.028.031.035.039.041.041.042 10CA72 Fly Ash.000.002.007.013.014.016.022.024.027.028.029.029

60 LEGEND I IC7 6 6ES7,70________________ 2 = 2CES7 7 -7W 7 z " 3C=-3C7 8 - 8CEL7 _ 0 A \ 3S ~ 3SG7 9 ~ 9S7 4 ~4SG7 10 IOCA7 n, 9 __I\" ~3;; I I II — "^ Z I ^ ^ I 5 1 - 5 ~5CS7 D 8,, (0 5 ~ ~ ~ ~ ~ ~ ~ ~6. 2C 0 00. 4 14 28 42 56 70 84 98 112 133 DAYS OF DRYING FIG. 35- MOISTURE LOSS- DRYING TIME RELATION

0.0 0 0.0 I IZ 9 o 0.02 - -- -- ---- -- -- (0 I I 0.037____ 3 3C 8_ 8CL - 0.04: 0.06 LEGEND in I = C7 6 =6ES7 0.07 3C3 C 7 8 = 8CEL7 3S = 3SG7 9 = 9S7 —. 4 =4SG7 I 0 =10CA7 5 = 5CS7 0 1 4 28 42 56 70 84 98 112 1.33 D A Y S OF DRYING I ro I FIG. 36 SHRINKAGE - DRYING TIME RELATION

-175 - of drying. It can be seen that there is no significant division in the performances of dense and lightweight aggregate products. However, it may also be seen that expanded slags (6ES7, 7W7 Waylite, and 8CEL7 Celocrete) lost smaller percentages of moisture in the first stage of drying than: cinders or dense aggregate products, but they suffered higher rates of loss in the later periodso Cinders lost moisture at higher rates than sand and gravel products which came next to the expanded slags during the early period of drying. Air-cooled slag showed the highest loss of moisture during this period, However, some of these proportions were changed at the later periods of drying which affected the total values as may be seen in Figure 35. Expanded slag products (6ES7, 7W7, and 8CEL7) losses of moisture at 133 days of drying were 89, 88 and 84 per cent respectively as compared with sand and gravel products 3SG7 and 4SG7 which lost 89 and 87 per cent, respectively. Cinder and air-cooled slag products showed a higher loss ranging between 91 and 95 per cent for the former and 93 per cent for the latter. Similar examination of Figure 36 will reveal the relative shrinkage performances of the various products. Again there is no significant division between dense and lightweight aggregate.: products. However, it can readily be seen that air-cooled slag (9S7 - dense) suffered the least shrinkage, while cinder product (except 3C7) shrinkages progressed at much higher rates. The products 3C7 showed much lower shrinkage, coming next to air-cooled slag; the reason of this will be discussed latero Sand and gravel products performed

differently with 4SG7 showing higher rates of shrinkage, close to cinder products, while 3SG7 shrinkage was relatively lower. Expanded slag products also exhibited different performances with 7W7 (Waylite) showing the lower rate of shrinkage while 6ES7 and 8CEL (Celocrete) showing, successively, higher rates. In fact, Celocrete showed the highest rate of shrinkage throughout the entire period of drying. The total shrinkage at the end of the drying, when the lengths were approximately stabilized, is of considerable significance. These values can be determined from Figure 36 at 133 days. It is apparent that Celocrete suffered the maximum shrinkage,.08 per cent compared with.024 per cent for air-cooled slag (9S7) which is the least. Cinder products (except 3C7 and 10CA7) came next to Celocrete but showing much lower values than the latter, ranging from.047 to.052 per cent. The product 3C7 showed a low value.031 per cent. Sand and gravel products had different values; a relative low.032 per cent for 3SG7 compared with a relative high.046 per cent for 4SG7. The remaining expanded slag products showed values close to those of cinder products with 7W7 (Waylite) having the lower shrinkage. Comparison of the two sets of curves in Figures 35 and 36 showed no correlation between moisture loss and shrinkageo For instance, air-cooled slag lost a large percentage of its moisture at the early stage of drying and continued to have its curve at the bottom of Figure 35 throughout the drying time but appeared at the top of Figure 36 with minimum shrinkage. Exactly the opposite held true

-175 - for Celocrete. These two products represent the extreme cases; others lie in between without permitting this correlation. This is also apparent in Table 25 and Figure 37 which contains the moisture contentshrinkage relation for the products. These curves are not parallel. Correlation of the shrinkage with other physical properties were also attempted. These properties are the previously determined unit dry weight, 24 hour absorption, modulus of rupture and Young's modulus of elasticity. No correlation was observed between shrinkage and any of these physical properties. This may be explained by the findings of other investigators which were discussed at the beginning of the chapter, that drying shrinkage is determined by the combined effects of the physical and chemical properties of the cement and aggregate, mix proportions, methods of curing...which influence the physical properties of concrete by varying degrees. But by such general consideration one can, in this study, explain the fact that lightweight aggregates products produced higher shrinkage than those of dense aggregates, except 4SG7 whbbse performance cannot be explained by such consideration. By the same token the comparatively low shrinkage of 3C7 (cinders) might be explained as caused by high-pressure steam curing which has been found to reduce shrinkage. Incidentally, the rest of the products were cured by normal-pressure steam at temperatures varying between 170-180 F (see Table 3). Another important observation pertaining to the drying shrinkage can be deduced from Figure 37 where it is apparent that at a moisture content of 40 per cent (the specification's limit) only a small portion of the potential shrinkage of the products had occurred,

TABLE 25. SHRINKAGE ( % OF LENGTH) = MOISTURE CONTENT ( % OF SATURATION) RELATIONS * Drying Time (Days) Specimen 1 4 7 14 21 35 49 63 84 98 112 133 1C7.000.005.014.024.027.028.034.036.034.043.046.048 62 19 15 10 10 10 10 10 10 10 8 8 2CES7.000.005.012.020.027.029.033.037.043.047.050.052 79 28 22 14 11 11 11 10 9 9 9 9 3C7.000.00 4.008.014.017.017.020.023.027.029.030.031 72 25 18 10 10 10 10 10 9 8 7 7 3SG7.000.003.010.020.021.022.025.028.031.031.032.032 55 25 21 16 15 15 15 15 14 13 12 11 4SG7.000.005.014.023.027.028.033.036.041.044.045.046 49 28 22 17 16 16 16 16 16 16 14 13 - 5CS7.000.006.014.026.029.030.033.036.041.044.046.047 54 22 19 13 12 11 11 11 10 10 9 8 6ES7.000.002.008.019.025.028.031.035.040.044. 046.48 68 35 29 20 17 17 17 17 15 13 12 11 7W7.000.001.006.04.019.021.025.028.033.039.041.041 63 35 28 20 17 17 17 16 15 14 12 8CEL7.000.005.015.033.044.049.055.060.066.071.074.077 80 41 35 27 24 23 22 22 22 20 18 16 9S7.000.002.006.011.012.013.014.017.021.023.024.024 59 17 14 9 9 9 9 9 9 9 8 7 10CA7.000.007.013.020.020.022.027.030.033.035.036.036 73 20 16 10 9 9 9 9 7 7 6 5 * Values are average results of two specimens. Shrinkage is on the first line, moisture content is on the second line.

FOR SRINKAGE - MOISTURE 0. 0 8 0.0 120 RELATION BELOW 50 % < M U RE CONEN % O A T MOISTURE CONTENT SEE LO - 0.0 3 - —, I0.0 0.04 0.05 0.04 0.06 — LEGEND 03 I =1 C7 6 6ES7 2 =2CES7 7 = 7W 7 0.0 7 3C = 3C7 8 =8CEL7 3S = 3SG7 9 = 9S7 =S 4 = 4SG7 10 10=IOCA7 5 = 5CS7 0.08 _ _ _ D 50 40 30 20 10 0 MOISTURE CONTENT 0/ OF SATURATION FIGo 37 SHRINKAGE - MOISTURE CONTENT RELATION (Example- For 8CEL7 containing 35% moisture, expressed as % of total absorption, will suffer a shrinkage when dried to 20% moisture content of 0.071-0o015 = 0.056%)

-178 - TABLE 26. TOTAL SHRINKAGE AND MOISTURE CONTENT AT THE END OF DRYING. WET AND DRY WEIGHTS OF SPECIMEN. Total Average Values Type at End of Drying Weight of Specimen Specimen of Shrinkage Moisture Saturated Oven-Dried No. Aggregate % of Loss % of gis gns Length Saturation 1C71 1C72 2CES71 2CES72 3C71 3C72 3SG71 3SG72 4SG71 4SG72 5CS71 5CS72 6ES71 6ES72 7W71 7W72 Cinders Cinders & Expanded Slag Cinders Sand & Gravel Sand & Gravel Cinders & Sand Expanded Slag. o48.052.031.034.046.046.048 92 91 93 89 1122 1176 1281 1269 1169 1120 1597 1699 1697 1657 1296 1298 14o6 1420 1355 1387 972 1002 1084 1090 976 941 1473 1578 1590 1540 1161 1143 1262 1270 1239 1270 92 89 88 Waylite 8CEL71 8CEL72 9S71' 9S72 10CA71 10CA72 Celocrete Air-Cooled Slag Cinders -i Fly Ash.077 1156 1169.024.036 93 95 1561 1500 1296 1236 990 1003 1428 1410 1092 1045

-17 9 - the greater portion of shrinkage occurring at moisture contents below 20 per cent. This observation was also reported by other investigators15 as was discussed previously in Chapters I and V. On the other hand, it should also be noted that the degree of drying obtained in this test is much higher than to be expected in the field. Menzel5 states that the moisture content of the concrete masonry unit in the field at equilibrium depends upon the relative humidity of the surrounding air. He also indicated that for Michigan, along with most other parts of the United States, the average annual relative humidity is 65 to 75 per cent which is much higher than the 50 per cent employed in this test.

CHAPTER XII THERMAL EXPANSION It was stated in Chapter XI that temperature and moisture variations are main causes of volume changes in concrete masonry. The temperature-volume variation property of concrete is commonly measured by the coefficient of linear expansion, which is usually defined as the increment of length in a unit length contributed by a rise in temperature of one degree. The numerical value of this coefficient commonly5 used for concrete design, 5.5 x 106 inch per inch per one degree F, does not agree with those determined for some lightweight aggregate concretes reported herein. Thermal expansion, like drying shrinkage, could cause high stresses in concrete. Tensile stresses are of great significance because of the inherent low tensile strength of concrete. Such stresses may develop by contraction of restrained concrete, or by its uneven heating or cooling when the outside face is exposed to sudden rise or drop in temperature. Internal stresses may also be developed in concrete containing aggregate of low coefficient of expansion; or when the coefficients of coarse aggregate and mortar are different. The coefficient of expansion of hardened cement paste was determined by many investigators to be between 6 and 12 x 10-6, which is higher than those of most of the commercially used aggregates (see references 49, 60, 61, 64) themselves. Thermal incompatibility of paste and aggregate occurs- when the latter.has unusually low coefficient. It may also occur between the mortar and coarse aggregate when the coefficient

-18l of the latter alone is comparatively low. Both incompatibilities were considered by some investigators as capable of causing destruction of concrete under certain conditions of exposure. The problem of thermal incompatibility of aggregate and matrix was studied by many investigators; among them are Pickett49 and later Glover.5 Coarse aggregate in concrete was considered approximated by spherical bodies embedded in a matrix of much larger volume and having a different coefficient of expansion. These investigators, separately, produced mathematical expressions for radial and tangential stresses on the matrix as caused by differential strain at the spheretdsboundary occurring by rise and drop in temperature and resulting from the difference of expansivities of- the aggregate and the matrix. It was shown that for the case under discussion (coefficient of matrix is larger than that of the aggregate) and by considering the mutual restraints exerted by the aggregate and the matrix, a large rise in temperature will produce radial tensile stress and tangential compressive stress of considerable magnitudes. An equal drop in temperature will produce an equal amount of stress in the reverse direction. In discussing these stresses, Callan 9'50 showed that for certain aggregate combinations the tensile stress might be much higher than the average strength of most concretes. He also indicated that stresses are expected to be even higher by concentration in the angles of the irregularly shaped aggregate particles, a factor that was not considered in the approximate solution. However, the amount of thermal stresses in concrete depends on several factors; among them are the amount and rate of temperature change, the coefficient of expansion of the constituents, the moisture content and the

-182 - degree of mutual restraint of coarse aggregate and mortar which is determined by the degree of their interfacial bond and their moduli 50 of elasticity. Robinson's mathematical expressions for the radial and tangential stresses at the aggregate boundary imply that the larger these elastic moduli the higher the stresses caused by a certain temperature change.* Concrete made with lightweight aggregates '-has: a comparatively low modulus of elasticity which is a favorable property to lower thermal stresses. If it were accepted that thermal incompatibility is the decisive factor in developing these internal stresses, consideration must then be given to the coefficients of expansion of aggregates when selecting a combination of two or more to be used with the mix, which is a common practice in the precast industry. Investigators are, however, divided on evaluating the effect 49 of thermal incompatibility on the durability of concrete. Callan4 conducted a large series of tests on aggregates, mortars, and concretes made with aggregates of varied coefficient of expansion. He also exposed concrete specimens containing these aggregates to rather severe cycles of accelerated freezing and thawingo By employing statistical methods in analysis of the results, he arrived at the conclusion that where the difference between the coefficients of expansion of coarse aggregate and mortar is large, the durability of the concrete may be considerably lower than would be expected from the qualities of aggregates as determined by the usual tests. Pearson arrived earlier at the same conclusion (considering, however, the thermal * For more discussion see the following references(1) K. Robinson, Journal of Applied Physics, Vol. 22, po 1045, 1950. (2) Ro Do Mindlin and Do Ho Cheng, Journal of Applied Physics, Vol. 21, p. 9319 1950.

-183 - incompatibility of aggregates and paste) by observing the unexpected failure after one winter of concrete in a certain structure which was made of aggregate of low coefficient of expansion, and later by conducting accelerated freeze-thaw tests on concrete made with the aggregate. Hornibrook63 also observed the same kind of failure of another structural concrete which was also made with such aggregate and agreed with Pearson's conclusions. Several other investigators, on the other hand, arrived at different or opposite conclusions to those described above, Bloem and Walker, also Willis and Reagel, in their discussion of the work of Callan disagreed with some of the conclusions of the latter author, and although still recognizing the possible effect of thermal properties on the durability of concrete, they minimized the significance of the apparent correlation of the incompatibility and durability as based on the statistical analysis produced by Callano They indicated that a similar analysis of test data also yielded a correlation between the coefficient of expansion of the mortar and the durability. Mitchell described a large program of tests conducted by the U. S. Bureau of Reclamation for the purpose of studying the effect of the thermal characteristics of aggregate on the durability of concrete. Tests were performed on concrete made with numerous combinations of aggregates having widely varied coefficients of expansions. It was concluded that no direct relationship existed between the thermal incompatibility of the aggregates and accelerated freeze-thaw durability of concrete (see also reference 51). But it was observed that the coefficient of expansion and the durability are both affected by the ease with which concrete becomes saturated and the degree of actual

-184 - 64 saturation. Walker, Bloem and Mullens6 studies also indicated lack of correlation between thermal incompatibility and durability. They indicated that resistance of concrete to temperature changes are related to two separate factors. These are the rate of temperature variation and the coefficient of expansion of the concrete itself. The findings were that over a given range a fast rate of temperature change is much more destructive to concrete than a slow rate, and that concrete of high thermal coefficient failed more rapidly than that of low thermal coefficient. They also concluded that deterioration of concrete during temperature change appears to result from strains set up due to a difference in temperature within the mass rather than due to a difference in coefficient of expansion- of the constituents. As was discussed in Chapter IX, the present tests, which were conducted on companion specimens, showed no correlation between the thermal coefficients of concrete and its resistance to accelerated freeze-thaw cycles. Also no deleterious effect was observed on the durability of products made with two aggregates of different coefficients of expansion such as 2CES6 (cinders and expanded slag) or 5CS6 (cinders and sand). Investigators agree on most of the other characteristics of the coefficient of thermal expansion. It was shown (see references 60, 61, 62) that the coefficient of a certain concrete is determined by the coefficients of its constituents and their proportions in the mix. Mullen thus suggested that the coefficient of expansion of concrete may be approximately calculated from the coefficients of its constituents. However, this is beset by the difficulty that the

coefficient of hardened paste is dependent on the moisture content. These investigators also agree that the coefficient of expansion of concrete ini bath the completely saturated and in the oven-dried conditions are approximately equal but are considerably higher at intermediate degrees of moisture content. Mitchell6 studied the variation of the coefficient of expansion of the hardened cement paste with moisture content. His results indicated that the coefficient of neat cement is minimum for vacuum saturated condition, slightly higher when oven dried, and maximum at some intermediate "optimum" moisture content which was estimated between 60 and 80 per cent of saturation. Results indicated that the coefficient increased by 100 per cent when the moisture content decreased from vacuum saturation by 10 to 20 per cent. The results of the present work indicate that the coefficient of expansion of the concrete immersed is higher than for oven dried. This may partially be caused by the effect of the moisture content of the immersed specimen being lower than that of the vacuum saturated. Mitchell's results also indicated that moist cured concrete has higher coefficient than the vacuum saturated or the oven dried. In the present study, coefficients of thermal expansion were determined for saturated (not vacuum saturated) and oven dried specimens by using two different methods of tests. A. Methods of Test. Two methods of test were used for measuring length changes: (1) the extensometer method, and (2) the variable resistance wire

-186 - strain gage (SR-4) method. The first method was used for saturated and oven dried specimens, while the second was used for oven dried specimens only. 1. Extensometer Methodo The same apparatus that was used for measuring shrinkage was used for measuring temperature length changes. The extensometer is equipped with a dial gage reading to a ten thousandth inch. It was described in Chapter XI and was shown in Figure 34. The specimens were also similar to those used for the shrinkage test~ they were bars having they dimensions 2 x 2 x 11 inches, sawed off solid masonry units (slabs) of the modular dimensions 3-5/8 x 7-5/8 x 15-5/8 inches (see Chapter II) and were equipped with stainless steel plugs at both ends for reference points. The procedure used for cementing these plugs was also similar to that described in Chapter XI. Two specimens were used per product. 2. The SR-4 Gage Method. Baldwin standard SR-4 gage type A-I was used in this method. This is the same type of gage used for measuring the strain in the elastic properties tests. It is a flat wire gage about 13/16 inch long, mounted on a paper base. More detailed description of the gage is to be found in Chapter VIII. Specimens were cut from companion samples of the solid masonry units used for the first method of test, to the dimensions 3-1/2 x 2-1/2 x 7 inches. The two opposite 3-1/2 inch faces of each specimen were then cleaned and prepared for cementing the gages. A single gage was cemented longittudinally on each of the two faces. The two gages were later connected in series. The procedures of preparing the surfaces and cementing

-187 - the gages were the same as those described in Chapter IX. Two specimens were used per product; both were considered "active"o The method of test for determining the thermal expansion here reported utilizes a differential principal wherein the gage usually referred to as the "dummy" is mounted on a copper bar of accurately known coefficient of expansion, The copper bar and associated gage (actually two in series were used) is maintained at the same temperature as the gage on the specimen and should therefore compensate for different responses at different temperatures but undergoes known expansion and contraction. A method was used for further compensation believed due to errors outside the gages themselves, and probably caused by the effect of heat variation on the electrical resistance of the lead wires or any such reasons. The method consists of providing the copper bar with another gage (also two in series were actually used), which is then considered as an active gage, i.e,, connected to the circuit in the same manner used for the specimens. The reading of this active gage (which is obtained by connecting the active and the compensating gages of the copper bar) is theoretically constant at all temperatures. Any difference between two successive readings during the test is believed to indicate the error discussed above. Since the indicated error is outside the gages themselves, it is assumed to affect the results of all the specimens under test by the same amount. To compensate for this error, its amount should, therefore, be subtracted algebraically from the indicated unit expansions of all the specimens. The copper bar, which is 1-1/2 x 1-1/2 x 14 inches, is shown in Figure 38. At one end of the bar is the compensating gage

-188 - Figure 38. Copper Bar with Compensating and Active SR-4 Gages for the Thermal Expansion Test.

-189 - and at the other end is the active gage. Each of these gages is connected in series with its companion gage, which is mounted on the opposite face of the bar. During the test the compensating pair were directly connected to the designated points of the SR-4 strain indicator, while the active pair were connected to the switch box, similar to those of the concrete specimens. The equation of the coefficient of expansion of the concrete 49 is therefore as follows: -. Rc - ARm + 9.8 AT (16) Where e = linear coefficient of thermal expansion of concrete x 10 per degree Fo T = T2 - T1 dego F where T2 and T1 are respectively the high and low temperatures of the cycle. ARc = difference between successive readings of the strain indicator at temperatures T2 and T1, inch x 10-6 per inch (considered negative when due to temperature rise). ARm = correction factor, equal to the difference between successive readings at temperatures T2 and T1, of the active gage of the copper bar, inch x 10 per inch. The number (9.8) in the numerator is the coefficient of thermal expansion of the copper bar x 10-6 per degree F. Since A T = 60 F in this test, equation 16 then becomes:

-190 - ARc - ARm + 588 e = - 60 (16a) B. Procedure of Test. As was previously stated, coefficients of expansion were determined for the saturated (not vacuum saturated) and oven dried concrete. 1. Saturated Specimen. The first method of test was used for this condition. Concrete bars were placed in water baths of varied temperature, between a minimum of 35 F and a maximum of 140 F. Length measurements were made for each complete cycle at the following temperatures: 35, 73, 110, 140 73 and 35 degree F heating cooling Coefficients were then calculated between these temperatures for each cycle (see Table 74, Appendix VII). Temperature was varied at a rather slow rate of 1/2 degree F per minute. The required temperature was held stabilized for one hour before lengths were measured. One cycle was completed every day to a total of 5 cycles. 2. Dry Specimen - Extensometer Method. Upon the completion of the first test the specimens were oven dried to a constant weight. They were then cooled to room temperature, placed inside an automatically controlled oven and heated to 100 F. This temperature was held constant for two hours, after which lengths were measured by Method 1. The temperature was then raised slowly to 162 F and was also stabilized for two hours before length measurements were made at this temperature. The same procedure was used in cooling back to 100 F which designated a complete

-191 - cycle. Coefficients were then calculated for both heating and cooling for the five cycles of tests (see Table 75, Appendix VII). 3. Dry Specimen - SR-4 Gage Method. Different size specimens were used for this test as was previously described in Method 2. The specimens which had already been dried before cementing the gages were placed in the oven, lead wires were placed outside the oven; those of the compensating gage were connected to the strain indicator directly, the rest were all active, including the active gage of the copper bar, and were all connected through the switch box selector (see Figures 39 and 40). The indicator, Baldwin's portable SR-4 strain indicator, type MB, is similar to that used for the elastic properties tests. More detailed description of the indicator is given in Chapter IX. The temperature inside the oven was raised slowly to 100 F and left constant to stabilize for 2 hours before length measurements were made. The switch selector is manually operated. Each active gage was separately switched with the compensating gage (of the copper bar) and the indicator was balanced. This operation took a relatively short time since half of the specimens were tested at a time. Following the same procedure, the temperature was rotated between 100 F and 160 F until five complete cycles were obtained. C. Discussion of Test Results. The coefficients of thermal expansion of the saturated and oven dried concrete are given in Table 27. Each result is the average of two specimens and five heat-cool cycles of test. More detailed data of the saturated concrete are given in Table 74 (Appendix VII) where it may be seen that the coefficients were calculated for three

-192 - Figure 39. Apparatus for Thermal Expansion Test Using the SR-4 Gage Method (Close-up with Oven Open). Figure 40. Apparatus for Thermal Expansion Test Using the SR-4 Gage Method (General View with Oven Closed).

stages of heating (35 to 73 F, 73 to 110 F and 110 to 140 F) and two stages of cooling (140 to 73 F and 73 to 35 F) for each cycle. Further discussion on Table 74 is given in the appendix. More detailed data is also given in Appendix VII for the dry concrete, Methods 1 and 2 of test, in Tables 75 and 76 respectively. Likewise, the coefficients were calculated for the heating and cooling stages of each cycle. Generally, results of Table 27 indicate that cinder products have lower coefficients of expansion than sand and gravel products or the slag products. The latter two have values of similar magnitude. Close examination of any set of the results in Table 27 such as those of the saturated condition which represents a larger range of temperature, reveals that 10CA6 (cinders plus fly ash) has the lowest coefficient 3.1 x 10-6 while Celocrete has the highest 5.7 x 106. All-cinder products (1C6 and 3C6) have low and equal coefficients of 3.4 x 106 while those mixed with expanded slag or sand have higher values (2CES6 = 4.3 x 10-6 and 5CS6 = 3.9 x 10-6 respectively). Sand and gravel products have higher values (4.9 and 5.2 x 10-6 ). These results are in agreement with the findings of other investigators, which were discussed at the beginning of this chapter, that the coefficient of expansion of the concrete depends on the coefficients of its constituents and their proportion in the mix. Table 27 also indicates that the coefficients of the saturated (not vacuum saturated) concrete is somewhat higher than those of the dry concrete. The relation of the coefficient of expansion and moisture content was also discussed in some detail at the beginning of this chapter. The difference is believed to be contributed

TABLE 27.* COEFFICIENT OF THERMAL EXPANSION (INCH PER INCH PER 1 F) X 10 Specimen No. ic6 2CES6 3C6 3SG6 4SG6 5CS6 6ES6 7W6 8CEL6 9S6 lOCA6 Type of Aggregate Cinders Cinders & Expanded Slag Cinders Sand & Gravel Sand & Gravel Cinders & Sand Expanded Slag Waylite Celocrete Air-Cooled Slag Cinders + Fly Ash Saturated Concrete Test Method #1 3.4 4.3 3.4 4.9 5.2 3.9 5.4 4.6 5.7 4.8 3.1 Oven-Dried Concrete Test Test Method #1 Method #2 2.8 2.9 3.7 3.9 2.7 4.1 2.6 4.0 4.3 3.2 4.5 4.2 4.9 3.8 2.5 3.5 5.1 4.3 5.0 4.6 2.5 * For detailed data see Tables 74, 75, and 76 in Appendix VII. mainly by the cement paste. It is sufficient to repeat here the findings of Mitchell, who observed that the coefficients of vacuum saturated and oven dried concrete are approximately equal and that it is higher at intermediate moisture contents reaching maximum at the "optimum" moisture content which is estimated between 80 and 60 per cent of vacuum saturation. Table 27 also indicates small differences

between the results of dry concrete as determined by the two methods of tests. This is believed to be caused from using different methods of test and from the fact that the specimens were taken outside the oven for length measurements in the first method which was not necessary in the second method of test.

CHAPTER XIII THERMAL CONDUCTIVITY A highly desirable characteristic of hollow concrete masonry, especially that made with lightweight aggregates, is its property of heat insulation. Concrete itself possesses a variable resistance to the flow of heat and this resistance is comparatively larger for that made with lightweight aggregates. Some of the latter concrete possesses sufficient insulation that it is now used in building construction specifically where such a property is considered desirable. In solids, heat is transferred mainly by conduction, defined as the flow of heat from one part of the body to another part under the influence of temperature gradient without significant displacement of the particles. The term "steady state" is used when the flow of heat is such that the temperature at any point in the body does not vary with time. The rate of flow of heat through a homogeneous and isotropic solid depends upon the temperature gradient and the thermal conductivity coefficient of the solid. The latter is a physical property of the material itself, It is defined as the time rate of heat flow, in a steady state, through a unit area of the homogeneous substance when acted upon by a unit temperature gradient in the direction perpendicular to the area. In engineering applications, the coefficient is most commonly measured as the number of British thermal units transferred per hour per square foot of area perpendicular to the direction 196

-197 - of heat flow per unit temperature gradient. The unit temperature gradient is expressed as the degrees Fahrenheit per inch of length of heat path. This will be abbreviated65 as "Btu per (hr)(sq ft) (OF per in.)" and is commonly denoted by "k". For wall sections that are not homogeneous, i.e., consisting of two or more different materials or containing air cavities, the time rate of heat flow is usually measured for the entire thickness of the wall. This coefficient is termed the "thermal conductance" and is measured by the number of Btu transferred per hour per sq. ft. of area perpendicular to the direction of heat flow per unit temperature difference between the two surfaces of the wall. This is usually abbreviated as "Btu per (hr)(sq ft)(l F) and denoted by "C". When the rate of heat flow between air on the inside and air on the outside of a wall is considered, the term "overall coefficient of heat transmission" (sometimes called "thermal transmittance") is used. It is measured by the number of Btu transferred per hour per sq. ft. of area perpendicular to the direction of heat flow per unit temperature difference between air inside and outside of the wall and is usually denoted by "US. The thermal conductivity of homogeneous materials is found to be affected mainly by density, moisture content, and the mean value of the temperature gradient, usually called the "mean temperature". Since concrete is actually a coalesence of different materials, its conductivity is further dependent upon the individual conductivities of the constituents; size distribution of the aggregates, mix proportion, water-cement ratio, etc. In concrete masonry construction,

-198 - the conductance is further affected by the geometrical design of the units, method of construction, property of the surface or surface finish...... The relation of the thermal conductivity of concrete to the factors mentioned above are not yet formulated to the degree of accuracy known for many of concrete's other physical properties. This is partially because of the number of variables involved and also because accurate determination of the coefficient by the usual methods of test requires expensive apparatus and meticulous effort. However, various important studies have been conducted in the past for determining the currently used thermal coefficients. Some of these will be discussed later. Although various methods of tests have been used, the principle involved in determining the thermal coefficient is the same. It consists of supplying a known amount of heat energy at one face of the specimen or built-up wall section while keeping the other face at lower temperature and then measuring the resulting temperature gradient between the two faces at steady conditions. The coefficient can then be calculated from the amount of heat supplied per unit time and the observed temperature gradient. The heat may be supplied by a specially designed electrically heated plate kept in contact with one face of the specimen, while the other face is cooled by being in contact with another plate with internally circulated water at lower temperature. The preceding method is commonly known as the "hot plate" method. The conductance of a wall section is usually determined by the "hot box" method. In the latter method, the wall section is built

between two chambers, one of which is heated by electrical heating elements while the second is kept at lower temperature by refrigeration coils or equivalent equipment. In some cases, the hot chamber is built entirely from the test materials. Other special methods have been used for determining the conductivity of water saturated concrete. However, all the methods follow the same principle described above. A brief review of the more important past research for determining the conductivity or the conductance coefficients of concrete under steady state of heat flow will follow. To facilitate comparison, all the coefficients presented will be in corresponding units as given at the beginning of the chapter. Norton's research at the Massachusetts Institute of Tech66 nology was among the earlier works on determining the thermal conductivity of concrete. Two aggregates were tested, cinders and stone. Three different methods were used: (1) The hot plate method, for low temperatures, consisting of concrete slabs which were applied to the faces and edges of thin electrically heated plates. The two outside faces of the concrete parallel to the hot plate were cooled by copper plates with internally circulated water. Since this apparatus is not insulated, it was necessary to run the apparatus several days before the steady state was attained. (2) For high temperature determinations, the specimens were cast in the form of a cylinder. The heat was supplied by passing a heavy current through a steel bar in the central axis of the cylinder. A calorimeter was then fitted closely over the concrete cylinder for measuring the heat flow. (3) For thicker sections, the hot box method was used. Cubical 36-inch concrete boxes of various wall thicknesses were made. During

-200 - the test, the box was heated with electrical heating elements sealed in the inside. The outside faces of the box were kept at room temperature. The conductivity coefficients were found to be 2.33 for cinder concrete and 6.25-8.42 for stone concrete, depending upon the mean temperature. Willard and Litchty, at the University of Illinois, employed the hot box method. The coefficients were determined for one concrete mix (1-2-4) at steady heat flow, and for still or moving air on the outside surface of the box. The speed and relative humidity of the air were varied. The box was shaped in a hollow column. For the moving air test, a hood was placed over the box and connected to an air duct at the top. A conductivity coefficient of 8.3 was assigned to the still air condition; other values resulted from the moving air tests. It was found that the conductance coefficient increased by an increase in speed of air or a large increase in the relative humidity. Other early investigations at the University of Illinois 68 were conducted by Carman and Nelson on dry concrete of various mix proportion. A special method called the "cylinder method" was used. Concrete specimens were cast in forms of 7-1/2 x 24 inch cylinders with a 1-1/2 inch hole along the central axis. Heat was furnished by an electrical coil placed in this hole along the entire length of the cylinder. No cooling was provided on the outside surface of the specimen. No insulation was provided; the heat flow was assumed radial near the mid-point of the length. At steady state, temperatures were measured near the heating coil, near the surface and at a mid-point in between. The investigation also included the variation of the

-2:01 - conductivity with the mean temperature, age of the specimen, mix proportion and the water-cement ratio. The authors suggested a value for k equal to 10.84 to be used as a practical value for sand and gravel concrete. 69 Houghten and Gutberlet studied the effect of age on the conductivity of concrete. A "double guard" hot plate apparatus was used. The plate was 2 feet square including the guard ring. The test specimen was inserted only on one side of the hot plate; the other side of the hot plate was kept in contact with an "auxiliary'" 3I:te itAch was kept at the same temperature as the hot plate thus acting as a double guard. A cold plate was placed on the opposite face of the specimen. Two specimens of different concrete mixes were tested at curing periods as much as 1300 days during which they were stored in the laboratory air. No allowance was made for the change in the moisture content of the specimen during the storage. The authors concluded that the conductivity coefficient of concrete decreased with time and changed most rapidly during the first 30 days. They also suggested a value for k equal to 12 for practical application. The U, S. Bureau of Reclamation investigated the thermal properties of concrete made with aggregate similar to those used in the Hoover Dam construction. Concrete was tested in saturated condition to simulate service conditions. The method of test, as reported by Rippon and Snyder,70 consisted of molding 8 x 16 inch cylindrical specimens with a hole, 1-1/2 inch diameter, along the central axis. An immersion heater was placed in this hole and the remaining space sealed with water. The cylinder, thus assembled, was then placed in a water bath of constant but lower temperature.

-202 - The current passing through the immersion heater and the temperatures at the.inside and outside surfaces of the specimen were measured at steady state for calculating the coefficient. The results indicated that the conductivity coefficient varied with the aggregate and its mineral composition. Probably the most comprehensive research on the conductivity of building materials are those done by Rowley and Algren at the University of Minnesota (see references 71, 72, 73). Reliable apparatus was developed by which the coefficients of thermal conductivity of the materials and the overall heat transmission coefficients of built-up walls were determined for a large variety of materials including plastic and "dry-tamped" concrete slabs and walls and concrete masonry made with various aggregates including sand and gravel, slags, cinders and Haydite. The study also included determining the coefficients of surface conductance in still and moving air, at various speeds and directions, and the conductance coefficients of various types of cavity construction. The effect of many factors, such as type of construction, moisture content and mean temperature on conductivity were also investigated. Both the guarded hot plate and the guarded hot box methods were used. The hot plate apparatus, used for homogeneous slabs having smooth surfaces, consisted mainly of a hot plate, two cold plates and the insulating frame. The electrically heated plate was one foot square including the central heater, 9 x 9 inches, and the guard ring, 1.5 inches wide including 1/8 inch air gap. The cold plates were kept at the required temperature by internal circulation of water. The guarded hot box, used for built-up wall sections, consisted of inner, or test, chamber and outer, or guard, chamber which

-205 - were built into one side of a cold storage room. This "double chamber" was built of insulating materials with one side omitted and was so constructed that the open face was in the same plane as the opening in the cold room wallo The wall under test was placed over the open side of the double chamber thus cutting off completely the passage of air between the test and guard chambers which were maintained at the same temperatures during the test. The findings offered were numerous and the coefficients determined were widely adopted and used by designers. Among the values given for the conductivity coefficients of concrete were the following; sand and gravel k = 12o4 to 13.0, cinders k = 4.45 to 5.75, and Haydite k = 3.73 to 4.03. The most comprehensive research on the coefficient of thermal conductivity of concrete made with lightweight aggregates was probably that sponsored by the Housing and Home Finance Agency and conducted separately by the Bureau of Reclamation (BR) and the National Bureau of Standards (NBS) (references 19, 20, 21). This research included poured concrete of various mix proportions made with a large number of aggregates including sand and gravel, expanded clay and shale, expanded slag, pumice, etc. Concrete specimens were tested dry at mean temperatures of 27, 83 and 130 F by the BR and 119 F by the NBS. The guarded hot plate method was used. The BR employed a circular plate 9 inches in diameter while the NBS employed an 8 inch square plate; both sets of apparatus conformed with the requirements of the ASTM Designation C177-45 (Standard Method of Test for Thermal Conductivity of Materials by Means of the Guarded Hot Plate).22 Results obtained by the two laboratories for concrete made with a given aggregate differ somewhat because of different sources

of materials, concrete mix proportions and method of manufacture, and because of different mean temperatures of the test. However, values of the following order of magnitude were found for the conductivity coefficients: sand and gravel k = 6.59 expanded slag k = 1.5 to 3.2, expanded clay and shale k = 2.9 to 4.0. It was also found that for lightweight aggregate concrete, good correlation exists between the unit weight and the thermal coefficient. The present author's investigation constitutes a comparative study of the coefficients of thermal conductivity for products made with both dense and lightweight aggregates. Guarded hot plate apparatus, designed and constructed at the Department of Mechanical Engineering of the University of Michigan was usedo The description of the apparatus, after undergoing certain improvements prior to the present tests, will follow later. A. Brief Theoretical Analysis. Fourier74 has stated that the instantaneous rate of heat flow, d, under the influence of a temperature gradient, is equal to the product of three factors: (1) the coefficient of thermal conducd.T tivity (k), (2) the temperature gradient (-3Tx) which is the rate of change of temperature (T) with respect to the length of heat path (x) (negative in the direction of heat flow), and (3) the area of the section (A) measured normal to the direction of the flow. Thus the mathematical expression of Fourier's law is as follows: dQ -kA dT (17) dt dx

-205 - For steady state, the rate of heat flow is constant, thus~ dQ = Q usually denoted as q = rate of heat flow in dt t 'Btu per hour. For unidirectional heat flow through a homogeneous and isotropic body the temperature gradient (- - ) is, dT T2 - T1 dx L Where, T2 = temperature of the hot side of the specimen. T1 = temperature of the cold side of the specimen. L = average thickness of the specimen in the direction of the heat flow. For the conditions given above, equation (17) will therefore become as follows: T2 - T1 q= k A( - (18) or by using the units given at the beginning of this chapter: k = -Ta ( (18a) T2 ') A (- ) = Btu per (hour)(sq. ft )(deg. Fahro per inch of thickness). or,qL k= A (1'2 T1) (18b) Equation (18b) above was used for calculating the coefficients of thermal conductivity in these tests. It is apparent that it is required to measure the ate of heat supplied by the central heater of the hot plate (q), the area of the central heater (test

-206 - area) (A), the thickness of the specimen (L), and the temperatures at the hot and cold sides of the specimens respectively, (T2 and T1). Since L and A are constants, it is required only to measure q, T2 and T1 instantaneously after realizing steady conditions. The temperatures can be measured by thermocouples, as will be seen later. The rate at which heat is supplied by the heater (q) can be calculated from the electrical power input. The latter is equal to the product of the voltage (V) and the current (I) which were measured instantaneously by a voltmeter and an ammeter respectivelyo Further discussion of the calculations of the results will followo B. Test Specimeno Test specimens were cut from solid concrete masonry units (slabs) having the modular dimensions 3-5/8 x 7-5/8 x 15-5/8 inches (see Chapter II). Since the hot plate assembly is 12 inches square, it was necessary to use two pieces to form one specimen having this area. Each piece was cut slightly larger than 6 x 12 inches and then ground on a large commercial grinding machine to this dimension and to the same thickness so that the two pieces forming each test specimen were identical, and the two test faces of the resulting specimen were plane. The specimen thus prepared had its surfaces in complete contact with the hot and cold plates during the test. After grinding, the pieces were cleaned thoroughly and dried in the oven at 210 F to a constant weight. Two test specimens thus prepared were used per product; each was placed on either side of the hot plate during the testo

C. Test Apparatus. As was stated earlier, the guarded hot plate method was used for this test. The apparatus consists of the hot plate cabinet, the control board, the source of power and the electrical instruments for the necessary measurements, All are shown in Figures 41 and 44. The hot plate consists of a central heater 9 inches square surrounded by a guard ring heater 1-7/16 inches wide. The inside edge dimensions of the ring are 91l/4 x 9-1/4 inches which leaves an air gap of 1/8 inch around the central heater The heater windings75 of the two parts are also separates each is insulated with a sheet of mica and protected with sheet copper. The copper sheath is clamped on each side over the mica sheets, forming a rigid plane unit, 12 inches square, which is called the "hot plate"o Sectional plan of this plate and other details are also presented in Appendix VIII. The two cold plates are identical. Each consists of a hollow box having outside dimensions of 12 x 12 x 3/4 inches, made with copper plates 1/4 inch thick. Cooling is accomplished by circulation of tap water through the plates, The hot and cold plates are housed inside a wooden cabinet which is insulated on the outside by a 1-1/2 inch cork board with transite lining. The inside width and height of the cabinet are 12 inches. The hot plate is held permanently at the center vertical plane of the cabinet. The cold plates are placed parallel to the hot plate. The top of the cabinet and its sides parallel to the plates are removable to facilitate the loading of the specimens. The cabinet is also provided with a device for clamping the plate assembly

-208 - together, consisting of two threaded rods placed horizontally in both sides of the cabinet at about mid-height (see Figure 41). The control panel is shown in Figure 41 and, separately, in Figure 43, where the circuit components and the instrument connections to binding posts are clearly labeled. Separate controls are provided for the central heater and the guard ring circuits, as may be seen in Figure 43. The central heater circuit control, across the top of the panel, consists mainly of three continuous rheostats, connected in series, having the following total resistances: No. 1 and 2 - 500 ohms each; No. 3 = 50 ohms. The guard heater circuit controls, placed on the lower line consist of the two continuous rheostats, also connected in series and having the total resistances of 250 ohms for No. 1 and 10 ohms for No. 2. The current enters the control panel at the lower left binding post marked "input". It is then divided between the two circuits at the right, passing through the rheostats and the switch to the binding post (marked "output") at left, where the measuring instruments are connected as labeled. Separate connections were made to the central heater and to the guard ring heater (see Figure 41). Measurements made include the voltage and the current of the central heater circuit, and the voltage of the guard circuit. The instruments are shown in Figures 41 and 42. They consist of two portable Weston D. C. voltmeters of 300 volts capacity, graduated to 1 volt, for measuring voltages of the two circuits. For measuring the current of the central heater, a Weston portable D. C. millivoltmeter, connected in series with a shunt was used; this instrument is

-209 - readable to 1/2 millivolt which provides measurement of the current to the nearest 0.01 amp. Temperature measurements were made by thermocouples which were manufactured of thin (No. 30 A.oWG) copper-constantan wires fused together at the hot junctions. The hot junctions of the couples are placed on the surface of the specimen while the other ends are connected to a switch box (see Figure 73 in Appendix VIII and Figure 41) which is in turn connected to a temperature indicator, The latter is a portable Leeds and Northrup potentiometer graduated directly in degrees Fahrenheit. Cold junctions were not used because the instrument compensates internally for room temperature. This compensation and the potentiometer balancing are accomplished manuallyo Source of power is a motor-generator set which is shown in Figure 44. The motor ( at right in the figure ) is three phase 15 horsepower, operating on alternating current of 220 volts, 60 cycles, which is obtained from the University lineo The output of the generator is 250 volts D, C. which is reduced to 110 Volts by a rheostat at the operating panel of the set (power panel), see Figure 44, The capacity of this set is much larger than that required; however, it provides a steady current, free of fluctuations that may cause unsteady heat flow which might cause erroneous results. D. Procedure of Test. The specimens were prepared as described in part B above. Before placing the thermocouples, the location of the test area (the 9-inch square opposite the central heater) and the points where the temperatures are to be measured were marked on the face of the specimen, The hot junction of a thermocouple was then placed on each of

-210 - Figure 41. General View of the Hot Plate Apparatus. Figure 42. Close View of Measuring Instruments and Switch Box for the Hot Plate Apparatus.

-211 - Figure 43. Control Panel for the Hot Plate Apparatus. Figure 44. Source of Electrical Current for the Hot Plate, Motor-Generator Set.

-212 - these points and the wire was held in place with masking tape. Ten thermocouples were installed on each specimen; six of these were placed on the face adjacent to the hot plate, and the remaining four were placed on the face adjacent to the cold plate. An equal number of thermocouples were placed on the companion specimen at the opposite side of the hot plate. The total number of thermocouples used was therefore 20. They were located as shown in Figure 73 in Appendix VIII. After placing the thermocouples, the two companion specimens were "loaded". This was accomplished by lifting the top panel of the apparatus, loosening the clamps and sliding one side of the apparatus including the cold plate away from the hot plate to provide room for the specimen. The specimen was then placed vertically in the cavity between the hot and cold plates; thermocouple leads were brought to the top; sheets of blotting paper, cut to face dimensions of the specimen, were slipped on the two sides of the specimen, and the side panels of the apparatus including the cold plates were pushed into contact with the specimen. The same procedure was used for loading the companion specimen on the other side of the apparatus. The assembly was then clamped tightly together as may be seen in Figure 45. The thermocouples were then connected and tested to see if any were damaged during the loading, The thermocouples of the central heater were connected and read separately and not differentially as has been sometimes done. Around the movable sides of the apparatus or any opening where heat loss was suspected, rock wool was carefully packed. A blanket of this material was placed on the top before inserting the top panel. Prior to starting the test the motor-generator

-213 - set was warmed up. The measuring instruments were then connected to the control panel as may be seen on Figures 41J and 42o The apparatus now was ready for the test, in which the following procedure was used. The general procedure is in agreement with the ASTM Designation C177-45 (Standard Method of Test for Thermal Conductivity of Materials by Means of the Guarded Hot Plate).2 Tap water was fed swiftly through the cold plates and was adjusted by two individual valves so that equal amounts ran through the two plates. For safety, the total resistance of the rheostats of each of the central heater and the guard heater circuits were inserted before the current was allowed in each circuit, The voltage was immediately adjusted to that required by removing some of the resistance from the circuit. The actual voltages required for the central heater and the guard to provide a quick and uniform heating differ according to the specimen tested; however, for most specimens 100 and 50 volts were used respectively. Smaller voltages were used for some of the lightweight products, but the 2.l ratio above was found satisfactory. During the heating period, temperature measurements were taken every fifteen minutes. When the required temperature of the hot plate was approached, but not quite reached, the voltages of the central heater and the guard were dropped to values which kept the temperature of the plates (heater and guard) at the level desired. The voltage required depends on the temperature and the type of concrete tested, and was determined by pilot testso After setting the voltages at the "running" values, temperature measurements were made at 15-minute intervals for the first two hours, during which current adjustments, if necessary, were made in small increments to bring the

214 Figure 45. Looking Inside the Cabinet with Hot Plate, Cold Plate and Specimen in Between, Clamped Together.

-235. temperatures of the central heater and the guard into equivalence. When this was accomplished, the apparatus was kept running while temperature measurements and the necessary adjustments were made at one-hour intervals to refine the temperature balance. After this the test continued for at least five hours, without current adjustment, before the final readings were taken. At the latter period the temperatures of the hot plate (heater and guard) and the cold plates were measured frequently to assure continued equilibrium. The conductivity coefficient was calculated from the final readings which included careful determination of the central heater input (voltage and amperage) and the temperatures of the hot and cold plates. The total time required per run was 15 to 18 hours, differing somewhat according to the type of concrete. Two runs were conducted per product at different mean temperatures. E. Calculations. In part A, the coefficient of thermal conductivity (k) was defined in equation (18b) as follows: qL k = A (T 2 T1 (18b) Where q is the rate of heat flow (Btu per hour) But, q = 3.412 x P = 3.412 x I x V (19) Where, P = electrical power (watt) I = rate of flow of the current (amp) V = emf (volts) L = length of heat path (inch)

-216 - Now, in equation (18b), A 2 (9 X 9) = 1.125 sq. fto total test area for the two companion specimens. Substituting above values of q and A in equation (18b)~ 3.412 x I x V x L k = 1.125 (T2- T1) or k 3033 x I x V x L(20) T2 - T1 (20) Results were calculated by equation (20) for two runs per product, shown in Table 28. In this table, the current input of the central heater (columns 2 and 3) were taken from Table 77 (Appendix VIII) which will be discussed in part F. The temperatures T2 and T1 in Table 28 are respectively the average temperatures of the hot sides and the cold sides of specimen No. 1 (columns 5 and 6) and No. 2 (columns 7 and 8) as has been discussed in part A. T2 is the average temperature indicated by four thermocouples located in the test area of the hot sides of each of the companion specimens; T1 is the average temperature indicated by four thermocouples located in the cold side of each of the companion specimens. The individual readings of the thermocouples, for all the specimens, are given in Table 77. The temperature difference (T2 - T1) in the calculations (column 9 of Table 28) is the average value of the two companion specimens. F. Discussion of Test Results. The final current and temperature determinations for the two runs of test of each product are given in Table 77 in Appendix VIII. These were determined (see part D) at the end of at least five hours of constant and steady state o.f heat flow in each test,

TABLE 28. CALCULATION OF THE COEFFICIENT OF THERMAL CONDUCTIVITY (k) BTU PER (HR)(SQ. FT.)(DEG. F/INCH) 1 2 3 4 5 6 7 8 9 10 11 Average Average Temp. Thickness Average Temperature Average Temperature Difference Current of Specimen #1 Specimen #2 for Both Average of Sample Input Specimens OF OF Specimens Two Runs No. Volt Amp. #1 and #2 Hot Side Cold Side Hot Side Cold Side OF k k (L) inch. T2 T1 T2 T1 (T2 - T1) 1C10 65.39 65.39 3.55 3.55 123.8 124.4 123.3 133.8 56.1 56.5 56.5 56.5 124.5 125.6 124.0 135.0 57.0 57.3 56.5 57.1 67.6 67.6 67.1 76.7 4.04 4.04 4.07 4.13 4.04 4.10 2CES10 65.39 70.42 3C10 65.38 70.42 3SG10 75.43 80.48 4SG10 80.48 80.48 5CS10 60.36 70.42 6ES10 55.33 60.36 7W10 50.30 60.36 8CEL10 50.30 60.36 9S10 60.36 70.42 3.55 3.55 3.55 3.51 3.43 3.43 3.43 3.53 124.6 139.3 109.8 117.0 114.3 115.1 115.5 130.4 127.0 136.5 113.1 134.6 119.6 146.3 121.5 139.3 104.8 134.0 56.0 57.3 64.3 64.5 59.6 60.3 85.5 60.5 53.0 52.5 59.6 55.5 52.0 53.5 57.0 55.8 57.8 60.0 126.0 139.0 110.6 117.6 115.0 115.9 116.3 131.1 127.8 136.6 114.1 136.6 120.4 147.4 121.5 138.6 57.5 57.5 64.5 65.1 60.0 61.0 58.5 59.3 52.5 52.0 60.4 56.0 52.3 55.5 57.0 55.3 58.5 60.5 68.6 81.8 45.8 52.5 54.8 54.9 57.4 70.9 74.6 84.3 60.0 79.9 67.9 91.3 64.5 83.4 3.88 3.87 7.49 7.88 7.54 7.53 4.01 4.42 2.53 2.67 2.60 2.81 2.30 2.46 3.59 3.77 3.88 7.69 HJ 7.54 4.22 2.60 2.71 2.38 3.68 lOCA10 55.33 3.51 70.42 105.3 134.3 46.9 73.9 4.12 4.24

-218 - which was proved by frequent test observations during the period. Complete data, including these observations and other readings made earlier in the test for current adjustments, consists of over 400 determinations per run and are therefore not given since they are not necessary for the calculations of the results. Further discussion of Table 77 is given in Appendix VIII. Calculations of k were made by equation (20) separately for each run and are presented in Table 28, which was discussed in part E. The results of each product are averaged in column 11 of the table. The average results show that (1) sand and gravel products have k values equal to 7.69 and 7.54 for 3SG10 and 4SG10 respectively, which are much higher than those for lightweight aggregates; (2) cinder products have values of k averaging between 4.04 and 4.22 which varied in proportion with their unit dry weight (see Table 29); (3) expanded and air-cooled slag products have lower values of k, equal to 3.68 for air-cooled slag (9S10) and average between 2.38 and 2.71 for expanded slags. These also vary in proportion to their unit weight. It should be noted that the average values of k discussed above do not correspond to the same average mean temperature although, as may be seen in Table 29, the difference between these temperatures is comparatively small. It may also be seen in Table 28 that, for most of the products, the average temperature (column 9) in the second run for the product is made higher than that for the first run in order to study the effect of increase in temperature on k. However, these temperatures are also not equal for all the products, although they are all in the same order of magnitude as that met by concrete

-219 - walls in normal service. The main reason is that it was not possible to keep the mean temperatures or the temperature gradients uniform for all the products is that the city water cooling the cold plates (see part D) was not maintained at a constant temperature. Nevertheless, it can be seen in Table 28 that values of k for a certain product are increased by increasing the mean temperature. This relation was also noticed by many investigators as was discussed at the beginning of this chapter. The only correlation observed between the thermal conductivity and the remaining physical properties of the concrete products, as determined in previous tests, is that between the thermal conductivity and unit dry weight. For products of the same types of aggregates, k is proportional to the unit weight. This can be seen in Table 29 in which the unit weights were calculated from the weights of the specimen before test. The latter weights and the weights after test are given in Table 78 in Appendix VIII. Correlation between the thermal conductivities and unit weights of lightweight aggregates were also reported by many investigators (see references 19, 20, 21, 65 and 73); others (reference 76) have indicated that this correlation is peculiar to concrete made with lightweight aggregates and that it does not always exist for that made with dense aggregates. As was mentioned in the brief historical notes, some investigators found that the coefficient increases with the age of concrete. This relation cannot be determined for the present products because the initial ages of the concrete are not known (see Chapter II). However, all the products were stored over two years in the normal laboratory air before being tested.

-220 - TABLE 29. COEFFICIENT OF THERMAL CONDUCTIVITY (k), AVERAGE MEAN TEMPERATURE, AND UNIT DRY WEIGHT (Waylite and Celocrete are the commercial names for expanded slag. ) Unit Average Sample Dry Mean Average No. Aggregate Weight Temp. k p.c.f. OF lC10 Cinders 89.9 90.7 4o04 2CES10 Cinders & 91.0 92.9 4.10 Expanded Slag 3C10 Cinders 88.1 94.7 3.88 3SG10 Sand & 129.1 89.0 7.69 Gravel 4SG10 Sand & 125.4 87.6 7.54 Gravel 5CS10 Cinders 93.2 91.2 4.22 & Sand 6ES10 Expanded 90.1 97.7 2.60 Slag 7W10 Waylite 100.0 91.2 2.71 8CEL10 Celocrete 82.8 93.4 2.38 9S10 Air-Cooled 116.9 93.2 3.68 Slag 10CA10 Cinders & 91.5 89.4 4.18 Fly Ash It should also be borne in mind that the results given were determined for oven dry concrete and still air conditions. Presence of moisture in concrete or motion of air across the surface, such as may be met in service markedly increases the rate of heat flow. The amount of this increase depends on the moisture content of concrete and the speed of air. This was discussed briefly at the beginning of the chapter.

CHAPTER XIV SUMMARY The tremendous growth in the use of concrete masonry units made with lightweight aggregates in the last decade has made desirable the accumulation of data regarding the performance of these building materials in order that intelligent decisions can be made regarding their proper use. This study attempts to fulfill this need in the Michigan area for a cross section of the products contemporarily available. In this concluding chapter, effort is made to provide reference to the salient features of the test results on masonry units obtained from representative dense and lightweight aggregate products. The overall average results of most of these tests are presented in Table 30 at the end of this chapter, with a bar diagram (Figure 146) for comparative study. Important conclusions are here summarized, (1) The adoption of the system of modular dimensions by the industry is evident in the results of the tests. Three products have units with deviations exceeding that permitted by the current ASTM standards. However, no single specimen of a given product showed a deviation from its own average exceeding the permitted value. (2) Weights of 8 inch masonry walls containing the lightweight aggregate units range between 29 and 36 pounds per sq. ft., and for the dense aggregate between 43 and 48 psf. (3) Compressive strength of the lightweight aggregate masonry units generally varied between 58 and 88 per cent of the strength of the dense units; no correlation was found between weight 221

and strength. Three lightweight products averaged just below the minimum ASTM requirement for load-bearing units (700 psi). (4) Lightweight aggregate and air-cooled slag products are more absorbent than sand and gravel products. This is an adverse property for lightweight masonry exposed to the weather. Cinders have the highest absorption followed by the slag products. Some correlation exists between the compressive strength and the absorption. High strength usually accompanies low absorption. (5) The moisture contents of the products when sampled varied between 19.4 and 62.6 per cent of the 24 hour absorption. Six products exceeded the 40 per cent maximum allowed by the ASTM specifications. The latter requirement is considered of value in limiting shrinkage and subsequent cracking which may result from using units of high moisture content. Most of the products are stored in open air and, therefore, their moisture content varies from time to time. Sampling was done during the relatively damp spring season, (6) There was no sharp line of demarcation between the dense and lightweight aggregate products with regard to the total shrinkage or the shrinkage-moisture loss performance. For a given product, and between products, the shrinkage experienced is not predictable from the actual moisture content. Generally, the shrinkage tended to be greater for those lightweight aggregate products (not autoclaved) having higher absorption or for those requiring higher cement content to obtain the required strength. Cinders and expanded slag products (except Celocrete) undergo similar shrinkages and are much higher than those of the air-cooled slag producto No apparent correlation exists between shrinkage and the unit weight, strength

-225 - or modulus of elasticity. However, it is suggested that lightweight aggregate products, because of their lower modulus of elasticity and reported higher creep, have less tendency for cracking. High-pressure steam curing (autoclaving) resulted in lower shrinkage. (7) Staining and popouts were determined by visual examination of the masonry units after 26 months of moist storage. One product (cinders) showed a popout believed to be caused by hydration of free lime. Three other cinder products showed stains that are apparently caused by oxidation of tramp iron particles. Stains and popouts did not appear on faces painted with a cement base paint indicating that the paint may have given at least temporary protection. (8) Flexural strength of dense aggregate products are 1.6 to 3 times higher than those of lightweight aggregates. However, it has been suggested that the flexural strength of an unreinforced masonry wall depends largely on the strength of the mortar joints. Very poor correlation exists between the compressive strength and the modulus of rupture possibly because of the differences in the particle shapes and surface texture existing for the wide range of aggregate types included in the present study. It is now well recognized that the flexural test is much more sensitive to the characteristics of the coarse aggregate than is the compressive testo (9) Elastic constants (Young's modulus of elasticity, E, modulus of rigidity, G, and Poisson's ratio, ) as determined by the static method are lower than those determined by the dynamic (sonic) method. This conclusion has been confirmed by all other investigators. Products made with dense aggregates have higher constants (E and G) than those made with lightweight aggregates. Young's moduli for

-224 - lightweight aggregate products range between 33 and 80 per cent of the lower value of the two sand and gravel aggregate products in the study. Rough correlation exists between the static modulus of elasticity and the compressive strength and is still better between products having similar aggregates. Rough correlation exists between the elastic modulus and modulus of rupture. Better correlation exists between the elastic modulus and the unit dry weight. (10) Lightweight aggregate products of highly cellular structure, unlike sand and gravel products, are vulnerable to a very limited number of accelerated freeze-thaw cycles when fully water saturated. This probably results from their high absorption and capability of retaining a large amount of freezable water in the open air voids in the porous concrete and distinct from the voids contained in the particles themselves. A special mix design was probably used to 'obtain this highly cellular concrete to promote light weight. However, the condition of full saturation is rarely met by masonry walls in normal service. A moisture content, of for instance, 40 per cent of the absorption (maximum allowed by the ASTM specifications) may therefore be established as a more appropriate limit. Tests on specimens sealed in copper jackets at the 40 per cent moisture content showed that lightweight products withstood 500 cycles of freezing and thawing well. Of the dense products, air-cooled slag exhibited inferior performance which was probably caused by deleterious particles in the coarse aggregate. However, this was not shown conclusively because poor performance of concrete cannot normally be assigned to a single cause, and a conclusive study would have required a larger program of tests than was possible. No correlation was observed between

-225 - durability and the other properties of the concrete, such as weight, strength, etc. (11) Coefficient of thermal expansion was determined for water-immersed and oven-dried conditions. The coefficients for the oven-dried concrete were approximately 10 to 20 per cent lower than for the water-immersed condition. This difference is believed to be contributed mainly by the cement paste. In general, the results indicate that cinders have a lower coefficient than sand and gravel or slag products. The latter two have values of similar magnitude. No correlation was observed between the coefficient of expansion and any of the remaining physical properties of the concrete. However, the coefficient appears to be influenced by the coefficients of its constituents and their proportion in the mix. This is observable for products made from a blend of two different aggregates, where the coefficients of concrete containing the individual constituents were determined. (12) Coefficient of thermal conductivity (k) was determined for oven-dried concrete by the hot plate method at mean temperatures between 88 and 98 F. The two sand and gravel products have (k) values of 7.7 and 7.5 Btu per (hr)(sq. ft.)(deg. F per inch), respectively, compared with (k) values for cinder products which averaged between 4 and 4.2 and the slags which averaged between 2~4 and 2.7' No correlation was observed between thermal conductivity and any of the remaining physical properties of the concrete except unit dry weight. On the basis of the previous summary it may be concluded that: Lightweight aggregate masonry units possess certain characteristics that are desirable in construction materials for types of

-226 - structures such as residential buildings, etc., where high strength is not of primary importance and where severe weathering in the presence of high moisture content is not normally encountered. The varied and appealing surface texture of these materials have architectural advantages. Their comparative lightness permits economy of construction by reducing the dead loads on other structural members. Their lower thermal conductivity results in economy in heating and air conditioning, and provides a more comfortable housing in hot seasons where artificial cooling is lacking. It has also been suggested that lower thermal conductivity may indicate a somewhat longer fire-endurance period for the masonry wall. Much work has been done to reduce the potential excess shrinkage of certain lightweight aggregate products which may cause subsequent cracking of the masonry; certain methods have been suggested and used with considerable success such as highpressure steam curing or drying of the units to a safe moisture content prior to use. It appears that the presently large and continuing growth in the use of lightweight aggregate concrete masonry units for housing and similar type construction is based on sound engineering and is well justified.

I (3 z -J LL 0 z Ir (U) FIELD MOISTURE CONTENT WATER ABSORPTION DRYING SHRINKAGE OVEN-DRY WEIGHT (NET VOLUME BASIS) 1500 I 1300 z Itj w 11 700 0 O 500 I PRODUCT I 2C 2H 3C 3S 4 5 6 7W 7B 8 9 10 COMPRESSIVE STRENGTH (GROSS AREA BASIS) z 0.08 0 0. PRODUCT I 20 3C 3S 4 5 6 7W 8 9 10 POISSONS RATIO (JL) STATIC (DASHED BARS FOR DYNAMIC VALUES) a. E 800 2400. 600 1800 aEr L. 400, f 1200 I IIIMI OF 3 (.100 o* I 600 o 4.o ~ 111 ~ 0 0 o w 0 I 2 3C 3S 4 5 6 7W 8 9 10 I 20 3C 3S 4 5 6 7W 8 9 10 MODULUS OF RUPTURE (R) STATIC YOUNGS MODULUS OF ELASTICITY E, AND MODULUS OF RIGIDITY G (DASHED) (C Z 80 32::)..I60:: 6.0 -Ii 20m m 2.0 LEGEND I = IC, CINDERS 2C - 2CES, CINDER ( EXPANDED SLAG,COMBINED 2H - 2H, HAYDITE 3C 0 3C, CINDERS 3S - 3SG, SAND t GRAVEL 4 - 4SG, SAND 4 (RAVEL 5 0 5CS, SAND (FINE).- CINDER COARSE 6 = 6ES,EXPANDED SLAG 7W- 7W,WAYLITE 78B 7BW,BESLITE t WAYLITE, COMBINED 8 - 8CEL,CELOCRET ( SAND 9 '9S, AIR-COOLED BLAST-FURNACE SLAG 10 IOCA, CINDER + FLY ASH z IL. 8.0 LL 6.0 I 4.0.Er 2.0 I i I ro -I\ I - w 3 II 111111 1 I 2 3 3S 4 5 6 7W 8 9 XC 30 36 4 5 6 7W 8 9 10 0 I 2C 3 3S 4 5 6 7W 8 9 10 I 2C 3C 3S 4 5 6 7W 8 9 10 FREEZE-THAW DURABILITY (500 CYCLES F * T) COEFFICIENT OF THERMAL EXPANSION FOR SATURATED t DRY CONCRETE (DASHED) COEFFICIENT OF THERMAL CONDUCTIVITY (K) FIGURE 46-OVERALL AVERAGE RESULTS OF TESTS ON CONCRETE MASONRY PRODUCTS

TABLE 30. SUMMARY OF THE OVERALL AVERAGE RESULMS OF TEM O: N CONCRETE PRODUCTS Unit Weight Absorption 24 Compress-ive Freeze-Thaw. Total fying D P,!oototool Wighottoo b of U.ttO of. hof O StCt O-gth Elootic Poooo-oioi,.tD.obil ty SoOioeopostioo 0A 7 5f00 & C ffto Ot Cf - l C-otiv Mot P.-od Mofobsip BOood B..od DyTootoi 73is( steti Mod. of Modolo mostobs ThebOO p btp-iobt Gross Dimensions Gross Net Gross Net Area of Per Content o on Modulus Modulus Poisson's ModulusModulus Poisson's Elasti- of Loss Water Oven Oven k of Volooo o 1- Ooto Amo A-. Dr AS Wato- Oveo- As Wt-sr Ovot- WOtoor C,.t.1 Goo Not Mdao.so of of (otis of of Uotio NOO- city 50ptCr, Sti.o- Shikus- Ooio o 33tt- Doy Doy i t 'I".. Pu y Pro- Type odular Cunits Cent Sam- Satu- Dry Sa- Satu- Dry per of Sampled- Area Area of p - of Elasti- Rigid- Elasti- Rigid- of After R ing age Loss rated hy;y.:eiht 0-i (Ohr) OSOO of thjLo- 0-1-h tOL of plod rmto plod -otod -.ft. Dfy pfr oob of of HbpoSro city Sty city ity Cyolor Toob, bPf!r toO p00000bt p-0 c- by 0000003-33-S obr (oqyt.) No. Aggregate ness Length Height Gross of Weight of todultr Modular R E G 0E G of Test Popouts of of Fxtenso- meter baoe (deo. t L h Ar0000 Coo-. Abeoop. 00it UDit _ _iti_1 I h Abrp Loo~tbr Soy 00000 to. to. to. i0 in" iOj t5 0 % pf 0 pr o c. ft. pcf i yoi yot poi 5o0 kSn kSn SoO 000031 psp % joobyof oobyoof pocI bC Cinders 7.70 15.68 7.74 935.6 477.2 121.0 61.7 51 26.8 29.3 25.1 97.1 106.1 90.9 15.1 16.6 41.9 698 1357 272 1002 430.16 91 396.16 500 96 423 Popout.049 92 3.4 2.8 2.9 *9.9 90.7 4.0 2CES Ciders & 7.67 15.63 7.67 913.4 517.3 119.8 67.5 56 29.7 32.1 27.9 99.1 106.7 93.1 14.0 15.0 43.7 1004 1778 224 1266 520.22 1017 426.19 500 o100 523 Stain.052 391 4.3 3.7 3.9 91.0 929 4.10 Expanded Slag 2H Hayd-te 7.68 15.65 7-67 922.5 515.2 120.3 67.2 55 27.8 29.4 25.1 93.1 98.7 84.2 14.1 17.3 62.6 1332 2316 -- -- -- -- - - -- - -- - None 30 Cinders 7.69 15-7 7.63 924.2 554.3 121.1 72.6 60 27.7 32.1 26.7 3883.5 100.1 83.4 16.8 20.2 19.4 754 1252 35S 904 400.24 832 360.15 500 90 567 Stain.031 93 3.4 2.7 26.,1 94.7 3.6 3SG Sand & 7.67 15.64 7.62 914.5 502.3 120.0 6509 55 39.6 41.1 38.2 136.3 141.2 132.3 8.5 6.4 45.4 12703 2212 510 2921 1260.16 2233 - - 500 o100 694 None.034 89 4.9 4.1.0 19.1 89.0 769 03 Gravel r\J 3G Saod & 7.72 15.74 7.71 937.6 469.4 121.6 60.9 50 36.4 38.3 35.5 134.1 140.8 129.2 10.3 7.9 37.5 994 1983 635 3674 1489.24 3061 1350.24 500 100 836 Noneo.046 87 5.2 4.3 0.7 12S. 776 7.5o U) Gravel 5CS Saod-Fine, 7.70 15.61 7.67 921.7 464.5 120.2 60.4 50 27.3 29.3 26.1 101.5 108.9 97.1 12.3 12.7 32.2 377 1728 344 0619 694.17 1325 451.17 500 97 425 Stain.046 92 3.9 3.2 3.5 93.2 91.2 o 02 Ciad,,,.6ES Expanded 7.70 15.65 7.70 927.7 459.0 120.4 59.5 50 28.8 30.9 26.9 100.4 116.3 101.3 14.9 14.7 48.6 695 1395 338 1792 775.16 1696 752.13 500 97 550 None.048 92 5.4 4.5 5.1 90.1 97.7 2.60 Slag 7W Waylite 7.68 15.61 7.67 921.3 474.4 120.1 61.8 51 27.3 29.7 26.7 99.3 108.3 97.3 11.0 11.3 19.7 777 1506 334 2412 977.23 2131 953.15 500 100 584 one.041 088 4.6 4.2 4.3 100.0 91.2 2.71 7BW Besl0te & 7-63 15.65 7.65 913.9 470.6 119.5 61.5 51 28.7 31.4 27.8 105.3 114.1 102.2 13.0 12.7 25.5 989 1923 - -- - - - - - -- - - None Wayllte, Combined 8CEL Celocrete 7.77 15.65 7.73 944.4 507.0 122.2 65.6 54 24.6 26.8 22.6 03.0 90.3 76.2 14.0 18.8 49.7 647 1U99 193 1313 559.10 1249 -- - 500 88 272 None.077 84 5.7 4.9 5.0 82.3 93.4 2.38 9S Air-Cooled 7.64 15.61 7.62 908.3 512.6 119.3 67.3 56 38.3 40.8 37.5 129.2 137.7 126.3 12.7 9.0 27.7 1243 2195 526 2562 1033.18 2250. 929.13 57 62 357 Mone.024 93 4.8 3.3 0.6 116.9 93.2 3.68 Slag 10A Cinders + 7.71 15.67 7.63 925.7 503.1 120.7 65.6 54 26.4 28.9 24.8 90.7 99.0 85.2 13.9 16.3 39.7 906 1663 277 1106 479.16 891 354.13 500 93 442 None.036 95 3.1 2.5 2.5 91.5 89.4 4.18 Fly Ash

APPENDIX I AGGREGATES A. Specific Gravity and Absorption. Tests for determining the specific gravity and absorption of coarse, fine and combined aggregates were discussed in Chapter III. Results were given in Table 6. The equations used for calculating these results will be presented here. Detailed data are given in Tables 31, 32 and 33. 1. Coarse Aggregate. The following formulas were used for calculations of results for coarse aggregate. Let A = weight in grams of B = weight in grams of C = weight in grams of w = unit weight of dry oven-dried sample in air saturated surface-dry sample in air sample in water, and compacted aggregate (in grams per cc). Then, Bulk Specific Gravity (oven dry basis) Bulk Specific Gravity (saturated surface-dry basis) A B-C B-C - B-C (21) (22) Apparent Specific Gravity Absorption (per cent of dry weight) Absorption (volume of water per unit volume of dry compacted aggregate, per cent) 229 A A-C B -A (B-A)w A (23) (24) (25)

-2350. 2. Fine Aggregate. It may be repeated that unlike coarse aggregates (see Chapter III), specific gravity and absorption of the fine aggregates were determined from two different samples which were both obtained from an original sample in a saturated surface-dry condition. For the specific gravity sample, let: a = weight in grams of oven dry sample in air, calculated from (b) (following)) and the absorption. b = weight in grams of saturated surface-dry samples in air. v = bulk volume (cc) of sample in the flask, which is equal to the difference between the final reading and the initial reading of the flask. For the absorption specimen, let A, B, and w be the same as for coarse aggregate. Then, a= 100 x b (26) 100 + absorption (% by weight) Bulk Specific Gravity (oven dry basis) = a(27) v Bulk Specific Gravity (saturated surface-dry = (28) basis) v Apparent Specific Gravity a (29) v-(b-a) Absorption (per cent of weight) = B-A x 100 A (B-A)w Absorption (per cent of volume) -- x 100 A

-351 - TABLE 31. CALCULATION OF SPECIFIC GRAVITY AND OF COARSE AGGREGATE ABSORPTION Aggregate 6ES19 7W9 7B9 8CEL19 ReSampl1e 1 2 1 2 1 2 1 2 marks A (gns) 730.0 892,0 1031.0 922.0 835.0 742.0 586.0 544.0 B (gns) 779.5 952o0 1109.0 992.0 953.0 847.0 729.0 675.0 C (gms) 284.0 347.0 412.0 370.0 392.0 352.0 120.0 104.0 B-C (gns) 495.5 605.0 697.0 622.0 561.0 495.0 609.0 571.0 A-C (gms) 446.0 545.0 619.0 552.D 443.0 390.0 466.0 440.0 B-A (gms) 49.5 60,0 78.0 70.0 118.0 105.0 143.0 131.0 B. S. Gi Oven Dry 1.47 1.47 1. 48 1.48 -I'49 1.50 0.96 o.98 Eq. Basis 21 Average 1.47 1.48 1.50 0.97 B. S. G. Saturated 1.57 1.57 1.59 1.59 1.70 1.71 1.20 1.21 Eq. Surface 22 Dry Basis Average 1.57 1.59 1.71 1.21 Apparent S. G, 1.64 1.64 1.66 1l67 1.88 1.90 1.26 1.28 Eq. Average 1.64 1.67 1,89 1.27 23 Absorption % of Dry 6.77 6.73 7.55 7.59 14.11 14.15 24.40 24.10 Eq. Weight 24 Average 6.75 7 57 14.13 24.25 B. Materials Causing Popouts of Concrete. Tests for determining the presence of deleterious materials in aggregate that are capable of causing popouts of concrete were discussed in Chapter III. Here, the detailed data used in calculating the concrete mixes and determining the results will be presented.

TABLE 32. CALCULATION OF SPECIFIC GRAVITY AND ABSORPTION OF FINE AGGREGATE A --- —-- -7I~-l --- —') --- —----;- -- _ --- ___- — I-._.- - Sample Initial Reading (cc) Final Reading (cc) U- o juy3Su 49ji u5 bES29 7W29 OCEL29 1OCA9 He1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 marks 0.10 0.20 0.20 1.10 0.30 0.60 0.40 0.70 0.10 0.30 0.50 0.70 0.50 0.80 0.20 0.70 0.20 1.10 - 19.10 19.10 18.70 19.50 19.10 19.35 19.00 19.30 19.40 19.00 19.20 19.50 19.50 19.40 20.80 21.20 19.10 20.10 v (cc) 19.00 18.90 18.50 18.40 18.80 18.75 18.60 18.60 19.30 19.20 18.70 18.80 18.50 18.60 20.60 20.50 18.90 19.00 a (ens) 32.32 32.26 32.14 32.08 49.26 49.25 49.40 49.35 32.35 32.35 33.30 33.30 33.30 33.29 30.59 30.59 32.29 32.28 Eq.26 b (gns) 35.00 35.00 35.00 35.00 50.00 50.00 50.00 50.00 35.00 35.00 35.00 35.00 35.00 35.00 35.00 35.00 35.00 35.00 B. S. G. 1.70 1.71 1.74 1.74 2.62 2.63 2.66 2.65 1.68 1.68 1.78 1.77 1.80 1.79 1.49 1.49 1.71 1.70 Eq.27 Oven Dry Basis Average 1.71 1.74 2.63 2.66 1.68 1.78 1.80 1.49 1.71 B. S. G. 1.84 1.85 1.89 1.90 2.65 262.6 9 2.69 9 1.81 1.82 1.87 1.86 1.89 1.88 1.70 1.71.85 1.84 Eq.28 (S. S. D.) Basis Average 1.85 1.90 2.66 2.69 1.82 1.87 1.89 1.71 1.85 Apparent 1.98 2.00 2.05 2.07 2.73 2.73 2.74 2.75 1.94 1.95 1.96 1.95 1.98 1.97 1.89 1.90 1.99 1.98 Eq.29 S. G. Average 1.99 2.06 2.73 2.75 1.95 1.96 1.98 1.90 1.99 B (gns) 455.00 423.00 512.00 578.00 580.50 699.00 947.00 872.50 503.00 461.00 406.00 305.00 506.00 482.00 500.00 460.00 380.50 431.50 A (gms) 420.00 390.00 470.00 530.00 572.00 690.00 935.00 861.00 465.00 426.00 387.00 290.00 481.50 458.50 437.00 402.00 351.00 398.00 B-A (gas) 35.00 33.00 42.00 48.00 8.50 9.00 12.00 11.50 38.00 35.00 19.00 15.00 24.50 23.50 63.00 58.00 29.50 33.50 Absorp. 8.30 8.50 8.90 9.10 1.50 1.50 1.28 1.32 8.20 8.20 5.10 5.12 5.10 5.13 14.42 14.40 8.40 8.42 Eq.24 % by Weight Average 8.40 9.00 1.50 1.30 8.20 5.11 5.12 14.41 8.41 ro

-234 - Data for the concrete mixes are given in Table 34. It may be repeated that although proportioning is done by volume, measuring was done by weight as shown in column 5 of Table 34. These proportions correspond to a constant ratio of 1:6, one part cement to six parts of a combined aggregate, by volume. The amount of water was determined from trial batches to give the slump shown in Table 34. Aggregates were combined in the same proportions as that used in the plant mixes, shown in column 3 of Table 34. The weight of the autoclave specimen before and after test is given in column 6 of Table 34~ It is obvious from the results that no loss in weight, waich migtht be caused by popouts or breakage of concrete, was suffered by the specimens; it was, therefore, concluded (see Chapter III) that the aggregates were free of deleterious popout materials.

- 235 - TABLE 34. POPOUTS TEST - LABORATORY CONCRETE MIXES 1 2 3 4 5 6.... Weight of Aggregate F. M. Laboratory Mix Concrete Spec. Type Proportion of w/C Water Cement Combined Slump Saturate, Specimen of for Combined by Aggregate Before After No. Aggregate Plant Mix Aggregate Volume gms rns gins inch Test Test gms gms 1C91 Cinders Combined 4.19 1.07 1016 982 3668 2.70 1093 1102 1C92 1081 1090 3C91 Cinders Combined 3.86 1.20 1116 982 3591 2.80 1121 1121 3C92 1124 1125 3SG91 Sand & Combined 3.67 1.23 851 768 5493 2.90 1667 1667 3SG92 Gravel 1682 1680 4SG91 Sand & Combined 3.32 1.31 703 725 5466 2.80 1698 1696 4SG92 Gravel 1700 1701 5CS91 Cinders 3100 lbs 4.41 1.15 961 896 3257 2.60 1117 1117 5CS92 & Sand 500 lbs 543 1144 1144 6ES19 Expanded 2500 lbs 4.78 1.04 811 896 3000 2.70 1316 1312 6ES29 Slag fine 450 lbs 600 1343 1344 coarse 7W91 Waylite 1867 lbs 3.88 1.19 763 768 1764 2.75 1024 1027 7W92 fine 933 lbs 882 1026 1030 coarse 7BW91 Waylite & 1000 lbs 3.84 1.20 1098 896 1387 2.80 1199 1200 7BW92 Beslite 1600 lbs 2219 1211 1210 8CEL91 Celocrete 1280 lbs 4.59 1.06 1397 1067 2296 2.75 925 923 8CEL92 & Sand Cel. fine 220 lbs 397 sand 320 lbs 574 935 936 Cel.coarse 10CA91 Cinders Combined 4.04 1.16 1024 939 3587 2.70 1068 1069 10CA92 1086 1086

APPENDIX II DIMENSIONS OF THE MODULAR UNIT,....-,.,.. _ Detailed data and calculations of the results are presented here. Definitions, procedure of tests, and discussion of test results are given in Chapter IVo In Table 35 data is given for determining the net volume of each modular unit, and the average results of 5 specimens for each product. Table 36 presents data for determining the dimensions (thickness, length and height), gross cross-sectional area and gross volume of units tested; average results for the 5 specimens used for each product are also given. Table 37 presents net cross-sectional area as percentage of gross cross-sectional area of each modular unit; average results for the 5 specimens of each product are also given. 236

TABLE 35. CALCULATION OF NET VOLUME OF THE MODULAR UNITS Wet Wt.- Wet Wt.Specimen Wet Suspended Suspended Net Net Specimen Wet Suspended Suspended Net Net No. Weight Weight Weight= Volume Volume No. Weight Weight Weight= Volume Volume Net Vol. Net Vol. gns gns cc. in3 ft3 gms gs cc. in3 ft 1Cl 13,365 5,585 7,780 474.77.275 6ES11 13,515 6,190 7,325 447.00.258 1C12 13,325 5,470 7,855 479.34.277 6ES12 13,725 6,255 7,470 456.00.264 lC13 13,435 5,485 7,950 485.14.281 6ES13 14,140 6,570 7,570 462.00.267 1C14 13,335 5,475 7,860 479.65.278 6ES14 14,050 6,475 7,575 462.30.268 1C15 12,990 5,335 7,655 467.14.270 6ES15 14,635 6,965 7,670 468.10.271 Average 7,826 477.21.276 Average 7,522 459.00.266 2CES11 14,910 6,315 8,595 524.50.304 7W11 13,520 5,735 7,785 475.10.275 2CES12 14,320 5,895 8,425 514.13.298 7W12 13,485 5,660 7,825 477.50.276 2CES13 14,655 6,260 8,395 512.30.297 7W13 13,735 5,940 7,795 475.70.275 2CES14 13,580 5,525 8,055 491.55.285 7W14 13,600 5,835 7,765 473.90.274 2CES15 15,270 6,355 8,915 544.03.315 7W15 13,135 5,435 7,700 469.90.272 Average 8,477 517.30.299 Average 7,774 474.40.275 2H11 13,325 4,825 8,500 518.70.300 7BWl 14,300 6,605 7,695 469.60.272 2H12 13,445 5,005 8,440 515.04.298 7BW12 14,380 6,610 7,770 474.20.274 2H13 13,240 4,805 8,435 514.74.298 7BW13 14,475 6,755 7,720 471.10.273 2H14 13,330 4,895 8,405 512.91.297 7BW14 14,435 6,705 7,730 471.70.273 2H15 13,450 5,015 8,435 514.74.298 7BW15 13,542 5,900 7,642 466.40.270 Average 8,443 515.23.298 Average 7,711 470.60.272 3C11 14,770 5,555 9,215 562.34.325 8CEL11 12,850 4,300 8,550 521.80.302 3C12 14,895 5,885 9,010 549.80.318 8CEL12 12,940 4,355 8,585 523.90.318 3C13 14,450 5,375 9,075 553.80.321 8CEL13 13,080 4,455 8,625 526.30.305 3C14 14,250 5,235 9,015 550.10.318 8CEL14 11,070 3,190 7,880 480.90.278 3C15 14,470 5,390 9,080 554.10.321 8CEL15 10,855 2,955 7,900 482.10.279 Average 9,079 554.30.321 Average 8,308 507.00.296 3SG11 18,370 10,275 8,095 493.99.286 9S11 18,270 9,975 8,295 506.20.293 3SG12 18,230 10,050 8,180 499.18.289 9S12 18,860 10,310 8,550 521.76.302 3SG13 19,130 10,760 8,370 510.77.296 9S13 18,740 10,280 8,460 516.26.300 3SG14 18,670 10,450 8,220 501.61.290 9S14 18,800 10,320 8,480 517.48.300 3SG15 18,745 10,455 8,290 505.89.293 9S15 17,965 9,750 8,215 501.31.290 Average 8,231 502.29.291 Average 8,400 512.60.297 4SG11 17,395 9,630 7,765 473.85.274 10CAll 13,165 4,960 8,205 500.70.290 4SG12 17,150 9,510 7,640 466.20.270 10CA12 13,060 4,825 8,235 502.50.291 4SG13 17,775 10,045 7,730 471.70.273 10CA13 13,010 4,810 8,200 500.40.290 4SG14 17,565 9,830 7,735 472.00.273 10CA14 13,070 4,810 8,260 504.06.292 4SG15 16,880 9,270 7,610 463.40.268 lOCA15 13,140 4,820 8,320 507.72.294 Average 7,696 469.40.272 Average 8,244 503.08.291 5CS11 13,465 5,870 7,595 463.50.268 5CS12 13,395 5,855 7,540 460.10.266 5CS13 13,190 5,615 7,575 462.30.268 5CS14 13,370 5,665 7,705 470.20.272 5CS15 13,300 5,655 7,645 466.50.270 Average 7,612 464.50.269 ~~~~~~.2. 9 -) -2

TABLE 36. CALCULATION OF GROSS AREA AND GROSS VOLUME OF THE MODULAR UNITS Thickness Length Height Gross Area Gross Vol. Gross Thickness Length Height Gross Area Gross Vol. Gross Specimen t L h t xL t x L x h Volume Specimen t L h t x L tx L x h Volume No. in. in. in. in2 in3 ft3 No. in. in. in. in2 in3 ft3 1C11 1C12 1C13 lC14 1C15 Average 2CESll 2CES12 2CES13 2CES14 2CES15 Average 2H11 2H12 2H13 2H14 2H15 Average 3C11 3C12 3C13 3C14 3C15 Average 3SGll 3SG12 3SG13 3SG14 3SG15 Average 4SGll 4SG12 4SG13 4SG14 4SG15 Average 5CS11 5CS12 5CS13 5CS14 5CS15 Average 7.69 15.60 7.70 7.70 15. 4 7.76 7.76 15.75 7.76 7.74 15.65 7.76 7.70 15.72 7.70 7.70 15.68 7.74 7.66 15.60 7.65 7.65 15.65 7.66 7.62 15.62 7.70 7.65 15.65 7.70 7.75 15.62 7.62 7.67 15.63 7.67 7.65 15.65 7.65 7.71 15.60 7.62 7.66 15.56 7.68 7.70 15.70 7.70 7.70 15.65 7.70 7.68 15.65 7.67 7.62 15.75 7.52 7.72 15.70 7.C0 7.70 15.72 7.65 7.70 15.78 7.70 7.70 15.78 7.60 7.69 15.74 7.63 7.65 15.60 7.60 7.72 15.62 7.63 7.68 15.70 7.67 7.60 15.62 7.62 7.70 15.65 7.60 7.67 15.64 7.62 7.70 15.75 7.70 7.72 15.75 7.75 7.72 15.71 7.71 7.72 15.75 7.76 7.75 15.65 7.75 7.72 15.74 7.73 7.65 15.60 7.65 7.75 15.61 7.65 7.66 15.62 7.70 7.75 15.62 7.71 7.68 15.62 7.65 7.70 15.61 7.67 119.96 120.43 122.22 121.13 121.04 120.69 119.50 119.72 119.02 119.72 121.06 119.80 119.72 120.28 119.96 120.89 120.51 120.27 120.02 121.20 121.04 121.51 121.51 121.06 119.34 120.59 120.58 118.71 120.51 119.95 121.28 121.59 121.28 121.67 122.06 121.58 119.34 120.98 119.65 121.06 119.96 120.20 923.69.535 6ES11 934.54.541 6ES12 948.27.549 6ES13 939.97.544 6ES14 932.01.539 6ES15 935.60.542 Average 914.18.529 7W11 917.07.531 7W12 916.48.530 7W13 921.86.534 7W14 922.43.534 7W15 918.40.532 Average 915.87.530 7BWll 916.53.530 7BW12 921.29.533 7BW13 930.85.539 7BW14 927.89.537 7BW15 922.49.534. Average 914.51.529 8CELll 912.15.533 8CEL12 925.99.536 8CEL13 935.60.541 8CEL14 923.45.534 8CEL15 924.14.535 Average 906.98.525 9Sll 920.07.532 9S12 924.82.535 9S13 904.56.524 9S14 915.84.530 9S15 914.45.529 Average 933.82.540 10CAll 942.32.545 10CA12 935.08.541 10CA13 942.92.546 lOCAl4 933.77.540 10CA15 937.58.543 Average 912.95.528 925.47.536 921.86.534 933.33.540 915.06.530 921.73.533 7.66 15.72 7.68 120.42 7.70 15.65 7.70 120.51 7.70 15.61 7.70 120.20 7.72 15.60 7.75 120.43 7.70 15.65 7.69 120.51 7.70 15.65 7.70 120.41 7.65 15.60 7.65 119.34 7.71 15.70 7.70 121.05 7.62 15.65 7.62 119.25 7.70 15.60 7.70 120.51 7.70 15.65 7.70 120.51 7.68 15.64 7.67 120.05 7.62 15.65 7.61 119.25 7.62 15.62 7.62 119.02 7.65 15.70 7.68 120.10 7.63 15.67 7.68 119.56 7.65 15.62 7.65 119.49 7.63 15.65 7.65 119.48 7.79 15.75 7.70 122.69 7.79 15.72 7.71 122.46 7.80 15.70 7.85 122.46 7.66 15.70 7.70 120.26 7.80 15.75 7.70 122.58 7.77 15.72 7.73 122.14 7.65 15.61 7.62 119.42 7.67 15.62 7.60 119.81 7.68 15.60 7.62 119.81 7.57 15.62 7.62 118.24 7.62 15.62 7.62 119.02 7.64 15.61 7.62 119.26 7.70 15.64 7.68 120.43 7.70 15.66 7.62 120.58 7.70 15.64 7.70 120.43 7.73 15.69 7.68 121.13 7.70 15.72 7.66 121.04 7.71 15.67 7.67 120.72 924.79.535 927.89.537 925.52.536 933.35.540 926.68.536 927.65.537 912.95.528 932.06.539 908.69.526 924.90.535 927.85.537 921.29.533 907,52.525 906.96.525 922.41.534 918.24.531 914.12.529 913.85.530 D 944.71.547 c 944.15.546 961.31.556 926.02.536 945.95.547 944.43.547 909.95.527 910.52.527 912.94.528 901.01.521 906.96.525 908.27.526 924.90.535 918.82.532 927.30.537 930.27.538 927.20.537 925.70.535

- 239 - TABLE 37. NET AREA AS PERCENTAGE OF GROSS AREA Net Gross Net Area Net Gross Net Area Area Area as % of Area Area as % of Specimen An A Gross Area Specimen An Ag Gross Area No. g No.2 in 2 in2 ( M)....... in 2 in....n, 1Cll 1C12 1C13 lC14 1C15 Average 2CES11 2CES12 2CES13 2CES14 2CES15 Average 2Hll 2H12 2H13 2H14 2H15 Average 3C11 3C12 3C13 3C14 3C15 Average 3SG1l 3SG12 3SG13 3SG14 3SG15 Average 4SG11 4SG12 4SG13 4SG14 4SG15 Average 5C01 5C12 5C13 5C14 5C15 Average 61.66 61.77 62.52 61.81 60.67 61.69 68.56 67.12 66.53 63.84 71.40 67.49 67.80 67.59 67.02 66.61 66.85 67.19 73-79 72.35 72.39 71.45 72.90 72.58 64.99 65.42 66.59 65.83 66.56 65.89 61.53 60.16 61.18 60.90 60.57 60.86 60.58 60.15 60.03 60.10 60.98 60.42 119.96 120.43 122.22 121.13 121.04 120.96 119.50 119.72 119.02 119.72 121.06 119.80 119.72 120.28 119.96 120.89 120.51 120.27 120.02 121.20 121.04 121.51 121.51 121. o6 119.34 120.59 120.58 118.71 120.51 119.95 121.28 121.59 121.28 121.67 122. o6 121.58 119.34 120.98 119.65 121.06 119.96 120.20 51 51 51 51 50 51 57 56 56 53 59 56 57 56 56 55 55 56 61 60 60 59 60 60 54 54 55 55 55 55 51 49 50 50 50 50 6ES11 6ES12 6ES13 6ES14 6ES15 Average 7W11 7W12 7W13 7W14 7W15 Average 7BWll 7BW12 7BW13 7BW14 7BW15 Average 8CEL11 8CEL12 8CEL13 8CEL14 8CEL15 Average 9Sll 9S12 9S13 9S14 9S15 Average 10CAll 1OCA12 10CA13 lOCA14 lOCA15 Average 58.20 59.20 59.99 59.64 60.86 59.54 62.10 62.02 62.43 61.54 61.02 61.82 61.70 62.23 61.34 61.42 60.96 61.53 67.76 67.94 67.05 62.45 62.61 65.56 66.42 68.65 67.75 67.91 65.79 67.32 65.19 65.94 64.99 65.63 66.28 65.60 120.42 120.51 120.20 120.43 120.51 120.41 119.34 121.05 119.25 120.12 120.51 120.05 119.25 119.02 120.10 119.56 119.49 119.48 122.69 122.46 122.46 120.26 122.85 122.14 119.42 119.81 119.81 118.24 119.02 119.26 120.43 120.58 120.43 121.13 121.04 120.72 48 49 50 50 51 50 52 51 52 51 51 51 52 52 51 51 51 51 55 55 55 52 51 54 56 57 57 57 55 56 54 55 54 54 55 54 51 50 50 50 51 50 - - --

APPENDIX III UNIT WEIGHT, ABSORPTION AND MOISTURE CONTENT A. Unit Weight. Average weight of the masonry units, as sampled, wet and dry conditions, are presented in Table 8 (Chapter V). Each value is the average of the results of 5 specimens tested for each product. Results of all specimens are given in Table 38 in grams (as tested) and in pounds. Unit weights are calculated from the weight of the specimens (Table 38) using their net volume (Table 35). Results are given in Table 39 from which the average results of Table 8 are taken. B. Absorption. Procedure of test for determining the absorption of the unit was discussed in Chapter V. Average results of the test are presented in Table 9. Each value in the table is the result of 5 specimens used for each product tested. Results of all the specimens are presented in Table 40 in two forms: (1) in pounds per cubic foot (pcf), and (2) in percentage of dry weight of the unit. The following equations were used: Let Ww = wet weight of the unit, grams (see Chapter V) Wd = dry weight of the unit, grams (see Chapter V) v = net volume of the unit, cc (see Chapter V) (W Wd.) 62.4 Absorption (pcf) (30) (W - Wd) 100 Absorption (per cent of = Ww - d(31) dry weight) Wd 240

TABLE 38. WEIGHTS OF MASONRY UNITS (SAMPLED, WET AND DRY CONDITIONS) Sample Sampled Weight Wet Weight Dry Weight Sample Sampled Weight Wet Weight Dry Weight No. gms lbs gnms lbs gmss IbQ No. gms lbs gms lbs gms lbs 1Cll 1C12 lC13 lC15 Average 2CES11 2CES12 2CES13 2CES14 2CES15 Average 2Hll 2H12 2H13 2H14 2H15 Average 3Cll 3C12 3C13 3C14 3C15 Average 3SGll 3SG12 3SG13 3SG14 3SG15 Average 4SGll 4SG12 4SG13 4SG14 4SG15 Average 5CS11 5CS12 5CS13 5CS14 5CS15 Average 13,320 12,204 12,280 12,170 11,970 12,179 13,865 13,320 13,560 12,330 14,380 13,491 12,530 12,820 12,540 12,540 12,680 12,622 12,793 12,880 12,508 12,130 12,697 12,602 17,785 17,560 18,625 17,920 18,200 18,018 16,570 16,208 17,230 16,840 15,930 16,556 12,610 12,580 12,280 12,240 12,280 12,398 27.1 13,365 29.5 26.9 13,325 29.4 27.0 13,435 29.6 26.8 13,335 29.4 26.3 12,990 28.6 26.8 29.3 30.5 14,910 32.9 29.3 14,320 31.6 29.8 14,655 32.3 27.1 13,580 29.9 31.6 15,270 33.7 29.7 32.1 27.6 13,325 29.4 28.2 13,445 29.6 27.6 13,240 29.2 27.6 13,330 29.4 27.9 13,450 29.6 27.8 29.4 28.1 14,770 32.6 28.3 14,895 32.8 27.5 14,450 31.9 26.7 14,250 31.4 27.9 14,470 31.9 27.7 32.1 39.1 18,370 40.5 38.6 18,230 40.2 41.0 19,130 42.2 39.4 18,670 41.2 40.0 18,745 41.3 39.6 41.1 36.5 17,395 38.3 35.7 17,150 37.8 37.9 17,775 39.2 37.1 17,565 38.7 35.1 16,880 37.2 36.4 38.3 27.7 13,465 29.7 27.7 13,395 29.5 27.0 13,190 29.1 26.9 13,370 29.5 27.0 13,000 28.7 27.3 29.3 11,450 11,455 11,435 11,430 11,200 11,394 12,945 12,555 12,850 11,800 13,085 12,647 11,270 11,420 11,340 11,400 11,500 11,386 12,250 12,600 11,960 11,760 12,060 12,126 17,275 17,050 18,090 17,450 17,710 17,515 16,100 15,745 16,735 16,405 15,460 16,089 12,100 12,030 11,700 11,695 11,680 11,841 25.2 25.3 25.2 25.2 24.7 25.1 28.5 27.7 28.3 26.0 28.8 27.9 24.8 25.2 25.0 25.1 25.4 25.1 27.0 27.8 26.4 26.0 26.6 26.7 38.1 37.6 39.9 38.5 39.0 38.2 35.5 34.7 36.9 36.2 34.0 35.5 26.7 26.5 25.8 25.8 25.7 26.1 6ES11 6ES12 6ES13 6ES14 6ES15 Average 7W11 7W12 7W13 7W14 7W15 Average 7BW11 7BW12 7BW13 7BW14 7BW15 Average 8CEL11 8CEL12 8CEL13 8CEL14 8CEL15 Average 9S11 9S12 9S13 9S14 9S15 Average 10CAll 10CA12 10CA13 10CA14 10CA15 Average 12,575 12,824 13,345 13,075 13,635 13,091 12,380 12,350 12,530 12,420 12,280 12,392 13,150 13,170 13,280 13,300 12,260 13,032 11,920 12,160 12,315 10,130 9,465 11,198 17,140 17,765 17,670 17,635 16,900 17,422 12,120 12,030 11,920 11,980 11,870 11,984 27.7 13,515 28.2 13,725 29.4 14,140 28.8 14,050 30.0 14,635 28.8 27.2 13,520 27.2 13,485 27.6 13,735 27.3 13,600 27.0 13,135 27.3 28.9 14,300 29.0 14,380 29.2 14,475 29.3 14,435 27.0 13,542 28.7 26.2 12,850 26.8 12,940 27.1 13,080 22.3 11,070 20.8 10,855 24.6 37.7 18,270 39.1 18,860 38.9 18,740 38.8 18,800 37.2 17,965 38.3 26.7 13,165 26.5 13,060 26.3 13,010 26.4 13,070 26.2 13,140 26.4 29.8 30.3 31.2 31.0 32.3 30.9 29.8 29.7 30.3 30.0 29.0 29.7 31.5 31.7 31.9 31.8 29.9 31.4 28.3 28.5 28.8 24.4 23.9 26.8 40.3 41.6 41.3 41.4 39.6 40.8 29.0 28.8 28.7 28.8 29.0 28.9 11,655 25.7 11,960 26.4 12,330 27.2 12,320 27.2 12,810 28.2 12,215 26.9 12,135 26.8 12,030 26.5 12,300 27.1 12,200 26.9 11,965 26.4 12,125 26.7 12,710 28.0 12,750 28.1 12,905 28.4 12,830 28.3 11,925 26.3 12,625 27.8 11,040 24.3 11,150 24.6 11,270 24.8 9,100 20.1 8,760 19.3 10,264 22.6 16,690 36.8 17,300 38.1 17,230 38.0 17,290 38.1 16,480 36.3 16,998 37.5 11,310 24.9 11,270 24.9 11,200 24.7 11,240 24.8 11,260 24.8 11,256 24.8 1)

TABLE 39. CALCULATION OF WEIGHTS OF MASONRY UNITS (SAMPLED, WET AND DRY CONDITIONS) IN POUNDS PER CUBIC FOOT (BASED ON NET VOLUME) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Net Net Weight of the Unit Volume Unit Weight Weight of the Unit Volume Unit Weight Sample (pound) of the pcf Sample (pound) of the pcf No. Sampled Wet Dry Unit Sampled Wet Dry No. Sampled Wet Dry Unit Sampled Wet Dry ft3 ft3 1C11 27.1 29.5 25.2.275 98.6 107.2 91.9 6ESll 27.7 29.8 25.7.258 1C12 26.9 29.4 25.3.277 96.8 105.9 91.0 6ES12 28.2 30.3 26.4.264 1C13 27.0 29.6 25.2.281 96.2 105.5 89.7 6ES13 29.4 31.2 27.2.267 1C14 26.8 29.4 25.2.278 96.4 105.9 90.7 6ES14 28.8 31.0 27.2.268 1C15 26.3 28.6 24.7.270 97.4 105.9 91.3 6ES15 30.0 32.3 28.2.271 Average 26.8 29.3 25.1.276 97.1 106.1 90.9 Average 28.8 30.9 26.9.266 2CES11 30.5 32.9 28.5.304 100.5 108.3 94.0 7W11 27.2 29.8 26.8.275 2CES12 29.3 31.6 27.7.298 98.5 106.1 93.0 7W12 27.2 29.7 26.5.276 2CES13 29.8 32.3 28.3.297 100.6 108.9 95.5 7W13 27.6 30.3 27.1.275 2CES14 27.1 29.9 26.0.285 95.4 105.2 91.3 7Wl4 27.3 30.0 26.9.274 2CES15 31.6 33.7 28.8.315 100.5 104.8 91.6 7W15 27.0 29.0 26.4.272 Average 29.7 32.1 27.9.299 99.1 106.7 93.1 Average 27.3 29.7 26.7.275 2H11 27.6 29.4 24.8.300 91.9 97.8 82.7 7BWll 28.9 31.5 28.0.272 2H12 28.2 29.6 25.2.298 94.6 99.4 84.4 7BW12 29.0 31.7 28..1.274 2H13 27.6 29.2 25.0.298 92.6 98.0 83.9 7BW13 29.2 31.9 28.4.273 214 27.6 29.4 25.1.297 93.0 99.0 84.7 7BW14 29.3 31.8 28.3.273 2H15 27.9 29.6 25.4.298 93.7 99.5 85.1 7BW15 27.0 29.9 26.3.270 Average 27.8 29.4 25.1.298 93.1 98.7 84.2 Average 28.7 31.4 27.8.272 3C11 28.1 32.6 27.0.325 86.5 100.0 83.0 8CEL11 26.2 28.3 24.3.302 3C12 28.3 32.8 27.8.318 98.1 103.2 87.3 8CEL12 26.8 28.5 24.6.318 3C13 27.5 31.9 26.4.321 85.9 99.4 82.3 8CEL13 27.1 28.8 24.8.305 3C14 26.7 31.4 26.0.318 84.8 98.7 81.4 8CELl4 22.3 24.4 20.1.278 3C15 27.9 31.9 26.6.321 87.1 99.4 82.9 8CEL15 20.8 23.9 19.3.279 Average 27.7 32.1 26.7.321 88.5 100.1 83.4 Average 24.6 26.8 22.6.296 3SG11 39.1 40.5 38.1.286 136.9 141.6 133.2 9S11 37.7 40.3 36.8.293 3SG12 38.6 40.2 37.6.289 133.7 139.1 130.1 9S12 39.1 41.6 38.1.302 3SG13 41.0 42.2 39.9.296 138.6 142.6 134.9 9S13 38.9 41.3 38.0.300 3SG14 39.4 41.2 38.5.290 135.8 141.8 132.5 9S14 38.8 41.4 38.1.300 3SG15 40.0 41.3 39.0.293 136.8 141.1 133.3 9S15 37.2 39.6 36.3.290 Average 39.6 41.1 38.2.291 136.3 141.2 132.8 Average 38.3 40.8 37.5.297 4SG11 36.5 38.3 35.5.274 132.9 139.8 129.4 10CAll 26.7 29.0 24.9.290 4SG12 35.7 37.8 34.7.270 132.2 140.1 128.6 10CA12 26.5 28.8 24.9.291 4SG13 37.9 39.2 36.9.273 138.9 143.5 135.1 10CA13 26.3 28.7 24.7.290 4SG14 37.1 38.7 36.2.273 135.6 141.7 132.4 10CA14 26.4 28.8 24.8.292 4SG15 35.1 37.2 34.0.268 131.7 138.7 127.0 10CA15 26.2 29.0 24.8.294 Average 36.4 38.3 35.5.272 134.1 140.8 129.2 Average 26.4 28.9 24.8.291 5CS11 27.7 29.7 26.7.268 103.4 110.7 99.4 5CS12 27.7 29.5 26.5.266 103.9 110.9 99.6 5CS13 27.0 29.1 25.8.268 101.0 108.7 96.4 5Cs14 26.9 29.5 25.8.272 99.0 108.3 94.7 5CS15 27.0 28.7 25.7.270 100.1 106.1 95.3 Average 27.3 29.3 26.1.269 101.5 108.9 97.1 107.0 115.2 99.3 106.9 114.7 99.9 109.8 116.6 101.7 107.6 115.8 101.5 110.7 119.1 104.2 108.4 116.3 101.3 99.1 108.4 97.3 98.3 107.6 95.5 100.2 110.0 98.5 99.6 109.3 98.1 99.4 106.5 97.0 99.3 108.3 97.3 106.5 116.0 103.1 105.6 115.5 102.4 107.2 117.0 104.3 107.2 116.5 103.6 99.9 105.6 97.3 105.3 114.1 102.2 86.9 93.8 80.6 84.3 89.8 77.4 88.9 94.7 81.6 80.1 87.7 72..1 74.6 85.7 69.2 83.0 90.3 76.2 128.8 137.5 125.6 129.5 137.7 126.3 130.1 138.2 127.1 129.6 138.4 127.3 128.2 136.5 125.2 129.2 137.7 126.3 92.1 100.1 86.0 91.2 98.9 85.4 90.6 98.9 85.1 90.5 98.7 84.9 89.0 98.5 84.4 90.7 99.0 85.2

- 243 - TABLE 40. ABSORPTION OF THE MASONRY UNITS 1 2 3 4 5 6 7 8 Absorbed Absorbed Water Water Sample Absorbed Absorbed per cent of Sample Absorbed Absorbed per cent of No. Water Water Dry Weight No. Water Water Dry Weight gns pcf % gms pcf % 1C11 1C12 lC13 lC14 lC15 Average 2CES11 2CES12 2CES13 2CES14 2CES15 Average 2Hll 2H12 2H13 2H14 2H15 Average 3Cll 3C12 3C13 3C14 3C15 Average 3SGll 3SG12 3SG13 3SG14 3SG15 Average 4SGll 4SG12 4SG13 4SG14 4SG15 Average 5CS11 5CS12 5CS13 5cs14 5CS15 Average 1,915 1,870 2,000 1,905 1,790 1,896 1,965 1,765 1,805 1,780 2,185 1,900 2,055 2,025 1,900 1,930 1,950 1,972 2,520 2,295 2,490 2,490 2,410 2,441 1,095 1,180 1,040 1,220 1,035 1,110 1,295 1,405 1,040 1,160 1,420 1,264 1,365 1,365 1,490 1,675 1,620 1,503 15.4 14.9 15.7 15.1 14.6 15.1 14.3 13.1 13.4 13.8 15.3 14.0 15.1 15.0 14.1 12.0 14.4 14.1 17.1 15.9 17.1 17.2 16.6 16.8 8.4 9.0 7.8 9.3 7.8 8.5 10.4 11.5 8.4 9.4 11.6 10.3 11.2 11.3 12.3 13.6 13.2 12.3 16.7 16.3 17.5 16.7 16.0 16.6 15.2 14.1 14.1 15.1 16.7 15.0 18.2 17.7 16.8 16.9 17.0 17.3 20.6 18.2 20.8 21.2 20.0 20.2 6.3 7.0 5.8 7.0 5.8 6.4 8.0 8.9 6.2 7.1 9.2 7.9 6ES11 6ES12 6ES13 6ES14 6ES15 Average 7W11 7W12 7W13 7W14 7W15 Average 7BWll 7BW12 7BW13 7BW14 7BW15 Average 8CEL11 8CEL12 8CEL13 8CEL14 8CEL15 Average 9Sll 9S12 9S13 9S14 9S15 Average 10CAll 10CA12 10CA13 10CA14 lOCAlS 10CA15 Average 1,860 1,765 1,810 1,727 1,825 1,797 1,385 1,455 1,435 1,400 1,170 1,369 1,590 1,630 1,570 1,605 1,617 1,602 1,810 1,790 1,810 1,970 2,095 1,895 1,580 1,560 1,510 1,510 1,485 1,529 1,855 1,790 1,810 1,830 1,880 1,833 15.8 14.7 14.9 14.2 14.8 14.9 11.1 11.6 11.5 11.3 9.5 11.0 12.9 13.1 12.7 13.0 13.2 13.0 13.2 13.0 13.1 14.3 16.6 14.0 11.9 11.2 11.2 17.8 11.3 12.7 14.1 13.6 13.8 13.8 14.1 13.9 16.0 14.8 14.7 14.0 14.3 14.7 11.4 12.1 11.7 11.5 9.8 11.3 12.5 12.8 12.2 12.5 13.6 12.7 16.4 16.1 16.1 21.7 23.9 18.8 9.5 9.0 8.8 8.7 9.0 9.0 16.4 15.9 16.2 16.3 16.7 16.3 11.3 11.4 12.7 14.3 13.9 12.3

C. Moisture Content. Procedure for determining the moisture content of the units, as sampled, was discussed in Chapter V. Average results of the test are presented in Table 9. These averages are taken from Table 41 in which the results of tests for all specimens are calculated. Moisture contents are presented as percentages of the absorption and were calculated by using the following equation: Let Ws = weight of the specimen as sampled Ww and Wd as before Thus, (W - Wd) 100 Moisture Content (per cent of = - (32) absorption) Ww Wd

- 245 - TABLE 41. CALCULATION OF MOISTURE CONTENT OF THE MASONRY UNITS AS SAMPLED 1 2 3 4 5 6 7 7.. Moisture Moisture Content Content Sampled as Sampled as Sample Moisture Absorption per cent Sample Moisture Absorption per cent No. Content of No. Content of gms gms Absorption gas gms Absorption.......~~~~~~~~~~~~~m 1Cll 1C12 1C13 ic14 1C15 Average 2CES11 2CES12 2CES13 2CES14 2CES15 Average 2Hll 2H12 2H13 2H14 2H15 Average 3Cll 3C12 3C13 3C14 3C15 Average 3SGll 3SG12 3SG13 3SG14 3SG15 Average 4SGll 4SG12 4SG13 4SG14 4SG15 Average 5CS11 5CS12 5CS13 5CS14 5CS15 Average 870 749 845 740 770 795 920 765 710 530 1,295 844 1,260 1,400 1,200 1,140 1,180 1,236 543 280 548 370 637 476 510 510 535 470 490 503 470 463 495 435 470 467 510 550 580 545 600 557 1,915 1,870 2,000 1,905 1,790 1,896 1,965 1,765 1,805 1,780 2,185 1,900 2,055 2,025 1,900 1,930 1,950 1,972 2,520 2,295 2,490 2,490 2,410 2,441 1,095 1,180 1,040 1,220 1,035 1,110 1,295 1,405 1,040 1,160 1,420 1,264 1,365 1,365 1,490 1,675 1,620 1,503 45.4 4o0.0 42.3 38.8 43.0 41.9 46.8 43.3 39.3 29.8 59.3 43.7 61.3 69.1 63.2 59.1 60.5 62.6 21.5 12.2 22.0 14.9 26.4 19.4 46.6 43.2 51.4 38.5 47.3 45.4 36.3 33.0 47.6 37.5 33.1 37.5 37.4 40.3 38.9 32.5 37.0 37.2 6ESll 6ES12 6ES13 6ES14 6ES15 Average 7W11 7W12 7W13 7W14 7W15 Average 7BWll 7BW12 7BW13 7BW14 7BW15 Average 8CEL11 8CEL12 8CEL13 8CEL14 8CEL15 Average 9Sll 9S12 9S13 9S14 9s15 Average lOCAll 10CA12 10CA13 lOCA14 lOCA15 Average 920 854 1,015 755 825 874 245 320 230 220 315 266 440 420 375 470 335 408 880 1,010 1,045 1,030 705 934 450 465 440 345 420 424 810 760 720 736 610 727 1,860 1,765 1,810 1,727 1,825 1,797 1,385 1,455 1,435 1,400 1,170 1,369 1,590 1,630 1,570 1,605 1,617 1,602 1,810 1,790 1,810 1,970 2,095 1,895 1,580 1,560 1,510 1,510 1,485 1,529 1,855 1,790 1,810 1,830 1,880 1,833 49.5 48.4 56.1 43.7 45.2 48.6 17.7 22.0 16.0 15.7 26.9 19.7 27.7 25.8 23.9 29.3 20.7 25.5 48.6 56.4 57.7 52.3 33.7 49.7 28.5 29.8 29.1 22.8 28.3 27.7 43.7 42.5 39.8 40.2 32.5 39.7

APPENDIX IV COMPRESSIVE STRENGTH Compressive strength of concrete masonry units is discussed in Chapter VI. Results are given in Table 10; each is the average of the results of test on 5 units. Detailed data is given in Table 43. Results are based on the gross cross-sectional area of the units which are determined by actual measurement of the overall dimensions of the unit. Results were also presented on the basis of the net cross-sectional areas of the units which are the average net areas of a similar specimen (five per product), as reported in Table 7 (Chapter IV). Therefore. Compressive strength (based on gross area) psi = compressive force (pound) (33) gross area (in2) Compressive strength (based on net area) psi = compressive force (pound) net area (in2)- (3 Detailed data of the gross cross-sectional area is given in Table 42; average results are also given in Table 10.

- 247 - TABLE 42. CALCULATION OF GROSS CROSS-SECTIONAL AREAS OF MASONRY UNITS, USED FOR COMPRESSIVE STRENGTH TEST 1 2 3 4 5 6 7 8 Gross Cross- Gross CrossLength Thickness Sectional Length Thickness Sectional Specimen L t Area Specimen L t Area No. t x L No. t x L in. in. in2 in. in. in 1C21 1C22 1C23 1C24 1C25 Average 2CES21 2CES22 2CES23 2CES24 2CES25 Average 2H21 2H22 2H23 2H24 2H25 Average 3C21 3C22 3C23 3C24 3C25 Average 3SG21 3SG22 3SG23 3SG24 3SG25 Average 4SG21 4SG22 4SG23 4SG24 4SG25 Average 5CS21 5CS22 5CS23 5CS24 5CS25 Average 15.71 15.63 15.60 15.63 15.69 15.65 15.62 15.60 15.62 15.65 15.62 15.62 15.62 15.63 15.65 15.6o 15.62 15.62 15.70 15.70 15.66 15.67 15.65 15.68 15.70 15.70 15.70 15.70 15.70 15.70 15.65 15.78 15.75 15.72 15.74 15.73 15.55 15.63 15.51 15.61 15.62 15.58 7.70 7.69 7.62 7.68 7.65 7.67 7.68 7.60 7.68 7.62 7.66 7.65 7.63 7.63 7.62 7.62 7.60 7.62 7.72 7.68 7.70 7.68 7.67 7.69 7.75 7.68 7.70 7.68 7.65 7.69 7.71 7.72 7.71 7.70 7.75 7.72 7.65 7.63 7.65 7.63 7.62 7.64 121.0 120.2 118.9 120.0 120.0 120.0 120.0 118.6 120.0 119.3 119.7 119.6 119.2 119.3 119.3 118.9 118.7 119.1 121.2 120.6 120.6 120.4 120.0 120.5 121.7 120.6 120.9 120.6 120.1 120.8 120.7 121.8 121.4 121.0 122.1 121.4 119.0 119.3 118.7 119.1 119.0 119.0 6ES21 6ES22 6ES23 6ES24 6ES25 Average 7W21 7W22 7W23 7W24 7W25 Average 7BW21 7BW22 7BW23 7BW24 7BW25 Average 8CEL21 8CEL22 8CEL23 8CEL24 8CEL25 Average 9S21 9S22 9S23 9S24 9S25 Average 10CA21 10CA22 10CA23 10CA24 10CA25 Average 15.65 15.55 15.58 15.58 15.63 15.60 7.68 7.70 7.71 7.72 7.61 7.68 15.55 15.62 15.62 15.66 15.64 15.62 15.59 15.58 15.60 15.62 15.62 15.60 15.71 15.70 15.75 15.73 15.75 15.73 15.59 15.60 15.62 15.56 15.60 15.59 15.70 15.66 15.65 15.63 15.70 15.67 7.65 7.62 7.66 7.66 7.65 7.65 7.72 7.65 7.65 7.65 7.66 7.66 7.75 7.76 7.72 7.76 7.69 7.74 7.65 9.62 7.63 7.62 7.62 7.63 7.65 7.72 7.70 7.70 7.68 7.69 120.2 119.7 120.1 120.3 118.9 119.9 119.0 119.0 119.6 119.6 119.7 119.4 120.4 119.2 119.3 119.5 119.7 119.6 121.8 121.8 121.6 122.1 121.1 121.7 119.3 118.9 119.2 118.,6 118.9 119.0 120.1 120.9 120.5 120.4 120.6 120.5 -- - -

- 248 - TABLE 43. CALCULATION OF COMPRESSIVE STRENGTH OF CONCRETE MASONRY UNITS (BASED ON GROSS CROSS-SECTIONAL AREA) 1 2 3 4 5 6 7 8 Gross Cross- Gross CrossCompressive Sectional Compressive Compressive Sectional Compressive Specimen Force Area Strength Specimen Force Area Strength No. No. lbs in2 psi lbs in psi 1C21 1C22 1C23 1C24 1C25 Average 2CES21 2CES22 2CES23 2CES24 2CES25 Average 2H21 2H22 2H23 2H24 2H25 Average 3C21 3C22 3C23 3C24 3C25 Average 3SG21 3SG22 3SG23 3SG24 3SG25 Average 4SG21 4SG22 4SG23 4SG24 4SG25 Average 5CS21 5CS22 5CS23 5CS24 5CS25 Average 83,000 89,ooo 85,500 73,000 89,000 83,900 145,000 107,500 127,500 118,000 102,000 120,000 162,100 172,700 157,600 152,600 148,100 158,620 92,500 95,500 79,000 102,500 85,000 90,900 125,000 177,000 95,000 171,500 160,350 145,770 124,000 113,500 140,000 135,000 91,000 120,700 113,000 113,000 89,000 102,000 105,000 104,400 121.0 120.2 118.9 120.0 120.0 120.0 120.0 118.6 120.0 119.3 119.7 119.6 119.2 119.3 119.3 118.9 118.7 119.1 121.2 120.6 120.6 120.4 120.0 120.5 121.7 120.6 120.9 120.6 120.1 120.8 120.7 121.8 121.4 121.0 122.1 121.4 119.0 119.3 118.7 119.1 119.0 119.0 686 740 719 606 740 698 1,209 907 1,063 990 852 1,004 1,360 1,448 1,322 1,284 1,248 1,332 6ES21 6ES22 6ES23 6ES24 6ES25 Average 7W21 7W22 7W23 7W24 7W25 Average 7BW21 7BW22 7BW23 7BW24 7BW25 Average 78,000 110,000 80,000 65,000 82,000 83,000 77,500 81,500 103,500 111,500 90,500 92,900 126,000 110,500 131,000 112,500 111,500 118,300 66,000 118,000 65,000 85,000 59,500 78,600 145,000 162,100 142,600 130,100 159,100 147,800 115,600 116,000 96,000 95,000 123,000 109,120 120.2 119.7 120.1 120.3 118.9 119.9 119.0 119.0 119.6 119.6 119.7 119.4 120.4 119.2 119.3 119.5 119.7 119.6 121.8 121.8 121.6 122.1 121.1 121.7 119.3 118.9 119.7 118.6 118.9 119.1 120.1 120.9 120.5 120.4 120.6 120.5 649 927 666 540 694 695 651 681 865 933 756 777 1,047 927 1,098 942 932 989 542 969 535 696 491 647 1,217 1,364 1,197 1,097 1,338 1,243 963 960 797 789 1,020 906 763 8CEL21 792 8CEL22 655 8CEL23 852 8CEL24 708 8CEL25 754 Average 1,027 1,468 786 '1,422 1,335 1,208 1,028 932 1,153 1,115 746 995 950 948 750 856 882 877 9S21 9S22 9S23 9S24 9S25 Average 10CA21 10CA22 10CA23 10CA24 10CA25 Average

APPENDIX V FLEXURAL STRENGTH In calculating the modulus of rupture (R), the assumptions made wereo (1) concrete is homogeneous elastic material and remains so near maximum stress, (2) the, transverse section of the beam plane before bending remains so after bending, and (3) that the moduli of elasticities are equal in tension and compression, and the section stress diagram is triangular. Assuming also that the breaking section is in mid-span, the following formula can be concluded for the stresses at this section of a simply supported beam.3 Let S = stresses at the extreme fiber of the beam (psi) P = transverse concentrated load at mid-span (pounds) M = bending moment caused by P, at mid-span (pounds) L = span length (inch) d = depth of the cross section (inch) b = width of the cross section (inch) c = d = distance of the extreme fibers from the neu2 tral axis (inch) I = d = moment of inertia of the section along the 12 neutral axis (in ) Then, = c (:1 ) t M= PL d b d3 but M= -, c = and I = 12 2' ~~12 249

-250 -Substitute in equation (1 ) and reduce to equation (.2 ) below: S 3PL ( 2 bd / When P is equal to the breaking force then, by definition, S is equal to the modulus of rupture (R). It is realized that the assumptions made are not all applicable to the breaking load; however, the approximate value obtained from the formula is useful for practical applications.

APPENDIX VI ELASTIC PROPERTIES In Chapter VIII the elastic constants E, G, and i were studied as determined by two methods of tests; namely, the dynamic (sonic) and the static. Detailed data and calculations will be presented here along with discussion of certain points in the tests. A. Dynamic Method. 1. Identification of the Mode of Vibration. Procedure of test for determining the resonant frequencies of the fundamental transverse and torsional vibrations was discussed in Chapter VIII. It was stated that each vibration is identified by the determination of its nodal points and phase difference between the driver and the pickup vibrations. This was accomplished by bringing the pickup needle in contact with the key point along the length of the specimen while observing the Lissajou's figure on the screen of the oscillograph. The identities of the two vibrations (transverse and torsional) are shown in Figure 47. Phase 1350 Angle= \ rd' - Trans verse qq0 - __ 135o 1\5 0 Vibration ofr^ ~ _ ^rat ion 131~25135\ \ 35~ 3 5~TorCd^ z \ ^ ^^ sional rd r-4 0 0 Phse ion Figure 47. Identification of Flexural and Torsional Modes of Vibration 251

-252 - 2. Calculations of the Elastic Constantso Calculation of the elastic constants E, G, and L have been discussed in Chapter VIII and summarized in Tables 14 and 15. Detailed data of the calculation are presented here in Tables 44 to 49 which were discussed in the previous presentation in Chapter VIII. B. Static Method. 1. Preliminary Investigationso Three methods of preparing the surfaces of the specimen before cementing SR-4 gages were discussed in Chapter VIII. It was stated that the second method (surface cavities filled with plaster of Paris) was adopted after a comparative study was made of the performances of gages cemented to faces of the specimen that have been prepared by this method and the first method (surface cavities filled with Duco cement). It was also stated that this study was performed on specimenstaken from the products 2CES5, 3SG5, and 9S5. The procedure of this test was described in Chapter VIII while the results are presented in the following (Tables 50, 51, and 52, and Figures 48, 49, and 50) which indicate that gages cemented to surfaces that had been prepared by the first and second method gave almost identical results. 2. Calculations. Calculations of the elastic constants were discussed in Chapter VIII and summarized in Table 17. Detailed stress-strain data for the vertical and horizontal (if present) gages of each specimen from which these results were obtained are presented here in Tables 53 to 73. These data were then plotted in Figures 51 to 71. Young's modulus of elasticity was then calculated for each

-253 -specimen from the slope of the straight portion of the stress-strain curve of the vertical gage; Poisson's ratio was calculated as the average ratio of the horizontal and vertical unit strain, and the modulus of elasticity in shear G was then calculated from E and by previously given (Chapter VIII) equation (15)o Values near zero stress are considered unreliable and were excluded from the calculations.

TABLE 44. DIMENSIONAL ANALYSIS OF SPECIMENS FOR THE SONIC TEST 1 2 3 4 5 6 7 8 9 Dimensions Area of of the Length Section Specimen Section L A=a x b a b L L No. a ab a a b in in in in 1C51 1C52 2CES51 2CES52 2.56 2.52 2.55 2.37 3.63 3.63 3.64 3.64 15.66 15.66 15.63 15.63 9.29.71 1.42 6.12 4.31 9.15.69 1.44 6.21 4.31 9.28 8.63.70 1.43.65 1.54 6.12 6.59 4.29 4.29 3C51 3C52 3SG51 3SG52 4SG51 4SG52 5CS51 5CS52 2.39 3.63 2.44 3.63 15.76 15.76 8.68.66 1.52 6.59 4.34 8.86.67 149 6.46 4.34 9.26.70 1.42 6.16 4.33 8.70.66 1.51 6.54 4.33 2.55 3.63 15.70 2.40 3.63 15.70 2.47 2.52 3.75 15.76 3.75 15.76 9.26 9.45.66 1.52 6.38 667 1.49 6.25 4.20 4.20 4.29 4.29 2049 3.64 15.63 2.55 3.64 15.63 9.06.68 1.46 6.28 9.28.70 1043 6.13 6ES51 6ES52 7W51 7W52 2.49 3.64 2.48 3.64 15.63 15.63 15.63 15.63 9.06.68 1.46 9.03.68 1.47 9.37.71 1.41 8.49.65 1.55 6.28 4.29 6.30 4.29 6.05 4.31 6.68 4.31 2.58 2.34 3.63 3.63 8CEL51 8CEL52 9S51 9S52 10CA51 10CA52 2.49 2.35 2.48 2.45 2.41 2.37 3.75 15.75 3.75 15.75 9.34.66 1.51 6.33 4.20 8.81.63 1.60 6.70 4.20 3.63 3.63 3.63 3.63 15.63 15.63 15.63 15.63 9.00.68 1.46 6.30 8.89.68 1.48 6.37 4.30 4.30 8.75.66 1.51 6.48 4.31 8.60.65 1.53 6.59 4.31 - -- -- --

-255 - TABLE 45. WEIGHTS AND RADIUS OF GYRATION - LENGTH OF SPECIMENS FOR THE SONIC TEST 1 2 3 4 5 6 Radius of Specimen Gyration Length r Weight Weight No. r L L W W in, in. gns pounds 1C51 1C52 1.049.1.049 2CES51 2CES52 1.052 1.052 3C51 3C52 1.049 1.049 3SG51 3SG52 4SG51 4SG52 5CS51 5CS52 1o049 1.049 1.0084 1.084 1.052 1.052 15.66 15.66 15.63 15.63 15.76 15.76 15.70 15.70 15.76 15.76 15.63 15.63 15.63 15.63 15.63 15.63 15.75 15.75 15.63 15.63 15.63 15.63 O0670.0670.0673. 673.0667.0667.0668.0668.0688.0688.0673.0673.0673.0673.0671.0671.0688.0688.0671.0671.0671.0671 3100 3040 3350 3140 2830 2860 4820 4425 5000 4920 3550 3620 3800 3725 3920 3510 3150 2890 4320 4200 3140 3000 6.84 6.70 7.39 6.92 6.24 6.31 10.63 9.76 11.03 1085 7.83 7.98 8.38 8.21 8.64 7.74 6.95 6037 9.53 9.26 6.92 6.62 6ES51 6ES52 7W51 7W52 8CEL51 8CEL52 9s51 9S52 lOCA51 10CA52 1.052 1.052 1o049 1.049 1.084 1.o84 1.049 1.049 1 049 1.049 - -- - -

-256 - TABLE 46. CALCULATION OF THE COEFFICIENT (C) FOR THE FLEXURAL VIBRATION 1 2 3 4 5 6 7 Driving in b direction Driving in a direction Specimen C';x'a*" C C x b* C No 2 2 L L.- b sec/in se c2/in 2 se/inc2/in sec2/in2 1C51 1C52 2CES51 2CES52 3C51 3C52 3SG51 3SG52 4SG51 4SG52 5CS51 5CS52 6ES51 6ES52 7W51 7W52 8CEL51 8CEL52 9S51 9S52 10CA51 10CA52 4.31 4.31 4.29 4.29 4.34 4.34 4.33 4.33 4.20 4.20 4.29 4.29 4.29 4.29 4.31 4.31 4.20 4.20 4.30 4.30 4.31 4.31.270.270.268.268.275.275.274..274 ~253.253.266.266.266.266 270.270.254.254.270.270.270.270.lo6.107 o105.113.115.113.108.114.102.100.107.104.107.107.105.115.102.108.109.110.112.114 6.12 6.21 6.12 6.59 6.59 6.46 6.16 6054 6.38 6.25 6.28 6.13 6.28 6.30 6.05 6.68 6.33 6.70 6.30 6.37 6.48 6.59.678.700.678.818.820.780.690.810.760.253.725.680.720.730.660.855.740.860.730.750.790.825.187.193.186 o225.226.215.190.223.203.192.199.187.198.201.182.236.197.229.201.207.218.227 * Values in columns (Figure 21). 3 and 6 are selected from Pickett's curve

-257 - TABLE 47. RESONANT FREQUENCIES FOR FLEXURAL AND TORSIONAL VIBRATION 1 2 3 4 5 6 7 8 Flexural Vibration Torsional Vibration Drive in Drive in Drive in b direction Drive in a direction b a Specimen n direction direction (n )2 No. n n2 n' n' cycle/sec (cy/sec)2 cycle/sec (cy/sec)2 cycle/sec cycle/sec (cy/sec)2 1C51 1C52 2CES51 2CES52 3C51 3C52 3SG51 3SG52 4SG51 4SG52 5CS51 5CS52 6ES51 6ES52 7W51 7W52 8CEL51 8CEL52 9S51 9S52 10CA51 10CA52 1,185 1,200 1,275 1,300 1,160 1,195 1,610 1,650 1,900 1,810 1,330 1,440 1,430 1,400 1,620 1,610 1,400 1,350 1,570 1,630 1,230 1,160 1,404,000 1,440,000 1,626,000 1,690,000 1,346,000 1,428,000 2,592,100 2,722,500 3,610,000 3,276,000 1,769,000 2,074,000 2,045,000 1,960,000 2,624,000 2,592,000 1,960,000 1,823,000 2,465,000 2,657,000 1,513,000 1,346,000 845 900 930 910 840 840 1,200 1,130 1,320 1,230 980 1,090 1,050 1,040 1,230 1,190 1,000 920 1,120 1,160 874 850 714,000 810,000 865, 000 828,000 706,000 706,000 1,440,000 1,276,000 1,742,000 1,513,000 960,000 1,188,000 1,103,000 1,082,000 1,513,000 1,416,000 1,000,000 846,ooo 1,254,000 1,346,000 764,000 723,000 1,600 1,660 1,700 1,700 1,540 1,580 2,220 2,200 2,410 2,300 1,820 1,990 1,960 1,930 2,230 2,145 1,880 1,720 2,120 2,175 1,670 1,600 1,600 1,660 1,700 1,700 1,540 1,580 2,220 2,200 2,410 2,300 1,820 1,990 1,960 1,930 2,230 2,145 1,880 1,720 2,120 2,175 1,670 1,600 2,560,000 2,755,600 2,890,000 2,890,000 2,372,000 2,496,000 4,928,000 4,840,000 5,808,000 5,290,000 3,312,000 3,960,000 3,842,000 3,725,000 4,973,000 4,601,000 3,534,000 2,958,000 4,494,000 4,731,000 2,789,000 2,560,000

-258 - TABLE 48. CALCULATION OF THE SHAPE FACTOR (R)* FOR TORSIONAL VIBRATION 1 2 3 4 5 6 7 8 9 Specimen 4a (a a) 2 b a a 2.52() R Specimeno. 4 b 4 - 2 E b No~~~.52(a) a 5 (2.52(2) No. 1C51 1C52 2CES51 2CES52 3C51 3C52 3SG51 3SG52 4SG51 4SG52 5CS51 5CS52 6ES51 6ES52 7W51 7W52 8CEL51 8CEL52 9s51 9s52 10CA51 10CA52.705.694.700.651.658.672.702.661.659.672.684.700.684.681.710.645.664.626.683.675.664.652 2.820 2.776 2.800 2.604 2.632 2.688 2.808 2.644 2.636 2.688 2.736 2.800 2.736 2.724 2.840 2.580 2.656 2.504 2.732 2.700 2.656 2.608.497.482.490.424.433.452.493.437.434.452.468.490.468.464.504.416.441.392.466.456.441.425 1.252 1.215 1.235 1.068 1.091 1.139 1.242 1.101 1.094 1.139 1.179 1.235 1.179 1.169 1.270 1.048 1.111.988 1.174 1.149 1.111 1.071 1.42 1.44 1.43 1.54 1.52 1.49 1.42 1.51 1.52 1.49 1.46 1.43 1.46 1.47 1.41 1.55 1.51 1.60 1.46 1.48 1.51 1.53 2.125 2.134 2.130 2.191 2.178 2.162 2.122 2.171 2.179 2.162 2.144 2.130 2.144 2.151 2.120 2.195 2.174 2.226 2.143 2.155 2.174 2.182 1.568 1.561 1.565 1.536 1.541 1.549 1. 566 1.543 1.542 1.549 1.557 1.565 1.557 1.555 1.570 1.532 1.545 1.516 1.558 1.551 1.545 1.537 1.355 1.367 1.361 1.426 1.413 1.396 1.355 1.407 1.413 1.396 1.377 1.361 1.377 1.383 1.350 1.433 1.407 1.468 1.375 1.389 1.407 1.420 * (R) is calculated by equation (14) following: (See Chapter VIII) a b a - b + - a 2.52 (a)2 b

-259 - TABLE 49. CALCULATION OF THE COEFFICIENT (B) FOR TORSIONAL VIBRATION* 1 2 3 4 5 6 7 g xA B Specimen R L 4 L R A = a x b 3 in3 sec2 No. -l sec in. in. in sec2 in.-~....,,',................... 1C51 1C52 1o355 15.66 1.367 15.66 1.361 15.63 1.426 15.63 2CES51 2CES52 84.88 85.63 85.09 89.15 89.08 88,00 9.29 9.15 9.28 8.63 8.68 8.86 3586 3335.0237.0267 3590.0236 3536.0242 3C51 3C52 1.413 1.396 15.76 15.76 3354.0266 3424.0257 3SG51 3SG52 4SG51 4SG52 5CS51 5CS52 1.355 15.70 85.09 1.407 15.70 88.37 1.413 1.396 15o76 15.76 89.08 88.00 9.26 8.71 9.26 9.45 9.06 9.28 3578 3366.0238.0263 3578.0249 3651.0241 3501.0246 3586.0237 1.377 15.63 86.09 10361 15.63 85.09 6ES51 6ES52 1.377 15.63 1.383 15.63 86.09 86.47 9.o6 9.03 3501 3489.0246.0248 7W51 7W52 1.350 15.63 84.42 1.433 15.63 89.59 1.407 15 75 88.64 1.468 15.75 92.48 8CEL51 8CEL52 9S51 9S52 10CA51 10CA52 9.37 8.49 9.34 8.81 9.00 8.89 8.75 8.60 1.375 1.389 1.407 1.420 15.63 15.63 15.63 15.63 3624.0233 3281.0273 85.97 86.84 88.64 88.78 3609 3404 3478 3435 3381 3323.0246.0272.0247.0253.0262.0267 * Calculated equation (13) follows: B = 4gA A where g = 386.4 in/sec2 gA g =

-260 - TABLE 50. PRELIMINARY INVESTIGATION - STRESS-STRAIN DATA (STATIC) FOR 2CES5 (CINDERS AND EXPANDED SLAG) Load pounds Stress psi 0 1000 2000 3000 4oo000 5000 6000 7000 8000 9000 0 log 109 217 326 435 543 652 761 870 978 Cavities Filled with Duco Cement Gage Strain Reading mic-in mic-in 6 1470 0 1360 110 1255 215 1145 325 1030 440 920 550 790 680 655 815 520 950 5 1350 1120 Cavities Filled with Plaster of Paris Gage Strain Reading mic-in mic-in 4 1850 0 1725 125 1600 250 1475 375 1342 508 1220 630 1080 770 935 915 790 1060 3 1600 1250

006 SN0lLV911S3ANI A8VNIWIN138d S3O aO 8 S3A8nO NIV.tlS-SS381S -8 ' HONI - OUOII NIV.LS 0gL 009 09Gt 0o0 O91 0 ---- ---------- ---- -~00 ----00 -- --— ^^.0-0 --- PA002 1 0 S00 C l l ll l ll l ll l ll l l-I - 009 u 0001 I

-262 - TABLE 51. PRELIMINARY INVESTIGATION - STRESS-STRAIN DATA (STATIC) FOR 3SG5 (SAND AND GRAVEL) Load pounds Stress psi 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000 14000 0 117 233 350 466 583 699 816 932 1049 1165 1282 1399 1516 1632 Cavities Filled with Duco Cement Gage Strain Reading mic-in mic-in 1630 0 1583 47 1535 95 1490 140 1440 190 1395 235 1350 280 1300 330 1250 380 1200 430 1150 480 1098 533 1045 585 990 640 935 695 Cavities Filled with Plaster of Paris Gage Strain Reading mic-in mic-in 4 1730 0 1680 50 1630 100 1581 149 1529 201 1576 248 1435 295 1387 343 1340 390 1292 438 1243 487 1189 540 1135 595 1078 653 1020 710 i _

-263 - 1600 1400 1000 (n 800 0. (0 (0 Iw) 600 w UJt~ -- - ----- - ------- n 400 ---- 0 0 50 300 40 600 750 900 STRAIN MICRO-INCH FIG. 49 - STRESS-STRAIN CURVES FOR 3SG5 PRELIMINARY INVESTIGATIONS

-264 -TABLE 52. PRELIMINARY INVESTIGATION - STRESS-STRAIN DATA (STATIC) FOR 9S5 (AIR-COOLED SLAG) Load pounds Stress psi 0 1000 2000 3000 4ooo 4000 5000 6ooo 7000 9000 10000 11000 12000 13000 14000 0 118 236 354 472 590 708 826 944 1062 1180 1298 1416 1534' 1652 Cavities Filled with Duco Cement Gage Strain Reading mic-in mic-in 6 1040 0 990 50 945 95 900 140 850 190 805 235 755 285 705 335 652 388 600 440 550 490 486 554 432 608 362 678 300 740 Cavities Filled with Plaster of Paris Gage Strain Reading mic-in mic-in 6 1230 0 1180 50 1135 95 1090 140 o040 190 993 237 950 280 898 332 842 388 792 438 740 490 685 545 629 601 560 670 510 72

-265 - 800 -., iS ~ _z_/_ n 600' w / a 400.. 200 0 50 300 450 600 750 900 STRAIN MICRO-INCH FIG. 50 - STRESS-STRAIN CURVES FOR 9S5 PRELIMINARY INVESTIGATIONS

-266 - TABLE 53. STRESS-STRAIN DATA (STATIC) FOR 1C51 (CINDERS) Load pounds Stress psi 0 1000 2000 3000 4ooo 4000 5000 6ooo 7000 8000 0 118 236 355 473 591 709 827 945 Vertical Reading mic-in 5 1760 1650 1530 1410 1290 1165 1030 895 760 Gage Strain mic -in 0 110 230 350 470 595 730 865 1000 Horizontal Reading mic-in 7 1280 1285 1295 1305 1320 1340 1362 1382 1403 Gage Strain mic-in 0 -5 -15 -25 -40 -60 -82 -102 -123

IS01 80J S3Adrno NItViS-SS381S - I '91 39V9 IVlNOZI8OH SOJ 006 OSL HONI-OHIOIYI NIVIS o00oz 001 o0 0 009 Ot' 002 0OS 0 091 0 0 Z 00~ 'o 'r U) U) m OS9- C 0) -0 009 (U 0gL 006 0901 0031

-268 - TABLE 54. STRESS-STRAIN DATA (STATIC) (CINDERS AND EXPANDED SLAG) FOR 2CES51 Load Stress Vertical Gage Horizontal Gage Reading Strain Reading Strain pounds psi mic - in mi c -in mic -in mic -in 5 5 0 0 -1730 0 1040 0 1000 117 1620 110 1050 -10 2000 233 1510 220 1060 -20 3000 350 1400 330 1080 -40 4000 466 1280 450 1100 -60 5000 583 1170 560 1120 -80 6000 699 1050 680 1150 -110 7000 816 920 810 1170 -130 8000 933 790 940 1240 -160 TABLE 55. STRESS-STRAIN DATA (STATIC) FOR 2CES52 (CINDERS AND EXPANDED SLAG) Load Stress Vertical Gage Reading Strain pounds psi mic-in mic-in 6 0 0 1100 0 1000 123 990 110 2000 245 870 230 3000 368 750 350 4000 490 630 470 5000 613 500 600 6000 735 370 730 7000 858 230 860 8000 981 110 990

-269 - 1600 1400 1000 - co 800 cn (I) QU It 400 w 0 (3 0 0 150 300 450 600 750 900 0 50 100 150 200 FOR HORIZONTAL GAGE STRAIN MICRO-INCH FIG. 52 - STRESS-STRAIN CURVES FOR 2CES51

-270 - 1600 1400 1200 --- — (000 800 0 150 300 450 600 750 900 --- a: 00l |tTSTRAIN MICRO OINCH FIG.53 - STRESS-STRAIN CURVE FOR 2CES02 FIG.53 STRESS-STRAIN CURVE FOR 2CES52 to N,. /-^ --- -- - ------ _ - -- >_ _ _ __ _ __ _ _ _ __ _ __ zoo - - - - - -- - - - - - - - - - - - - - - -

-271 - TABLE 56. STRESS-STRAIN DATA ( STATIC (CINDERS) ) FOR 3C51 Load Stress Vertical Gage Horizontal Gage Reading Strain Reading Strain pounds psi mic- in mic -in mic-in mic-in 5- 6.,.. 0 0 1630 0 1480 0 1000 121 1475 155 1510 -30 2000 242 1330 300 1530 -50 3000 362 1180 450 1560 -80 4000 483 4 1015 615 1585 -105 5000 6o4 -1850 780 1610 -130 6000 725 1685 945 655 -175 7000 845 1.510 1120 1700 -220 8000 966 1330 1300 1750 -270 TABLE 57. STRESS-STRAZN DATA (STATIC) FOR 3C52 (rClDEBRS ) Load Stress Vertical Gage Horizontal Gage Reading Strain Reading Strain pounds psi mic - in mic-in mic~-in mic-in 0 0 1420 0 1700 0 1000 121 1285 135 1710 -10 2000 242 1150 270 1730 -30 3000 362 51010 410 1750 -50 4000 483 -1865 555 1780 -80 5000 6o4 1720 700 1810 -110 6000 725 1560 860 1840 -140 7000 846 1400 1020 1880 -180 8000 967 1240 1180 1920 -220

-272 - 1200 1050 - 900 0 750 600 0 450STRAIN MICROINCH a. 300 0 150: 00 450 600 750 900 FOR HORIZONTAL GAGE 50 100 150 200 250 300 STRAIN MICRO-INCH FIG. 54 - STRESS-STRAIN CURVES FOR 3C51

-273 - 1200 1050 900 750 0 600 1.. _ _,-' --- —--— @CL - l450 / t LU) c - - I X, 300 --- 0 150- I- _ 0 150 300 450 6 0 50 100 150 2 STRAIN MIGRO-INCH i00 750 900 00 FOR HORIZONTAL GAGE FIG. 55 - STRESS-STRAIN CURVES FOR 3C52

-274 - TABLE 580 STRESS-STRAIN DATA (STATIC) FOR 3SG51 (SAND AND GRAVEL) Load Stress Vertical Gage Reading Strain pounds psi mic-in mic-in 7 0 0 1140 o l000 115 1095 45 2000 230 1050 90 3000 345 995 145 4000 46o 950 190 5000 576 900 240 6000 691 850 290 7000 806 800 340 8000 921 750 390 goo9000 1036 695 445 10000 1151 640 500 11000 1266 585 555 12000 1381 525 615 13000 1496 468 672 14000 1612 410 730 I - -- ---- -- -- --- ~L --- — I _I I_ -g __

-275 - 0 150 300 450 600 750 900 STRAIN MICRO -INCH FIG. 56 - STRESS-STRAIN CURVE FOR 3SG51

-276 - TABLE 59. STRESS-STRAIN DATA (STATIC) FOR 3SG52 (SAND AND GRAVEL) Load Stress Vertical Gage Reading Strain pounds psi mnic-in mic-in 7 0 O 1191 0 1000 123 1140 51 2000 246 1085 106 3000 369 1030 161 4000 493 970 221 5000 616 910 281 6000 739 855 336 7000 862 792 399 8000 985 730 461 9000 1108 667 524 10000 1232 600 591 11000 1355 535 656 12000 1479 478 713 13000 1602 4.0 781 14000 1724 343 848 I L --

-277 - 1600 1400 1000 _ _ ___ v _ _ /________________ u- 800 -- we r w _ _ 600 STRAIN MICRO-INCH FIG. 57 - STRESS-STRAIN CURVE FOR 3SG52 o 5 6; 0 5 0 /i ~ SRANMCR IC FI.57 SRSSSRIN GRE-.FR!S5

-278 - TABLE 60o STRESS-STRAIN DATA (STATIC) FOR 4SG51 (SAND AND GRAVEL) Load Stress pounds psi. ~,.,,. ~,,.,,,L _ _,,~~~~~~~~~. Vertical 0 1000 2000 3ooo 4000 5000 6ooo 7000 800ooo 9000 10000 11000 12000 13000 1400o 0 118 235 353 470 588 705 823 940 1058 1175 1293 1410 1528 1645 Vertical Reading mic-in 1670 1630 1595 1560 1530 1495 1460 1420 1388 1354 1380 1285 1250 1215 1185 -:Gage Strain mic-in 0 40 75 110 140 175 210 250 282 316 350 380 420 455 485 Horizontal Gage Reading Strain mic-in mic-in -480 0 1485 -5 149o -10 1500 -20 1510 -30 1515 -35 1523 -43 1532 -52 1540 -60 1552 -72 1565 -85 1579 -99 1590 -110 1620 -140 1650 -170

-279 - 75 0 50 100 150 200 FOR HC STRAIN MICRO-INCH FIG. 58 - STRESS-STRAIN CURVES FOR 4SG51 i0 900 )RIZONTAL GAGE

-280 - TABLE 61. STRESS-STRAIN DATA (STATIC) FOR 4SG52 (SAND AND GRAVEL) Load. Stress Vertical Gage Reading Strain pounds psi mic-in mic-in 6 0 0 1537 0 1000 118 1500 37 2000 235 1460 77 3000 353 1420 117 4000 470 1380 157 5000 588 1335 202 6000 705 1290 247 7000 823 1248 290 8000 940 1205 332 9000 1058 1160 377 10000 1175 1115 422 11000 1293 1070 467 12000 1410 1025 512 13000 1528 975 557 14000 1645 930 607

-281 - 1600 1400 1200 1000 w II I I I I i I-. 400 9o 600- - FIG. 59 STRESS-STRAIN CURVE FOR 4SG52 o400, -— 1 — FIG. 59 - STRESS-STRAIN CURVE FOR 4SG52:

-282 - TABLE 62. STRESS-STRAIN DATA (STATIC) FOR 5CS51 (CINDERS AND SAND) Load Stress Vertical Gage Horizontal Gage Reading Strain Reading Strain pounds psi mic-in mi c-in mic-in mic-in 5 0 0 1605 0 1120 0 1000 118 1505 100 1130 -10 2000 235 1400 205 1145 -25 3000 353 1290 315 1162 -42 4000 470 1180 425 1180 -60 5000 588 1070 535 1200 -80 6000 706 960 645 1220 -100 7000 824 850 755 1240 -120 8000 942 740 865 1260 -140 TABLE 63. STRESS-STRAIN DATA (STATIC) FOR 5CS52 (CINDERS AND SAND) Load Stress Vertical Gage Reading Strain pounds psi mic - in mic - in o 0 1845 o 1000 116 1780 65 2000 231 1710 135 3000 347 1630 215 4000 463 1560 285 5000 578 1485 360 6000 694 1410 435 7000 809 1330 515 8000 925 1250 595.............~59

-283 - 1200 1050 900 750 t) 600 0 -C) LJ 4) 450 U) Ln 0 0 15 300 150 0 0 150 300 450 600 750 900 0 50 100 150 200 FOR HORIZONTAL GAGE STRAIN MICRO-INCH FIG. 60 - STRESS-STRAIN CURVES FOR 5CS51

1200 1050 900 - -- - - -- ---- - --- - -- - - 750 --- —---------— P - - -- cn 600 --- -- 0o 450 4060709 75O — C F 300 -.0 150 300 450 600 750 900 STRAIN MICRO-INCH FIG. 61 - STRESS-STRAIN CURVE FOR 5CS52

-285 - TABLE 64. STRESS-STRAIN DATA (STATIC) FOR 6ES51 (EXPANDED SLAG) Load Stress Vertical Gage Horizontal Gage Reading Strain Reading Strain pounds psi mic-in mic-in mic-in mic-in 6 7 0 0 1190 o 1115 o 1000 128 1100 90 1120 -5.2000 256 1020 170 1125 -10 3000 384 940 250 1135 -20 4000 512 865 325 1145 -30 5000 640 790 400 1155 -45 6000 768 720 470 1170 -55 7000 896 640 550 1185 -70 8000 1024 560 630 1195 -80 TABLE 65. STRESS-STRAIN DATA (STATIC) FOR 6ES52 (EXPANDED SLAG) Load Stress Vertical Gage Reading Strain pounds psi mic-in mic - in 5 0 0 1780 0 1000 126 1710 70 2000 253 1635 145 3000 379 1560 220 4000 505 1480 300 5000 632 1405 375 6000 758 1330 450 7000 884 1250 530 8000 1010 1170 610

-286 - 1200 1050 900 750 T 600 -Lii — / Cn 450 C,) a. 300!! /' 0 0 150 300 450 600 750 900 0 50 100 150 200 FOR HORIZONTAL GAGE STRAIN MIGRO-INCH FIG. 62 - STRESS-STRAIN CURVES FOR 6ES51

600 - iL cff rJ 450 a. 300 0, 150 0 150 300 450 600 750 900 STRAIN MICRO-INCH FIG. 63 - STRESS-STRAIN CURVE FOR 6ES52

-288 - TABLE 66. STRESS-STRAIN DATA (STATIC) FOR 7W51 (WAYLITE) Load Stress Vertical Gage Horizontal Gage Reading Strain Reading Strain pounds psi mic-in mic-in mic-in mic-in `'~ 6..7 0 0 10o0 0 950 0 1000 121 1040 60 960 -10 2000 242 980 120 965 -15 3000 363 930 170 975 -25 4000 483 885 215 980 -30 5000 604 820 280 990 -40 6000 725 770 330 1000 -50 7000 845 710 390 1010 -60 8000 966 660 440 1025 -75 TABLE 67. STRESS-STRAIN DATA (STATIC) FOR 7W52 (WAYLITE) Load Stress Vertical Gage Reading Strain pounds psi mic-in mic-in 6 0 0 1430 0 1000 126 1370 60 2000 253 1305 125 3000 379 1240 190 4000 505 1180 250 5000 632 1120 310 6000 758 1060 370 7000 884 1000 430 8000 1010 940 490

-289 - 0 150 300 450 600 750 900 0 50 1 00 1 50 200 FOR HORIZONTAL GAGE STRAIN MICRO-INCH FIG. 64 - STRESS-STRAIN CURVES FOR 7W51

-290 - 1200 1050 900 750 c) 600 Q cn o) Ucn 1 -o) 450 LU > C) LU a 300 CD, LI --- /'-~-z zz Xt~~~~~....IIL Ii if W5TII~ir t -4 /. / 0 0 15( 3 n 0 150 300 450 600 750 900 STRAIN MICRO-INCH FIG. 65 - STRESS-STRAIN CURVE FOR 7W52

-291 - TABLE 68. STRESS-STRAIN DATA (STATIC) FOR 8CEL51 (CELOCRETE) Load Stress -Vertical Gage Reading Strain pounds psi mic-in mic-in 7 0 0 1130 0 1000 125 1000 130 2000 249 890 240 3000 374 790 340 4000 498 700 430 5000 623 610 520 6000 749 520 610 7000 874 430 700 8000 999 340 790 TABLE 69. STRESS-STRAIN DATA (STATIC) FOR 8CEL52 (CELOCRETE) Load Stress Vertical Gage Reading Strain pounds psi mic-in mic-in 5 0 0 1130 0 1000 122 1020 110 2000 244 900 230 3000 367 800 3-0 4000 489 690 440 5000 611 4 580 550 6000 733 -1470 660 7000 855 1370 760 8000 978 1250 880 '..... '....,... _,,,

-292 - 1200 1050 900 750 - 600 - - Ll- -_1 1 1 1 1- -1'II I-, / I I- -II I I - - i- ________ __________ _ MICR0- INCH -- 600 0 550 CL 300 0 150 300 450 600 750 900 STRAIN MICRO-INCH FIG. 66 - STRESS-STRAIN CURVE FOR 8CEL51

-293 - 1200 1050 900 750 Cn 600 (n co IU) 450 4~0 X 300 Z 0 0 150 0 150 300 450 6 STRAIN MIGRO-INCH STRESS-STRAIN CURVE FC 00 750 900 FIG. 67 )R 8CEL52

TABLE 70. STRESS-STRAIN DATA (STATIC) FOR 9S51 (AIR-COOLED SLAG) Load pounds 0 1000 2000 3000 hooo 5000 6ooo 6000 7000 8000 9000 10000 11000 12000 13000 14000 Stress psi 0 121 242 362 483 604 725 845 966 1087 1208 1329 1449 1570 1691 Vertical Gage Reading Strain mic - in 2 1890 1830 1775 1725 1670 1615 1560 1500 144o 1375 1310 1240 1170 1100 1020 mic -in 0 6o 115 165 220 275 330 390 450 515 580 650 720 790 870 - Horizontal Gage Reading Strain mic-in mic-in a 5 1230 0 1235 -5 1240 -10 1244 -14 1250 -20 1260 -50 1265 -35 1270 -40 1280 -50 1290 -60 1300 -70 1315 -85 1330 -100 1350 -120 1370 -140 -- -- ---

-296 - TABLE 71. STRESS-STRAIN DATA (STATIC) FOR 9S52 (AIR-COOLED SLAG) Load Stress Pounds psii 0 1000 2000 5000 4000 5000 6ooo 7000 8000 9000 10000 11000 12000 13000 14000 0 118 236 354 472 590 708 826 944 1062 118o 1298 1416 1534 1652 Gage Reading;mic -in 6 1040 990 945 900 85o 805 755 705 652 600oo 550 486 432 362 300. mic -in 0 50 95 140 190 255 285 535 388 44a| 490 554 6o8 678 740 Strain

-297 - 1600 1400 1000 n 800 --- —-- a. (yI 600 Ir a 400 0 0 150 300 450 600 750 900 STRAIN MICRO-INCH FIG. 69- STRESS-STRAIN CURVE FOR 9S52

-2918 - TABLE 72. STRESS-STRAIN DATA (STATIC) (CINDERS PLUS FLY ASH) FOR 10CA51 Load Stress Vertical Gage Horizontal Gage Reading Strain Reading Strain pounds psi mic-in mic-in mic-in mic-in 6 5 0 0 1280 0 545 0 1000 126 1150 130 550 -5 2000 253 1000 280 570 -25 3000 379 840 440 590 -45 4000 505 675 605 610 -65 5000 631 5 495 785 655 -110 6000 758 -1310 970 700 -155 7000 884 1100 1180 755 -210 8000 1010 840 1440 840 -295 TABLE 73. STRESS-STRAIN DATA (STATIC) FOR 10CA52 (CINDERS PLUS FLY ASH) Load Stress Vertical Gage Reading Strain pounds psi mic-in mic-in 0 0 1370 0 1000 126 1245 ' 125 2000 253 1125 245 3000 379 990 380 4000 505 860 510 5000 632 720 650 6000 758 570 800 7000 884 400 970 8000 1010 -1220 1150 _ ~ ~.:........

-299 - 1200 1050 900 750 - 6001 - - -{ ' 4/00 4U --- -------- 0 -- a-300 0 -— _ __ _ 150 --- —------------- 0 150 300 450 600 750 900 0 50 100 15O 200 FOR HORIZONTAL GAGE STRAIN MICRO-INCH FIG. 70- STRESS-STRAIN CURVES FOR I0 CA51

-300 - 1200 1050 900 750 0 600 / ~1 —11111 a,, EEETT 1 I I 11 I I I I w 450 ---w o 7 150 0 150 300 450 600 750 900 STRAIN MICRO-INCH FIG. 71 - STRESS-STRAIN CURVE FOR IOCA52

APPENDIX VII THERMAL EXPANSION As was previously discussed in Chapter XII, the coefficients of thermal expansion were determined for saturated (not vacuum saturated) and oven dried specimens. Two methods of tests were used for measuring length change. These were the dial gage extensometer and the variable resistance wire gage (SR-4) which were both described in Chapter XII. Both methods of test were used for the dry specimen but the first method only was used for the wet specimen. Table 74 gives the coefficients of the thermal expansion of saturated concrete for the various stages of heating and cooling. Each value represents the average of the 5 cycles of the test; their average is considered the thermal coefficient of the specimen. It may be seen from the table that the coefficients when heating are slightly higher than those when cooling. The actual extensometer readings are not given. Table 75 gives the coefficient of thermal expansion for the dry specimens by the first method of test. The discussion given above holds for this table except that the cycle in this latter test consists of one stage of heating (100 to 162 F) and one stage of cooling (162 to 100 F). Table 76 gives the coefficients of expansion of dried concrete by the second method of test. Results were calculated from equation (16a) which was presented in Chapter XII. The values AR

-302 -anad Ar in the table are the average of 5 complete heat-cool cycles. Again the actual readings of the strain indicator were not given.

-303 - TABLE 74. CALCULATION OF THE COEFFICIENT OF THERMAL EXPANSION OF WATER SATURATED SPECIMEN (inch/inch/l F) x 10I Heating Cooling Average Specimen 35-73~ 73-110~ 110-140~ 140-73~ 73-35~ Individual Both F F F F F Specimens 1C61 1C62 3.4 3.5 3.4 3.5 2CES61 2CES62 4.3 4.4 3C61 3C62 3SG61 3SG62 4SG61 4SG62 5CS61 5CS62 6Es61 6ES62 7W61 7w62 8CEL61 8CEL62 9S61 9S62 10CA61 10CA62 3.5 5.5 4.7 5.0 5.0 5.5 3.9 3.8 5.5 5.4 4.5 4.6 5.6 5.7 4.7 4.8 3.1 3.0 3.4 3.5 4.7 5.0 5.1 5.5 4.1 3.7 5.5 5.5 4.6 4.7 5.6 5.7 4.7 4.9 3.2 3.0 3.5 3.6 4.5 4.6 3.4 3.5 4.8 5.2 5.1 5.3 4.1 3.8 5.5 5.7 4.0 4.3 3.2 3.4 4.6 4.9 4.9 5.2 3.9 3.6 5.3 504 4.1 4.3 3.3 3.3 3.4 3.5 3.4 4.3 3.2 3.5 4.7 4.9 5.0 5.5 4.0 3.7 5.3 5.5 4.4 4.6 5.6 5 6 3.5 3.5 4.7 5.0 5.0 5.5 4.9 5.2 4.0 3.7 4.6 4.7 5.6 5.8 5.5 5.6 5.5 5.5 4.5 4.6 5.6 5.7 4.7 4.9 3.1 3.0 5.4 4.6 5.7 4.8 4.8 4.9 4.6 4.9 3.2 3.0 3.0 3.0 3.0 3.0 3.1 - --

-305 - TABLE 76. CALCULATION OF THE COEFFICIENT OF THERMAL EXPANSION OF DRY SPECIMEN Test Method No. 2 (inch/inch/1 F) x 106 Coeff. of Thermal Expansion -ARc* ARm* -Rc + ARm Average x 106 Saple x 6 x 106 + 588 Individual** of 2 x 106 Specimens --, i - ~^ '_ 1C61 1C62 2CES61 2CES62 3C61 3C62 3SG61 3SG62 4SG6l 4SG62 5cs61 5Cs62 6ES61 6ES62 7w61 7W62 8CEL61 8CEL62 9S61 9S62 425 420 360 360 430 440 350 355 305 520 385 380 295 280 350 320 290 285 320 315 3 0 166 171 3 0 231 231 2.8 2.9 3.9 3.9 2.7 2.5 4.0 3.9 3 0 3 0 3 0 3 0 161 151 241 236 286 271 3 0 3 0 3 0 206 211 296 311 241 271 301 306 3.4 3.5 4.9 5.2 4.0 4.5 50o 501 4.5 4.6 2.5 2.5 2.9 3.9 2.6 4.0 4.7 3.5 5.1 4.3 5.0 4.6 2.5 3 0 3 0 271 276 151 151 10CA61 10CA62 *Values of Rc and Rm are averages **See equation 16a (Chapter XII). of 5 complete cycleso

APPENDIX VIII THERMAL CONDUCTIVITY Ao Hot Plate Apparatus. The cut-away cross-sectional plan of the hot plate is shown in Figure 72. The windings of the central heater and of the guard ring are exposed for the right half of the plate, which is symmetrical with the left half. The heater wires are embedded in 1/16 inch grooves, cut in 1/4 inch transite board, spaced 3/8 inch apart. As has been discussed in Chapter XIII, the winding of the central heater is completely separate from that of the guard ring. The forme74 consists of 28.5 ft. of No. 33 nickel-chromium wire, which has a resistance of 5.739 ohms per foot; the latter consists of approximately 15 ft. of No. 25 Chromel-A wire. The windings are covered with sheet mica for insulation and the whole covered in turn with a copper plate 1/4 inch thick for protection and to minimize surface temperature differences. The copper plate consists of two parts: the central portion 9:-inch square covering the central heater, and the 1-7/8 inch wide outside (guard) ring. They are separated with a 1/8 inch air gap. The opposite side of the hot plate is identical to the sfde shown in Figure 72. Other information is given in Chapter XIII. B. Thermocouple Location. In the present tests, temperatures were measured by thermocouples made of thin (No. 30 A.W*,G) copper-constantan wires. The 3i06

-307 - FIG.72 CUT-AWAY PLAN FOR HOT PLATE

-308 - manufacture and the procedure of using these thermocouples were discussed in parts C and D of Chapter XIIIo Ten thermocouples were used for each of the two companion specimens, tested simultaneously, for each product. The locations of these thermocouples are shown in Figure 73. In this figure, the companion specimens are denoted by numbers 1 and 2, and the thermocouples are numbered from 1 to 20 designating the pole numbers of the switch box to which each thermocouple was connected. C. Discussion of Test Results. Results were calculated in Table 28 and summarized in Table 29; both tables are presented in Chapter XIII. Discussion of the results was also presented in part F of the chapter. However, the final individual current and temperature determinations are presented here, in Table 77. This table contains the measurements for two runs per product. In each case the values shown are those at the end of at least five hours of steady flow of heat through the specimens. The thermocouple's numbers in this table designate their locations, which may be seen in Figure 73. It may be noted (see Table 77) that thermocouples No. 3 of the central heater and No. 4 of the guard ring consistently indicated lower temperatures than the remaining thermocouples. Pilot tests showed that this temperature variation is localized to a small area around the thermocouples, and is believed to be caused by some deficiency of the winding at this points Because the area concerned is very small, the effect of this variation on the results should not be important. The actual readings of thermocouple No. 3 were included in the calculations of'the results.

30 9 COLD PLATE HOT PLATE SFECIMEN NO. I I' I I I LI~J i d~~~~ I *s -- - -. — 0 -..... Il........ _.. n COLD PLATE HOT PLATE SPECIMEN NO. 2 FIG.73 LOCATION OF THERMOCOUPLES TWO COMPANION SPECIMENS FOR

TABLE 77. FINAL CURRENT AND TEMPERATURE DETERMINATIONS FOR THERMAL CONDUCTIVITY TESTS (Two Runs per Product) Current Input Temperature in Degrees Fahr. Sample Central Heater Guard Thermocouples of Specimens No. 1 Thermocouples of Specimens No. 2 Room Ring Hot Plate Cold Plate Hot Plate Cold Plate Temp. No~ OF volt amp volt 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 65.39 33 124 124 121 121 125 125 56.5 56 56.5 55.5 124 124 124 124 125 125 57 58 57 58 78 65.39 35 125.5 125 121 121 125.5 125.5 56 56 57 56 12.5.5 125.5 124 126.5 126.5 126.5 57.5 57 57.5 57 79 65.39 36 124 124.5 121 121 124 124 57 56 57 56 124 124.5 124 124.5 124 124 57 56 56 57 79 70.42 38 135 135.5 130 130 135 135 56.5 56.5 57 57 135 135 135 135 135 135 56.5 57.5 57.5 57 80 3C10 65.38 32 126.5 126.5 120 120 126 126 56.5 55.5 56.5 55.5 126 126.5 126 126.5 126 126 58 57 57 58 78 70.42 36 140 140.5 137.5 138 139.5 140 57 56 57 59 139 139.5 139 139.5 139 139 58 57 58 57 79 3SGO1 75.43 39 111 111.5 o106 107 111 ill 64 64 65 64 111 111.5 111.5 111.5 111 109 64 65 65 64 78 80.48 41 119 120 110 110.5 119.5 118.5 65 64 65 64 119 119.5 119.5 120 119 113 65.5 65 65 65 80 4SG10 80.48 40 114.5 115 113 114 114.5 115 60 59 60 59.5 114.5 115 115 115 115 115.5 60 59.5 60.5 6o0 79 h 8o.48 40 115.5 116.5 113.5 114 116 115.5 61 60 61 60 116 117 115.5 116 116 116 61.5 60.5 62 6o0 78 60.36 32 115 116 116 116.5 115 116 59 58 59 58 114 115 117 117 117 117 59 58 59 58 79 70.42 38 130 131 130.5 131 131.5 130.5 61 6o 61 6o 128 129 132.5 133 132 132 59 59 60 59 79 6Esio 55.33 32 127 127.5 127 127 127 127 53.5 52.5 53.5 52.5 128.5 130 126.5 127 128 128 53 52 53 52 81 60.36 32 136.5 136.5 137 136 136.5 136.5 52 53 53 52 137.5 137.5 136 135 136.5 136.5 52 52 52 58 79 7WlO 50.30 27 113 114 111.5 111 114 114 54 53 54 53 114 114 114 114 114.5 114 54 53.5 54 53.5 79 0 6o.36 32 135.5 135.5 132 132 132.5 132.5 56 55 56 55 136.5 137 136.5 137 137 156.5 55.5 56.5 55.5 56.5 79 8CEL10 50.30 25 120.5 120 117 118 120.5 120.5 52.5 51.5 52 52 120 120.5 120.5 121 120.5 120.5 55 52 52 52 79 6o.36 34 147 147 143 144 147 148 54 53 54 53 146.5 147.5 147.5 148 147.5 148 56 55 56 55 79 o 6o.36 34 122 122.5 121 121 121 122 57.5 56.5 57 57 123 123 122 122 119 122 57 57 57.5 56.5 79 70.42 36 140 139 137 139 139.5 14o.5 56 55.5 56 56.5 138.5 139 139.5 140 137 139.5 56 55 55 55 79 0CA10 55.33 28 105.5 106 103 103 105.5 105 57.5 58 58 57.5 105 105 105.5 o106 105 105.5 59 58 59 58 79 70.42 38 135 136 131 130 135 135 60 61 61 6o 134 135 134 135 154.5 134.5 6o 61 61 60 8o

-311 - TABLE 78. THERMAL CONDUCTIVITY TEST WEIGHTS OF SPECIMEN BEFORE AND AFTER TEST (pounds) Sample Noo 1C10 2CES10 3C10 3SG10 4SG10 5cso1 6ES10 7W10 8CEL10 9S10 10CA10 Specimen No. 1 Before Test After Test 25.12 25.16 25.50 25.54 27.59 27.61 40.20 40.22 38.89 38,94 28.40 28.43 23.96 23596 27.14 27.16 25.90 25.96 36.40 36.42 24.76 24.79 Specimen No. 2 Before Test After Test 28.14 28.16 28.38 28.41 24.51 24,57 36.16 36.20 35.11 35.14 26.11 26.10 27.52 27.55 30.06 30.10 21.58 21.62 32.40 32.41 28.96 28.99

LIST OF REFERENCES 1. Christensen, Eniar, "Cinder as Concrete Aggregate," Report of A. C, I. Committee 203, A. C. I. Proceedings, Vol. 27, p. 583, February, 1931. 2. Pit and Quarry, p. 123, November, 1952. 3. Smithwyck, S. Carl, "Expanded Shale," Rock Products, p. 136, January, 1955. 4. "Zonolite," Rock Products, p. 60, July, 1934. 5. Pit and Quarry, p. 121, January, 1955. 6. Pit and Quarry, p. 198, January, 1947. 7. Ehle, J. C,, "Development in the Manufacture and Technology of Concrete Masonry Units," A C.. Proceedings, Vo 45, p. 613, April, 1949. 8. Woodworth, P. M., "Recommended Practice for the Manufacture of Concrete Blocks and Building Tiles," Report of A. C. I. Committee 708, A, C. I. Proceedings, Vol. 26, p. 491, February, 1930. 9. Allan, W, M., "Shrinkage Measurements of Concrete Masonry," A, C. I. Proceedings, Vol. 26, p. 699, April, 1930. 10. Richart, F. E., "Structural Performance of Concrete Masonry Walls," A. C. I. Proceedings, Vol. 28, p. 363, February, 1932. 11. Menzel, C. A., "The Strength of Concrete Masonry Walls After Standard Fire Exposure," A. C, I. Proceedings, Vol. 29, p. 113, November, 1932. 12. Wendt, K, F., and Woodworth, P. M., "Test of Masonry Units Using Tamping and Vibration Molding Methods," A. C. I. Proceedings, Vol. 36, p. 121, November, 1939. 13. Seaton, S. G., "Study of Causes and Preventions of Staining and Popouts in Cinder Concrete," A. C. I. Proceedings, Vol. 44, p. 361, January, 1948. 14. "Physical Properties of High-Pressure Steam Cured Concrete Blocks," Progress Report of A. C. IX Committee 716, A. C. IA Proceedings, Vol. 49, p. 745, April, 1953. 15. Kalousek, G. L., and Others, "Relation of Shrinkage to Moisture Content of Concrete Blocks," A. C. I. Proceedings, Vol. 50, p. 225, November, 1953. 312

- -313 - LIST OF REFERENCES (Cont ) 16. Richart, F. E., and Jensen, V. P., "Test of Plain and Reinforced Haydite Concrete," ASTM Proceedings, Vol. 30, Part 2, p. 674, See also by same authors: "Construction and Design Features of Haydite," A. C, I. Proceedings, Vol. 27, p. 151, October, 1930. 17. Washa, G. W., and Wendt, K. F., "The Properties of Lightweight Structural Concrete Made With Waylite Aggregate," A. C. I. Proceedings, Vol. 38, po 505, June, 1942. 18. Petersen, P. H., "Burned Shale and Expanded Slag Concrete With and Without Air-Entraining Admixture," A. C. I. Proceedings, Vol. 45, p. 165, October, 1948. 19. "Lightweight Aggregate Concrete," Housing and Home Finance Agency Booklet, August, 1949. 20. Kluge, R. W., Sparks, M. M., and Tuma, E. C., "Lightweight Aggregate Concrete," A. C. I. Proceedings, Vol. 45, p. 625, May, 1949. 21. Price, W. H., and Cordon, W. A., "Test of Lightweight Aggregate Concrete Designed for Monolithic Construction," A, C. I. Proceedings, Vol. 45, p. 581, April, 1949. 22. ASTM Standards, Part 3, 1952, and its 1953 and 1954 Supplements: Designation C29-42: Designation C33-52: Designation C40-48: Designation C90-52: Designation: C117-49: Designation C127-42: t-'andard Method::of:':Test for: Unit 'Weight of Aggregate, " ' 55. "Tentative Specifications for Concrete Aggregates," p. 897. "Standard Method of Test for Organic Impurities," p. 942. "Standard Specifications for Hollow Load-Bearing Concrete Masonry Units," p. 457. "Standard Method of Test for Amount of Material Finer Than No. 200 Sieve in Aggregate," p. 933. "Standard Method of Test for Specific Gravity and Absorption of Coarse Aggregate," p. 967.

-314 - LIST OF REFERENCES (Cont.) Designation C128-42: "Standard Method of Test for Specific Gravity and Absorption of Fine Aggregate," p. 969. Designation C136-46: "Standard Method of Test for Sieve Analysis of Fine and Coarse Aggregates," p. 954. Designation C140-52: "Standard Methods of Sampling and Testing Concrete Masonry Units," p. 465. Designation C151-52: "Standard Method of Test for Autoclave Expansion of Portland Cement," p. 52. Designation C177-45: "Standard Method of Test for Thermal Conductivity of Materials by Means of the Guarded Hot Plate," p. 848. Designation C215-52T: "Tentative Method of Test for Fundamental Transverse and Torsional Frequencies of Concrete Specimens," p. 1072 Designation C293-54: "Tentative Method of Test for Flexural Strength of Concrete (Using Simple Beam With Center-Point Loading)," 1954 Supplement, p. 264~ Designation C331-53T: "Tentative Specifications for Lightweight Aggregates for Concrete Masonry Units," 1953 Supplement, p. 233. Designation C341-54T: Designation D75-48: "Tentative Method of Test for Change of Concrete Products," plement, p. 283. "Standard Methods of Sampling Slag, Gravel, Sand, and Stone Use as Highway Materials," p. Volume 1954 SupStone, Block for 9490 Designation E12-27: "Definition of Terms Relating to Specific Gravity," p. 1594. 23. "Iron Blast-Furnace Slag Production, Properties and Uses," U. S. Department of the Interior, Bureau of Mines Bulletin #479, 1949.

-315 - LIST OF REFERENCES (Cont.) 24. "Production of Lightweight Concrete Aggregate from Clay, Shale, Slate and Other Materials," U. S. Department of the Interior, Bureau of Mines Bulletin # R 14401, November, 1948. 25. "Report on Significance of Tests of Concrete and Concrete Aggregates," ASTM Special Technical Publication # 22A, 1943. 26. "Symposium on Mineral Aggregates," ASTM Special Technical Publication # 83, 1948. 27~ "American Standard Basis for the Coordination of Dimensions of Building Materials and Equipment," A. S. A, Publication # A62-1, 1946. 28. "American Standard Sizes of Clay and Concrete Modular Masonry Units," A. S. A. Publication # A62-3, 1946. 29. "American Standard Building Code Requirements for Masonry," U;. -:S.?Department of iCommerce, National Bureau of Standards liscellaneous Publication #211, July, 1954. 30. Tucker, John Jr., "Statistical Theory of the Effect of Dimension and Method of Loading Upon the Modulus of Rupture of Beams," ASTM Proceedings, Vol. 41, p. 1072, 1941, 31. Wright, P. J. F., and Garwood, F., "Effect of Method of Test on the Flexural Strength of Concrete," Magazine of Concrete Research (London), p. 67, October, 1952. 32. Nielsen, K. E. C., "Effect of Various Factors on the Flexural Strength of Concrete Test Beams," Magazine of Concrete Research (London), p. 105, March, 1955. 33. Reagel, F. V., and Willis, T, Fo, "The Effect of the Dimensions of Test Specimens on the Flexural Strength of Concrete," Public Roads, Vol. 12, p. 37, April, 1931. 34. Timoshenko, S., and MacCullough, Element of Strength of Materials, Part I, p. 120, 1949. 35. Timoshenko, S., Vibration Problems in Engineering, D. Van Nostrand and Company, Inc., pp. 331-342, July, 1937. 36. Pickett, G., "Equations for Computing Elastic Constants from Flexural and Torsional Resonant Frequencies of Prism and Cylinders," ASTM Proceedings, Vol. 45, pp. 846-865, 1945.

-316 - LIST OF REFERENCES (Cont ) 37. Du Mont Cathode-Ray Oscillograph (Type 292) Operating and Maintenance Manual, " A Manual by the Allen B, Du Mont Laboratories, 1951. 38. Obert, Lo, and Duval, We I., "Discussion of Dynamic Methods of Testing Concrete With Suggestions for Standardization," ASTM Proceedings, Vol. 41, p. 1053, 1941. 39. Philleo, R. E., "Comparison of Results of Three Methods for Determining Young's Modulus of Elasticity of Concrete," A. C. I. Proceedings, Vol. 51, p. 461, January, 1955. 40. Simmons, J. C., "Poisson's Ratio of Concrete: A Comparison of Dynamic and Static Measurements," Magazine of Concrete Research, Volo 7, # 20, p. 61, July, 1955. 41. Gilkey, H. J., and Vogt, F., on Discussion of Paper by R. E. Davis and G. E. Troxell (Modulus of Elasticity on Poisson's Ratio for Concrete, and the Influence of Age and Other Factors Upon These Values), ASTM Proceedings, Vol. 29, p. 703, 1929. 42. Woods, H., "Observations on the Resistance of Concrete to Freezing and Thawing," A. C. I. Proceedings, Vol. 51, p. 345, December, 1954. 43. Powers, T. Co, and Helmuth, R. S., "Theory of Volume Changes in Hardened Portland Cement Paste During Freezing," Proceeding of Highway Research Board, Vol. 32, p. 285, 1953. 44. Withey, M. 0., and Others, "Symposium of Freezing and Thawing Tests of Concrete," ASTM Proceedings, Vol. 46, p, 1199, 1946. 45. Walker, S., and Bloem, D. L., "Performance of Automatic Freezing and Thawing Apparatus for Testing Concrete," ASTM Prodeedings, Vol. 51, p. 1120, 1951. 46. Kennedy, T. B., and Mather, K., "Correlation Between Laboratory Accelerated Freezing and Thawing and Weathering at Treat Island, Maine," A. C. I. Proceedings, Vol. 50, po 141, October, 1953. 47. Jackson, F. H., 'Long Time Study of Cement Performance in Concrete. Chapter 9 - Correlation of the Results of Laboratory Tests With Field Performances Under Natural Freezing and Thawing," A, C. I. Proceedings, Vol. 52, p. 159, October, 1955. 48. Sweet, H. S., "Research on Concrete Durability as Affected by Coarse Aggregate," ASTM Proceedings, Vol. 48, p. 988, 1948.

-317 - LIST OF REFERENCES (Cont. ) 49. Callan, E. J., "Thermal Expansion of Aggregate and Concrete Durability," A. C. I. Proceedings, Vol. 48, p. 485, February, 1952. 50. Walker, S., and Others, Discussion of Paper by E. J, Callan (reference 49), A. C. I. Proceedings, Vol. 48, p. 504, 1952. 51L Higginson, E. C., and Kretsinger, D. Go, "Prediction of Concrete Durability from Test of Aggregate, " ASTM Proceedings, Vol. 53, P. 991, 1953. 52. Legg, F. E. Jr., "Freeze-Thaw Durability of Michigan Concrete Coarse Aggregates," Submitted for Publication in Highway Research Board Proceedings, Vol. 35, p. 950. 53. Kalousek, G. L., "Fundamental Factors in the Drying Shrinkage of Concrete Blocks," A. C. I. Proceedings, Vol. 51, p. 233, November, 1954. 54. Washa, G. Wo, Saemann, J, C,, and Warren, C., "Effect of Curing on the Properties Affecting Shrinkage Cracking of Concrete Block," A. C. I. Proceeding, Vol 51, p. 833, May, 1955. 55. Easterly, W. Jr., "Correlation of Shrinkage and Curing in Concrete Masonry Units," A, C. I. Proceedings, Vol. 48, p. 393, January, 1952. 56. Pickett, G., "Shrinkage Stresses in Concrete," A. Ce I. Proceedings, Vol. 42, Part I, p. 165, and Part II, p. 360, January and February (respectively), 1946. 57. Powers, To C,, and Brownyard, T.L,, ', -tuis ic X P....lical:. 'PrIortis of Hardened Cement Paste - Part 3. Theoretical Interpretation of Absorption Data," A, C. I. Proceedings, Vol. 43, p. 469, December, 1946. 58. Menzel, C. Ao, "A Method for Determining the Moisture Condition of Hardened Concrete in Terms of Relative Humidity," ASTM Proceedings, Preprint # 83, 1955. 59. 'Concrete Manual, Us S Department of the Interior, Bureau of Reclamation, 1955. 60. Mitchell, L. J,, "Thermal Expansion Tests on Aggregates, Neat Cements and Concretes," ASTM Proceedings, Vol. 53, p. 963, 1953.

-318 - LIST OF REFERENCES (Cont.) 61. Bonnell, D. R., and Harper, F. C., "The Thermal Expansion of Concrete," Journal of the Institution of Civil Engineers, No. 4, p, 320, February, 1950. 62. Pearson, J. C., "A Concrete Failure Attributed to Aggregate of Low Thermal Coefficients," A. C. I. Proceedings, Vol. 38, p. 29, September, 1941. 63. Hornibrook, F. B., and Others, Discussion of Paper by J. C. Pearson (reference 62), A. C. I. Proceedings, Vol. 38, p. 36, June, 1942. 64. Walker, S., Bloem, D. L., and Mullen, W. G., "Effect of Temperature Changes on Concrete as Influenced by Aggregates," A. I. Proceedings, Vol. 48, p. 661, April, 1952. 65. Heating, Ventilating, Air-Conditioning Guide, American Society of Heating and Air-Conditioning Engineers, Vol. 33, 1955. 66. Norton, C. L., "Some Thermal Properties of Concrete," Journal of the American Society of Mechanical Engineers, Vol. 35, p. 1011, 1913. 67. Willard, A. C., and Lichty, L. C., "A Study of the Heat Transmission of Building Materials," University of Illinois Engineering Experiment Station Bulletin #102, 1917. 68. Carman, A. P., and Nelson, R. A., "The Thermal Conductivity and Diffusivity of Concrete " University of Illinois Engineering Experiment Station Bulletin 122, 1921. 69. Houghten, F. C., and Cutberlet, C., "Conductivity of Concrete," ASHVE Transaction, Vol. 38, p. 47, 1932. 70 ippon. S., and Snyder, L. J., "Thermal Properties of Mass Concrete," A C. I. Proceedings, Vol. 30, p. 35, SeptemberOctober, 1933. 71. Rowley, F. B., "Transmission of Heat Through Building Materials," University of Minnesota Engineering Experiment Station Bulletin 3, 1933. 72. Rowley, F. B., and Algren, A. Bo, "Heat Transmission Through Building Materials," University of Minnesota Engineering Experiment Station Bulletin #8, 1932. 73. Rowley, F. B., and Algren, A. B., "Thermal Conductivity of Building Materials," University of Minnesota Engineering Experiment Station Bulletin #12, 1937.

-319 - LIST OF REFERENCES (Conc.) 74. McAdams, Heat Transmission, McGraw-Hill Publishing Company, Third Edition, pp. 1-30, 1954. 75. Cejka, J. B., Hunt, M. W., and McKinnon, J. C., "Report on the Guarded Ring Heat Transfer Apparatus," 1939. Unpublished report made by the authors, to whom the development of the apparatus was originally assigned, to the Department of Mechanical Engineering, University of Michigan. Report obtained by private communication from F. N. Calhoon, Professor, Mechanical Engineering, University of Michigan. 76. Mitchell, L. J., "Thermal Properties," ASTM Special Technical Publication 169 (Significance of Tests and Properties of Concrete and Concrete Aggregates), 1956.

UNIVERSITY OF MICHIGAN 3 9015 02829 9561