ENGINEERING RESEARCH INSTITUTE THE UNIVERSITY OF MICHIGAN Technical Report TESTS OF FLEXICORE SLABS MADE OF ELASTIZELL CONCRETE Alex Eo Mansour Engineer, Elastizell ebrpio ieonoi~ America and L. M. Legetstit Professor of Civil Migiirke.*ing University of Mithijn. Project 2326 ELASTIZELL CORPORATION OF AMERICA ALPENA, MICHIGAN November, 1957

TABLE OF CONTENTS Page INTRODtJCTION..... *. 1 SUMMARY............... 2 TEST SPECIMENS..... *. 3 TEST PROCERE.......... 4 TEST RESULTS........... 5 DISCUSSION............ 6 DESIGN COMPUTATIONS........... 8 CONCLUSION. 12 SUPPL1iENT...... *. ~ ~ ~ ~ 13 APPENDIX........ 1... * * * 17 Figure 1. Total Load vs. Deflection at Center Line.... 18 Figure 2. Total Load vs. Deflection at Center Line.... 19 Figure 3. Total Load vs. Strain in Concrete at Center Line... 20 Photo. 1. Test Specimen after Failure. * 21 Photo. 2. View of Loading System.... 21

TESTS OF FLEXICORE SLABS MADE OF ELASTIZELL CONCRETE INTRODUCTION "Flexicore" is the trade name applied to a type of hollow precast concrete slab used for floor and roof construction. In structural action the Flexicore member is essentially a T beam. The concrete stress almost never controls in T beams. Therefore, if a concrete having other desirable properties, such as light weight and lower thermal conductivity, is to be used in Flexicore it need not be high strength concrete. It must be strong enough to resist the calculated T beam stresses and stiff enough to satisfy deflection requirements. Elastizell concrete is a cellular concrete which can be made in a wide range of densities by replacing a varying amount of fine aggregate by tiny air cells, thus reducing the weight of the mix. As the weight of the material is decreased by the addition of air the compressive strength and the thermal conductivity are also decreased. A mix having adequate strength for Flexicore members can be made at a weight saving of about 25% and with a 45% decrease in thermal conductivity without the necessity of storing and using lightweight aggregates.

The tests discussed in this report were made to study the structural behavior of Flexicore slabs made of Elastizell concrete. The work was sponsored jointly, by Elastizell Corporation of America and the Price Brothers Company, Michigan Flexicore Division. The test specimens were made at the Price Brothers' plant in Livonia, Michigan,, SUMMARY Four specimens of the Flexicore standard section, designation S44, were made of Elastizell concrete on July 23, 1957 and tested on September 3 and 4. The specimens had been air dried after an initial 12 hour steam cure. The test results indicate that, for most spans and member sizes, Elastizell concrete may be used in Flexicore slabs to reduce dead load and.still carry the same allowable live load as when heavy concrete is used. For certain spans, when deflection controls the design, the allowable live load may be slightly less than for the same slab made of heavy concrete, Comparative calculations are shown for the analysis of a Flexicore slab made of stone aggregate concrete and one made of Elastizell concrete. A supplement is included describing the first commercial application of Elastizell concrete to the manufacture of Flexicore slabs.

-3TEST SPECIMENS The four test specimens were manufactured at the Flexicore plant in Livonia, Michigan and were of the following cross section: ~JYM. /o'' ~'4 "/- -' 3' 7 d The Elastizell concrete used in these specimens was proportioned to weigh 110 per out of the mixer. However, the unit weight of the concrete in the beams after drying was 120 pcf and in the test cylinders was 11 perf. This means that with handling and vibration, the Elastizell concrete increased in density by approximately 10 to 15 pe.e This problem can be solved in future work by proportioning the concrete to weigh 10 to 15 pcf lighter at the mixer than is desired in the beam. With handling and vibratiorn, the concrete will increase in weight to yield the desired density in the beam.

-4The text specimens weighed 9 pounds per lineal foot less than similar beams made with ordinary concrete. This represents a 16,% saving in dead load instead of 23% as intended. The theoretical computations based on the cylinder strengths and moduli are not representative of the concrete in the beams. Past experience has shovrn that iwith a 5 pcf increase in density, the strength of this concrete increases 500 psi and the rmodulus of elasticity increases 285,000 psi. Since the strength anc moduli obtained Prom the cylinders are used in the conmputations, the allowable loads presented for the beams are somewhat in error on the safe side. After vibration, the specimens w-ere cured in a low pressure steam kiln for approaximiately 12 hours. After remc.val from the forms, the specimens were stored in an outdoor storagte area until several days prior to testing. TEST PROCEDURES The s)ecimens were placed on simiile supports at 12'-0 centers and were loaded with equal1 concentratC ed linear loads at the 1/3 points. The following' measurements were made at each increment of load. 1. Deflection of specimlen at the center line, 2. Strain in the concrete at the center line. The photographs in the Appendix diow ths testing appsatus and the method cf applying lcad. Photograph No. 1 shows a test specimen after failure in diagonal tension at the far end. The wires leading from the specimen are attached to

an SR-4 Strain Ga-ge located on the compression face of the member at the center' line, Tlis -age was used to rmeasure thle strain in the concrete at each increiment of load. A deflection cage is located under the specimen at the center line and was used to measure the deflection of the mrember at each increment of load. Photograph NTo. 2 sl-iows in creater detail the method of applying load to the specimens. The rollers beneathi the two load beamrs were not used with Spoe cilin TTo. 1. TEST RESULTS.Each specimen was prosrressively loaded to failure with results as illustrated on Figures 1, 2, and 3 in the Appendix SPECIMEN NO. 1 This specimen failed at a total live load of 4,074-. The initial failure was caused by yieiding of the steel in the center 1/3 of the beam and was followed by failure of the concrete in conrrpression in the center 1/3. ThLere were no rollers under the load beams at the 1/3 points for this specimenm only. SPECIIEN NO. 2 At a total live load of 3,784#, this specimen failed by yielding of the steel in the center 1/3 of the beama A diagonzal crack formed at 18" from one end of the beam at a total live load of 3,63L-, and at failure, this craclk was very much advancedo

-6SPECIMEN NO. 3 This specimen was damaged prior to testing. It was badly cracked on its upper surface, apparently from shrinkage and aggravated by rough handling. A horizontal crack was visible along the middle of the keyway. The failure of this specimen occurred in diagonal tension at a total load of 3,234V- and extended along the horizontal crack in the keyway. SPECIMEN NO. 4 This specimen also failed in diagonal tension with the center of the diagonal crack 16" from the support. The total live load at failure was 3,584#. DISCUSSION As may be seen in the Design Computations on Pages 8, 9, 10, 11 and 12 and also in Figures 1, 2 and 3 in the Appendix, Elastizell concrete may be used in Flexicore slabs. The advantage to be gained from the use of Elastizell concrete in Flexicore slabs is primarily a reduction in the weight of the slabs with no sacrifice of live load. There may actually be an increase in the allowable live load as shown in the Design Computations, The reduction of dead load in any precast concrete deck system produces economies throughout the structure in which such a system is used. The reduction of dead load in a floor or roof naturally permits a saving in the supporting beams, columns, and footings.

-7The production of the test specimens at the Livonia plant required no change in the sequence of manufacturing operations nor in the equipment used. For continuous production of Elastizell Flexicore slabs, the only additional equipment required is a foam generation unit which may be operated from the compressed air source presently located in the Livonia plant. The use of Elastizell concrete in Flexicore slabs normally will require the use of stirrups to a greater extent than is presently required with the use of ordinary concrete. This point is clearly illustrated in the tests. Specimens No. 3 and 4 failed in Diagonal Tension. The design of Flexicore slabs using Elastizell concrete is identical to that for ordinary concrete except that the proper structural values for Elastizell concrete must be used. These values have been determined by test. Since Elastizell concrete displays a higher drying shrinkage than ordinary concrete, it may be desirable to use welded wire fabric rather than bars as compressive reinforcement. Properly selected welded wire fabric normally is more effective than bars in controlling shrinkage cracks.

-8DESIGN COMPUTATIONS CROSS SECTION OF FLEXICORE SLAB S44 b- /s5/32Z r- -—';1.zz,$x. "-) - ilASTIZELL CONCRETE REGULAR CONCRETE * sJ = 1976 (Avg. 7 Cy.) r= 3750 psi J = 1.75 x 106 pSi (3v. 2 Cyl;):C = 3 75 x 106 psi n 17.15 n =8 SECTION PROPERTIES *:d = 1.71" kd = 1.27" j =.89 j=.91 I = 103 in4 (Transformed Section) I = 56.47 in4 TOTAB ALLOVWABLE BENDING MOMENT *'Nc = -- - 3,600"# IM nCs= o 3- 0,,- Ms 3270'i: or 39,250"# Ns Controls Values for Regular Concrete taken from Flexicore Tables

ELASTIZELL CONCRETE REGULAR CONCRETE ( continued ) (cont inue d ) TOTAL ALLOWABLE SHEAR (Neo i.eb Reinforcing) V = vcbjd V = vcbjd = (o03) (1976) (3.03)(. 89)(4.87) = (.03)(3750) (3.03)4.44 77C-, = 1515# DEAD LOAD CONDITIONS V:t. of!o'beber = 47.1.,/" t. of T.Member = r (.2;// (!v. of L specimens) 1o"x,,oa- >hear = 6 x 471- 2'.26it Dead Load Shear = 6 x 56.2 Dead;:cA la d'YOit =':? lLj, = 337.2# Dead.d Lod oment =... D. L. V. = C:47'# D. L..i = 10101'# LIVE LOAD CONDITIONS Allowable Live Load YMoment: Total,Allowable Mfoment = 3167 Total Alloable Moment -- 3270 Dead Loa.d M oment = - 847 Dead Load Moment = -1010 A L Loadr aI Al lowable Live Load o1mlent = 2320 M noment = 2260t'i Allowvlrable Live Load Shear: (No.eb Reinforcing) Total Allowable Shear = 778 Total illowable Shear =!515 Dead I,cad Shlear = -23 Dead Load Shear = - 337 A'' b-llci-'AevSea owable ive Load Shear --- 49# Shar r 1178P-'

-10g ELASTIZELL CONCRETE REGULAR CONCRETE (continued) (continued) SHEAR AND MOMIENT DIAGRAMS (LIVE LOAD) ci 2 ALLOWABLE THIRD POINT LOADING Shear: P P 2 2 -?5 P 1178 = 22~~v = 236 - 9 9 9jk PrV 2 356 Moment: (Steol Controls) PsL PsL = 2320 = 2260 6 ~b PS = 1160,- PS 1130# Shear Controls Moment Controls Assume properly designed web re info rQing. Then Allowable Pq = 1160#

ELAST I ZEL CONCRETE REGULAR CONCRETE (continued) (continued) ALLOWABLE UNIFORMLY DISTRIBUTED LOAD w,'"ith l'eb Reinforcing ~,jT -(2)= (-) (LLM) "LL j- ) (LLM)`LL -(._8_)(2320) =(_) (2260) 144 1~44 129 z/, = 125.5 #/' Or Or 96*7 psf 9L.1 psf VLL = 129 x 6 = 774# VLL = 125.5 x 6 = 753#7 VDL = 28 3 VDL = 3 37 Total V 1057i// Total V = 109O/ Allowable V = 77" (Page 9) Allowable V = 115 Stirrups Recquired for V' = 27 9y' No Stirrups Required Stirrup Details: 778 a = (72) =l19 "1 6 = 4-7/8" Stirrups Recuired for 23-7/8" V _ bji1057...s v= v- )T.897TT ) ps=

-12ELASTIZELL CONCRETE REGULAR CONCRETE (continued) (continued) I! 805 =.o041 fc < 0.06 fc Maximum spacing = 2 = 2*44 Say 2-3/8" V's 279 x 2.375 A v fvjd 20000 x.89 x 4.875.00764 sq. in. O1 Av = (.0015)(3.03)(2.375) -.01078 sq. in. Controls --- (ACI 807) Use 11 - #13 wire U stirrups at 2-3/8" centers at each end. CONCLUSION The following conclusions summarize the results of the tests on wtlich this report is based. 1. Elastizell concrete may be used in Flexicore slabs without decreasing the allowable live load of the slabs. For specific spans it may be found that deflection controls the design in which case the allowable live load may be somewhat less when Elastizell concrete is used. 2. The use of Elastizell concrete in Flexicore slabs would reduce the unit weight of these slabs.

3. No changles w:oulcl be required in the manufacturing sequence or in the equipment usecd. It would be necessar- to install a focasi. (ieneration unit.whic.h can1 be operated fromrl the,...'stin: source of compressed air. L40 only -minor chaniLjes need be mjade in fabrication. Such changes would includ(e the increased use of stirrups and possibly the use c w relded wire fabric in place of defor,:ned bars for corprossive reiLforcement. 5)..-J!<lastizell concrete increases in density T,.Lth handling and vibration. Surplus -air miust be added at the m:ixer to allow for this loss of air. SUPPLEMENT Since the preparation of the foregoing report the first commercial application of Elastizell concrete for Flexicore slabs has been completed. It consisted of a roof system for a 3500 sq. ft. residence. The maximum span was 26 ft. with other spans varying from approximately 12 ft. to 18 ft. The slabs were manufactured at the Price Brothers Company's Michigan Flexicore Division. The work included both 6" x 16" and 8" x 16" sections. The previous tests of Flexicore slabs made with Elastizell concrete showed that the concrete would increase in density during handling and vibrating. Based on a study

-14of those test results, the mix for these members was proportioned to give 100 pcf concrete at the mixer, with the expectation that the wet density after placing and vibrating would be 115 pef which would yield a final dry density of 110 pef f. The mix was designed to yield a concrete weighing 110 pef dry, having a compressive strength of 2000 psi at 28 days and a modulus of elasticity of 1.75 x 106 psi. The cement content was 7 sacks per cu. yd. at a density of 110 pef wet. Test cylinders were made at random intervals during the operation and vibrated with the slabs. The test cylinders were weighed, measured and broken at 7, 8 and 9 days after pouring to determine whether the concrete was sufficiently cured to permit shipping the slabsQ The test results are shown in the following table.

-15Air Dry Density Cyl. Date Date Age at at Test Date, Compressive No. Poured Tested Test, Days perf Strength, psi 1 10-9-57 10-18-57 9 109.2 1830 1A 10-9-57 10-18-57 9 113.0 2640 X 10-9-57 10-18-57 9 110.5 2495 XA* 10-9-57 10-18-57 9 113.5 2595 2 10-10-57 10-18-57 8 107.3 1823 2A 10-10-57 1O-lE-57 0-157 8 108.3 2015 3 10-10-57 10-18-57 8 10900 2075 3A 10-10-57 10-18-57 8 107.0 2010 4 10-11-57 10-18-57 7 111.0 2370 4A 10-11-57 10-18-57 7 112.3 2080 These batches had double the usual amount of dispersing agent. The following two production problems, neither of which is serious, were encountered in the manufacture of these slabs. 1. The greater fluidity of the Elastizell concrete requires a tight fitting gate on the bucket used to transport the concrete from mixer to forms. 2. The usual production procedure used in this plant required that the rubber core tubes be removed in about 4 hours for reuse. Since the Elastizell concrete has a somewhat longer setting time it had to be accelerated if it

-16was to conform to the usual production plan. It was found that the set could be sufficiently accelerated by turning off the vapor jets in the kiln for about 1-1/2 hours immediately after placing the slabs in the kiln. However, care must be taken to prevent too rapid drying during this period. After the initial steam curing period the members should be protected from extremes of weather conditions. This is equally true of members made of regular dense concrete. The Elastizell concrete members for this Job were kept rmoist in the yard for about a week. This first commercial application of Elastizell concrete to Flexicore slabs has demonstrated that it-is possible to make lightweight Flexicore slabs on a production basis using Elastizell concrete. The obvious advantages to be gained are 1. A 25% reduction in weight of members, with no reduction in live load capacity. 2. A 45% decrease in thermal conductivity. 3. Manufacture of lightweight structural grade concrete without the use of lighitweight aggregates.

-17APPENDIX The design loads shown on Figures 1, 2, and 3 are (1) Ps = 160# = the sum of 2 equal 1/3 point loads which would cause a calculated stress of 20,000 psi in the positive reinforcement and (2) Pv = 990# = the sum of 2 equal 1/3 point loads which would cause a calculated diagonal tension stress of 59.3 psi in the concrete adjacent to the supports. Figures 1 and 2: Total Load vs. Deflection at the Center Line - These figures show the relationship of total load to measured deflection at mid span. The equation D = L/360 is shown on each curve to indicate the actual loading at which the live load deflection at mid span was equal to 1/360 of the span. Figure 3: Total Load vs. Strain in Concrete at Center Line - This figure shows tihe relationship of total load to measured strain in the concrete at the center of the span. The symbol fc is shown on each curve to indicate the actual loading at which the maximum stress in the concrete was equal to 0O45 of the compressive strength of the concrete.

T_ FT J + 4 44 T- I- 4II I 17, H Iowa 4j" At 4- 44 -41'A-1 F A L.+_4 F++ 4- — H 14 + 7 4+ H+H+ 4- 4 + 44 _LJ _L4 I LIT 4 I 161 H-4-j;-Hi 1 -— ] I i I I I 1+4 4 i-44 -444 — + I L I I JIL j-H P.-I _f+ tj 4+ 4-4 4 -7 IT] 7 + T, I I IT T14fill f+ T I 4 — 4 -T'r-i+h'+4+ H+i Ah4viz_'-' —- 4+4. 44_+ 47 IT All f i l l I f i t P:T. fit fill H —1 7 IL I i;;! L i f T H -H - 4':T: 44-4- - Hi f __FF_ 7 t -,-Ttr — - H —4H+-1'_ -17'_HFT_ -T.4 4 —it I 1H if I I I II -4if 1, TTr _44H- L EJ a it I t T" t-tHi — T-44 -4 -4 T + -F, T +H -4- Z-4t I f t i f f.; I i I 1 + fit ++ 14 H+44 —- I., I 411- 441 Ea

44 4 +-~ ~ ~ ~ ~ ~il + -K ~ 2K 4-i~~~~~~~~~~~~~~41 4-1 t- -1 —i F 1~II- 7- L~ A~~~~~~~~~~~~~~~~~~~~~~1i A - I A4- -TjA4A I-h ~~~ ~~ ~ ~~~~~~4-V A 7F 7 4- V 4 47 A-~~~~~~~~~~~~~~~~~~~~~~~ -444 4- -,~ ~~-t-r'p T F2 ~~~~~~~~~~~~~~~~~~~~~~~~~~~ F —i~~~~~~~tT -t F t4~~~~~~~~~'LF 4 4- rr - ~~~~~~~~~~~.....- 44~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4 F I 4 44 +l l i - l I! tI - 11 i F t t t44I 5 4- A r I- d- - ~j - I 47 I 41I~~~~~~~~~IIi --- - 77vr7 A3-F 7 f- (i i-T- y -~ Vt.0 itfA ~ ~ ~ ~ ~ l-4 -r 4! +i- 11 I- L~Ir11 LII 44-V4 ~-H- ----- b A J- 44-~ ~ A-~ j-4-t f L4- A4 i4 TL L r —4j I I p 4~~~~~~~~~~~~~~~~V -F-T T II- t ~1 F -4 T jrr4Ti~~~~~~~~~~~~~14 L T-iI L ~ A — ~ ~~~ tj* L ~~~~~~~~~~~~~~~~~~~~~~~1~S 4- 41 F-LH

4 - -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4 ~~V$ tzv. Krt~~~~~~~~~~ 41 1 ~~n. I T-4 24 V -4- Ti-~~~-4 — I-, -U - 14-.4~~~~~- }~-+~i~~,4 — 4-. 4- -,~~~~~~~~~~~~~~~~~~~~~~~~~~~4 -4-4 ~ ~ ~ ~.! -+ —+ -i-4 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~. -4 AA 4! 4 -~~~~~~~~~~~~~~~~~~~~_ i 4- ++. _TT~~~~~~~~~4'44 — ~ + + -— K j ~~ 44-44, 4-4444f 1-~,-4~~1~4 4 mI-I — - - 1 ~L4 -4~ ~ ~ ~ ~ ~~ — 47 -4~ ~ ~ ~~~~~~~~~~~~~~~~~4- r4- 4-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~-ratio~I'I-5 4P 22I{''i44-4 4-4-4-! I~4H- I+ fI.1I1 jJ14 - 4

-21-. Test Specimen after