ENGINEERING RESEARCH INSTITUTE THE UNIVERSITY OF MICHIGAN ANN ARBOR Technical Report ELASTIZELL CONCRETE OVER MILCOR CELLUFLOR PANEL TYPE BB 16-16 Alex E. Mansour, Jro Project 2326 ELASTIZELL CORPORATION OF AMERICA ALPENA, MICHIGAN May 1957

The University of Michigan * Engineering Research Institute INTRODUCTION The Milcor Celluflor Panel Type BB 16-16 is one of a series of corrugated structural steel sections produced by the Inland Steel Products Company. One of the design economies of these sections is that they decrease the dead load of the structures in which they are used. In the use of Milcor Celluflor Panels, concrete is normally placed over the panel to provide a plane floor surface, to distribute concentrated loads, to provide fire protection, to cover miscellaneous electrical, plumbing, and heating services, and to stiffen the floor section. Elastizell concrete is a cellular concrete which can provide structural strengths at densities less than the density of ordinary concrete. Since both Elastizell and Celluflor are designed for lightweight structural use, a successful combination of the two products should greatly increase the potential of both products. With this thought in mind, the Elastize1 Corporation of America and the Inland Steel Products Company jointly sponsored the tests which are summarized in this report. The project was under the direction of Professor L. M, Legatski of The University of Michigan. SUMMARY The tests included four sections of Panel, Type BB 16-16, three of which were covered with Elastizell concrete and one with regular dense concrete. The results of these tests indicate that Elastizell concrete may be substituted for ordinary concrete over Milcor Celluflor corrugated panels in building construction with the advantage of carrying the same allowable live load as may be carried when ordinary concrete is used and simultaneously greatly decreasing the dead loado TEST SPEC IMENS The four test sections were prepared March 10, 1956 at The University of Michigan, Willow Run Laboratories. Each specimen consisted of one type BB 1616 panel covered with concrete to a depth of 2-1/2 inches over the corrugations.

The University of Michigan ~ Engineering Research Institute The Elastizell concrete was mixed in a 17-cubic-foot Elasticrete Mixer yielding the following properties:.,. ~., _ Material 28-Day Section Dry Density P bnCompressive E Sand Water Elastizell Elastimulse (f') psi J-1-2 94 2.3 o.45 0.005 0.02 898 J-24 104 2.3 o045 0.003 0.02 1912 1.57x106 KX1-2 95 2.55 0.45 0.005 0.02 1218 o.875X106 1-1 fyield 5000 to 5500 psi at 28 days and was supplied by the Ann Arbor Construction Nompany Its average 28-day compressive strength (ftC) was 4576 psi and its modulus of elasticity (Ec) was 3.31 x 106 psi. TEST PROCEDURES After placing and finishing the concrete, the test sections were mo.st-cured for six days and air-dried for thirty-six days. Standard 6 by 12 in. cylinders were taken of the concretes and were moist-cured for twenty-four days and dried to constant weight. The cylinders were used to determine the o of elasticity and the 28-day compressive strength of each concrete. The test sections were placed on simple supports at 9 ft-6 in. centers and were loaded by approximately equal concentrated linear loads at the 1/3 points. The following measurements were nade at each increment of load. l, Deflection at the center line 20 Strain in steel at center lineo 35 Strain inin concrete at center liner ~os.t~urd orsi dysan ar-rid orthrt-sx ay. tadad b 2

The University of Michigan * Engineering Research Institute 4. Horizontal movement of the concrete relative to the steel at the end of the section. The following photographs illustrate the method of loading and the location and types of gages. Photo - 1 This photograph illustrates the use of hydraulic jacks at the 1/3 points over a specimen in the loading frame. The gage at the extreme right is measuring the horizontal movement of the concrete relative to the steel. The loads were measured through the use of dynamometers. Photo - 2 This photograph illustrates in greater detail the measurement of relative movement of the concrete with respect to the steel. Photo - 3 This photograph shows the deflection gage at the center line of the span and also the SR-4 Strain Gage on the concrete at the center line. Another SR-4 Strain Gage is directly under the visible gage and was used to measure the strain in the steel. Ph oto- 1

The University of Michigan * Engineering Research Institute.-... F. i;........... i.E -E.LEE-....... 2. 2........ _EO ii~~~~~~~~~~~~klb_ Mi~~~~~~~~~~i 1~~~~~~~~~~~~~~~~~~~l I": _ ~ ~ ~ ~~~ ~~~~~ ~ ~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~....._'e.,, _;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...... Photo 2 l l l l l _~~Pht

The University of Michigan * Engineering Research Institute TEST RESULTS Each specimen was progressively loaded to failure with results as illustrated on Figs. 1 to 4 and in the Appendix. SPECIMEN J-1-2 The first crack was a horizontal crack noticed just above the steel at the slab edge at a total load of approximately 3900 lb. These horizontal cracks were visible for approximately 6 in. from the ends of the specimen. A complete bond failure occurred at a total load of approximately 7500 lb. Bond failure was visible from the supports to the load points. Diagonal cracks appeared outside the load points and tension cracks (vertical) appeared near the center of the span. The concrete did not crush until the load was at the maximum. SPECIMEN K-1-2 The first horizontal crack occurred at the north end of the specimen when the total load was approximately 3800 lb and at the south end at a total load of approximately 4200 lb. Vertical cracks also appeared in the concrete at this loading at a distance of 15 in. from the north end and 10 in. from the south end. Flexural cracks (vertical) appeared in the concrete at the center of the span under a total load of approximately 5300 lb. Complete bond failure occurred after about one minute under a total load of 9633 lb. Vertical cracks appeared about 8 in. outside the bearings, but no flexural cracks were visible in the center third of the specimen. The compression flange of the steel was buckling in the center third. SPECIMEN J-2-4 The first horizontal crackoccurredin bond for about 12 in. at the north end under a total load of approximately 6100 lb. A flexural crack (vertical) also appeared at the center line. The specimen failed in bond with no evidence of failure in the steel at a total of 11,733 lb.

The University of Michigan ~ Engineering Research Institute REGULAR CONCRETE This specimen also failed in bond at a total load of 18,153 lb. Cracking, although not visible, could be heard at a total load of approximately 6000 lb. DISCUSSION OF TEST RESULTS The results of these tests indicate that Elastizell concrete may be substituted for regular concrete over Milcor Celluflor corrugated panels with the advantage of carrying the same or greater live load and simultaneously reducing the dead load. The factor of safety for each specimen is as shown below and is defined as the ratio of the ultimate load to the design load. Specimen Design Load* Ultimate Load** Safety Factor J-1-2 1752 7505 4.29 J-2-4 1752 11733 6.70 K-1-2 1752 9633 5.50 regular concrete 1752 18153 10.37 *~Design load is the sum of two equal concentrated loads at the 1/3 points of the span which would cause a deflection of 1/360 of the span. **Ultilate load is the sum of two equal concentrated loads at the 1/3 points of the span which caused failure of the test specimen. In the above table the design loads are the same for all specimens be cause the deflection limitation controls the design. The design of Celluflor slabs is based on the assumption that the steel alone is the resisting section. See pages 7, 8, and 9 for design computations. DESIGN LOADINGS CONDITIONS Simple Span 9 ft —6 in, co to co of supports~ Panel Type BB 16-16:

The University of Michigan ~ Engineering Research Institute Wsteel 7o5 lb/sq ft I = 2.42 in. o4/ft width S = 1.45 in.3/ft width ALLOWABLE MOMENTS Uniformly distributed load: (a) For maximum allowable fiber stress, fs = 18000 psi Mt = f5 S = 18000 x 1.45 = 26,100 in.-lb or 2,175 ft-lb (b) For maximum deflection at Q = span x 1/360 d = _ L 9.5 x 12 36o0 360 48EId 48 x 29.5 x 106 x 2.42 x.317 - 5L2 5 x 114 x 114 ~m = 16,667 inlb or 1391 ft-lb Equate Allowable Moments For uniformly distributed load to allowable moments for equal concentrated loads at 1/3 points: Call P total live load in tests of 2 ft wide sections. I is for 1 ft width. Then P/4 = concentrated load for 1 ft width. PL W1 L2 12 8 Where W11 is equivalent uniform live load. P = 1.5 Wll L or W11 = 2P/3L

The University of Michigan ~ Engineering Research Institute CONCRETE VOLUME (See Cross-Section Dimensions in Inland Steel Products Company Catalcg No, 270) Concrete cross-sectional area 24 x 2.5 = 60.0 + (2.125 + 2.375) (1.5)(4) = 13.5 2 Area = 73.5 sq in. Volume = 73.5 x 9.5 x 1/144 = 4.85 ft3 Volume per sq ft = 94.85 =.255 ft3/ft2 9.5 x 2 DEAD LOAD MOMENTS Mdl = Msteel + Mconcrete Mstee =- Ws L = 84.6 ft-lb 8 Wc L2 Mconcrete = 8 Wc Mc Ms Mdi Section lb/ft2 ft-lb ft-lb ft-lb J-1-2 24 271 84.6 355.6 J-2-4 26.5 299 84.6 3$3. K-1-2 24.2 273 84.6 357.6 regular concrete 36.7 414 84.6 498.6 ALLOWABLE TOTAL LIVE LOADS Mll = Mt - Mdl

The University of Michigan * Engineering Research Institute P = 1.5 W11 L See page 7 (a) For fs = 18000 psi: Mt..M11 W1, P Section Mt Mdi ft-lb ft-lb ft-lb lb/ft2 lb J-1-2 2175 355.6 1819.4 161.3 2300 J-2-4 2175 38356 1791.4 159 2265 K-1-2 2175 357.6 1817.4 161 2295 regular concrete 2175 498.6 1676.4 148.7 2120 (b) dt = span x 1/360 Section MQ Wl P ft-lb lb/ft2 lb J-1-2 1391 123.0 1752 J-2-4 1391 123.0 1752 K-1-2 1391 12350 1752 regular concrete 1391 123.0 1752 ALLOWABLE STRAIN CONDITIONS Steel: fs = 18,000 psi Es = 29.5 x 106 psi ea 18 x 103 18 x.00061 in./in. Es 29.5 x 106 Concrete: Specimen f fc = 0.45 f Ec e = fc/Ec lb-sq in. lb-sq in. lb-sq in. in./in. J-1-2 898 404 -- J-2-4 1912 860 1.57 x 106 o00oo548 K-1-2 1218 547 0.875 x 106 o000625 regular concrete 4576 2060 3.31 x 106.000623,.... j,....,,9

The University of Michigan * Engineering Research Institute APPENDIX The design loads shown on Figs. 1, 2, and 3 are (1) P360 = the sum of two equal 1/3 point loads which would cause a calculated deflection of L/360, assuming the concrete to provide no rigidity and (2) Ps = the sum of two equal 1/3 point loads which would cause a stress in the steel of 18000 psi, assuming the concrete to provide no rigidity. Figure 1: Total load versus mid-span deflection. This figure shows the relationship of total load to deflection at the center of the span. The equation D = L/360 is shown on each curve to indicate the actual loading at which the center line deflection was equal to 1/360 of the span. As may be seen for all specimens, the actual loading causing such deflection is greater than the design loading. The deflections of specimen J-2-4 plot as a straight line above a load of about 2000 lb. The curved portion near the origin is probably due to uneven bearing on the supports. The dashed line drawn parallel to the straight part of the curve through the origin should be used in predicting the deflection of this specimen. Figure 2: Total load versus strain in steel at center line. This figure shows the relationship of total load to the strain in the steel section at the center of the span. The symbol Fs is shown on each curve to indicate the actual loading at which the maximum stress in the steel was equal to 18000 psi. As may be seen for all specimens, the actual loading causing such a stress is greater than the design loading. Figure 3: Total load versus strain in concrete at center line. This figure shows the relationship of total load to the 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 10

The University of Michigan * Engineering Research Institute equal to 0.45 of the 28 day compressive strength. Such data are not available for specimen J-1-2. No theoretical loading has been computed to cause Fc = 0.45 f'c since it is assumed in this type of construction that the concrete is not a part of the structural system. Figure 4: Total load versus concrete movement relative to steel. This figure shows the relationship of total load to the horizontal movement of the concrete relative to the steel at the end of the specimen. Only the curve for J-2-4 follows the expected pattern. The instrumentation may have been responsible for these erratic curves. 11

ULT. 11733 ULT 1815 2 2 3'- 2 3' 2 3'-2 10 9'-6 ULT. 9633 < KII' — 2. KI2~~~~~~I ro- I 0 ULT. 7505 24 24REGULAR CO CONCRETE z Z~~~~~~~~~~~~ = r~o I1~~~~~~~. =. 0 6 a. 0 0 a.. 0 _j~p 0 1"'1~ 4) if 0 / II ~ p w~~~~~~~~~~~~~~~~~~~~ <~~~~~"p DESIGN _.: I — P A LOADS'"~' <m o 3 f0 3 /0 P360 P36 0 0 0 0.2 0.4 (J-l.2) 0 0.2 0.4 (J.2.4) 0 0.2 0.4 (K.I.2) 0 0.2 0.4 (R.C.) CENTER LINE DEFLECTION INCHES Fig. 1. Total load versus mid-span def~lection.

p ~~~p ULT. 11733 UT 2 - /- ~~~~~~~~~~~~~~~~~~113.-I GAGE 10 ULT. 9633' LLT~~~~~~~~~~ to K'I'2 0'2 8 0 x ~~~~~~~~~~ULT. REGULAR 7505 ~~~~~~~~~CONCRETE o3 LA" 06 a. 6~~~~~31 0m 0) <44 0 uL~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~-..j~~~~~~~~ / 30 2006 0 600 3KI2 0 200 400 600 (J-24) STRAIN IN STEEL AT CENTER LINE- MICRO INCHES PER INCH Fig. 2. Total load, versus strain in steel at center line.

U LT. 18 15 3 ULT. 11733 - 11 300 I 0~~~~~~~~ J 2-4 ULT. 9633 r') ~~REGULARto ~~~CONCRETE K I-2x ~~~~~~~~~~~~~~~~~~~~~~ULT. 7505 o o0 z L 0 6 a0a. LLP 44 0, GAGE 4 0 360~ DESIGN LOADS 0 ~ 200 400 600 800 1Q00 1200 1400 STRAIN IN CONCRETE AT CENTER LINE - MICRO INCHES PER INCH Fig. 5. Total load versus strain in concrete at center line.

ULT. 18153 ULT. 11733 I0'~' ~f —-xULT. 9633 10 ~~0 ~~zx~~ I -' II 8 _ x -F II/ ULT. 7505 p sl~w CO) Z m. z 0 6 a. x i~~~~~~~~~~~~~~~~~~~~~~~~~~~~ H0.. Q ~ ~ ~ ~ ~~~~ ___ m Z t 0 S~~~~~P 0 2__ X J2.4 I- 2 ~P. o~~~o DE~sIGN LOADS 360 6 + K.I.2 O~ J'l2 ~ / 2-I 00 0 2 4 6 8 10 12 14 CONCRETE MOVEMENT RELATIVE TO STEEL INCHES X I0 Fig. 4. Total load versus concrete movement relative to steel.