POST-TENSIONED FLAT PLATE TO COLUMN CONNECTION … Journal/1974/May-June... · square) with a single column stub in the center and reinforced with various amounts of bonded reinforcement

POST-TENSIONED FLATPLATE TO COLUMNCONNECTION BEHAVIOR

Stephen W. SmithProject EngineerCook & Huitt, Inc.Engineers-Planners-SurveyorsDallas, Texas

Ned H. BurnsProfessor of Civil EngineeringThe University of Texas at AustinAustin, Texas

Three post-tensioned flat plate specimens (9 ftsquare) with a single column stub in the centerand reinforced with various amounts of bondedreinforcement were loaded to failure. Test data wereobtained on shear capacity, flexural strength,and general behavior in the column connection area.

The results were compared with the current ACIBuilding Code (318-71) design equations andprevious flat plate test results. The supplementarybonded reinforcement enhanced overall behaviorproviding crack control and increasing shear andflexural strength.

This study represents only the first phase of acontinuing research program on post-tensioned flatslabs. The results of the two later nine-panel modeltests will be reported upon in the near future.

74

Stephen W. Smith Ned H. Burns

The behavior of post-tensioned flatplates in the critical column area is ofgreat concern to designers. Unfortu-nately, specific design provisions forshear strength and flexural strength ofunbonded prestressed slabs are not pro-vided in the present ACI Code (ACI318-71).1

In order to provide answers to prac-tical design questions, an extensivepost-tensioned flat plate test program iscurrently being conducted at The Uni-versity of Texas at Austin. The primaryobjective of this test series is to deter-mine the overall behavior of a repre-sentative flat plate structure over thetotal range of loading. Data obtainedcan then be used to determine the re-liability of analysis and design tech-niques presently being used in practice.The entire series will consist of threeisolated column-slab model tests re-ported herein and two nine-panel scalemodels (one-third and one-half) de-signed to simulate practical bay ar-rangements, tendon layouts, and load-ing conditions.

The primary objective of the three

isolated column model tests was to gaina better understanding of the behaviorand strength in the critical column areaand to investigate the influence of vari-ous amounts of supplementary bondedreinforcement placed in the column vi-cinity. Specific objectives of the pro-gram may be summarized as follows:

1. To observe the physical behaviorof each test specimen, including de-flections, crack patterns, crack widths,and mode of failure.

2. To study the effects on perfor-mance and contribution to strength ofsupplementary bonded reinforcementplaced in the critical area around thecolumn.

3. To examine the tendon stress atultimate and check the reliability of theequations provided by ACI 318-71.

4. To compare the flexural strengthwith the corresponding values obtainedby using ACI 318-71.

5. To compare shear strength withthe values predicted by ACI 318-71and some empirical formulas tested byother researchers.

PCI Journal/May-June 1974 75

DESCRIPTION OF TESTPROGRAM AND SPECIMENS

The three specimens tested were iso-lated panels with column stubs repre-senting the region around an interiorcolumn of a multiple panel slab (Fig.1). The models were designed follow-ing the "Tentative Recommendationsfor Prestressed Concrete Flat Plates,"as reported by ACI-ASCE Committee423.2

All three one-third scale slab speci-mens (S-1, S-2, and S-3) were identicalin dimension, tendon layout, and ten-

don profile. The only variable was theamount of supplementary bonded rein-forcement located in the column region.Each specimen was of normal weightconcrete, 9 ft square, 2 3/4 in. thck, withan 8 in, square column stub in the cen-ter of the panel (see Fig. 2).

All specimens were post-tensioned inboth directions with 18 tendons eachway, as shown in Fig. 3a and 3b. Ten-dons were ASTM A-416, Grade 250, ¼in. diameter, 7-wire stress-relievedstrands, which for one-third scale cor-respond to two 1/z in. diameter strandsin the prototype. All strands were mas-tic coated and paper wrapped to pre-

C

ISOLATEC

I

ID

I– I

an

PANEL

SLAB THICKNESSMODEL = 2.75 in.

// PROTOTYPE = 8.25 in.

A

/ LINE OF SYMMETRY

/ MODEL10 10 10' ^2.5

PROTOTYPEI' 30, + 30' 30' .T^ .5j

Fig. 1. Plan view of prototype slab and model slab.

76

M TN T

1_ 1 1' 1 J

I ^

OO 0tt

0)

I

o- - - -o- -[---0-

t_ 4^0°r

9 0

LINE LOAD LINE LOAD

___ 1 :1

I LiFig. 2. Isolated panel.

vent bonding to the concrete. The ten-don spacing corresponds to a 70 per-cent distribution in the column stripand a 30 percent distribution in themiddle strip, with two tendons passingthrough the column in each direction.

The design average concrete pre-stress force, P/A, after losses was to be325 psi, corresponding to a tendonforce of 5.40 kips per strand. To mini-

mize variation in stress due to frictionaleffects, alternate tendons were stressedfrom opposite ends.

Design concrete strength for all spec-imens was 4000 psi, using Type III ce-ment and 3/s in. maximum size aggre-gate. Actual f', values are given in Ta-ble 1. The column stub for all speci-mens was reinforced with four No. 6bars, one in each corner, with plain No.2 ties spaced at 6 in.


P

9 I0" 7 3/g 13 at 4 I/4'= 4'7 I/4 7/g 10^ 9^

IA

L

A

MID. STRIP I COLUMN STRIP (5^-0 11 ) I MID. STRIPplu

9-0

Fig. 3a. Tendon layout.

Supplementary reinforcementAll supplementary, bonded reinforce-

ment was No. 2 deformed bars with across-sectional area of approximately0.05 sq in. and a yield strength of 55.8ksi.

Specimen S-1 contained no supple-mentary bonded reinforcement andserved as a control specimen.

Specimen S-2 contained the mini-mum amount of bonded reinforcementrecommended in Section 2.10.1 of the

ACI-ASCE 423 Recommendations. Thereport recommends that a minimumarea of bonded reinforcement equal to0.15 percent of the cross-sectional areaof the column strip should be providedin the top of the slab in both directionsover the column. To satisfy this require-ment, Specimen S-2 contained five No.2 deformed bars, 3 ft 6 in. long, spacedat 4 in. on center in each direction, asshown in Fig. 4.

In Specimen S-3, bonded reinforce-

78

HIGH DIRECTION (N-S)

38233I2.75

SECTION B - B

LOW DIRECTION (E-W)

275.00 1.95

SECTION A - A

Fig. 3b. Tendon profiles.

ment amounting to 0.24 percent of thecolumn strip area was provided corre-sponding to a 60 percent increase overthe amount in Specimen S-2. Eight No.2 deformed bars were used, with alength of 3 ft 6 in. and spaced at 2 2/s in.on center in each direction, as shown inFig. 5. All No. 2 bars in Specimens S-2and S-3 were placed in between theunbonded tendons at the same depth asthe tendons in both directions.

TESTING

LoadingLoading of the slab was accom-

plished by pulling downward with awhiffle-tree apparatus producing a lineload 4 ft square, as shown in Fig. 2.This loading was chosen to provide asymmetrical load around the columnsimulating loading conditions at a typi-cal interior column. The load line was

located a sufficient distance away fromthe column so it would not directly af-fect the strength or cracking behaviorin the column vicinity and so the col-umn area would remain completely vis-ible for observation during testing.

A separate whiflie-tree was used foreach side of the load line providingfour load points along each line. Eachwhiffle-tree was loaded with a 30-tonhydraulic ram, all of which were con-nected to a common manifold to pro-duce equal loading along all four sides.The loading frame and whiffle-tree ar-rangement are shown in Fig. 6.

InstrumentationPrestress forces in the tendons were

measured by 12 load cells placed at theholding ends of the tendons, as shownin Fig. 7. Six load cells were providedin each direction and were monitoredduring stressing and loading.


Table 1. Summary of test data.

Specimen

S-1 S-2 S-3

Concrete Strength at Time of Testf' (psi) 4350 4193 4625

Supplementary Reinforcement None 5-#2 NW 8-#2 EW

Design Load (Applied Load kips) 13.6 13.6 13.6

1st Cracking Load (kips) 16.3 16.3 19.3

Failure Load (kips) 25.3 27.3 30.3

Moment Capacity

Ultimate Load Based onMoment Capacity (kips)

1'u(test)25.25 27.31 30.41

P(calc) ACI Codeu24.29 27.06 27.49

P (testRatio 1.04 1.01 1.11Pu(calc)

Shear Strength

Ultimate Shear Stress

V

i;-0295 319 350

V()= 4.45 % = 4.93 = 5.16 f,

°(calc) = 4 f` ACI Codeu 264 259 272

°(test)/Vu(calc)U 1.12 1.23 1.29

U°(test) /°u(calc) x 1.32 1.45 1.52

v (calc) ACI Eq. 11-12cw 329 324 332

°u(test)/°cw(calc) 0.90 0.985 1.05

V (test) /Vcw(calc) x 1.05 1.16 1.24

°u(LSM) Lin-Scordelis-May Eq. 302 304 316

°u(test)^ u(LSM) 0.98 1.05 1.11

Vu(GV) Grow-Vanderbilt Eq. 360 360 360

V (test) u(GV) 0.82 0.89 0.98

Strains in the bonded reinforcementwere measured by SR-4 foil strain gageswith a gage length of 1/4 in. The gageswere located for Specimens S-2 and S-3as shown in Figs. 4 and 5.

Concrete strains near the columnwere measured with 0.80 in. paperbacked electrical resistance type strain

gages applied directly to the concretesurface. The locations of these gageswas the same for all three specimensand are shown in Fig. 8.

Deflections of the specimens underload were measured at 18 points, 9points in each direction, by dial gageswith anaccuracy of 0.001 in. The loca-

80

tions of the dial gages are shown in Fig.8, and the support frame for the dialgages is shown in Fig. 9. The support-ing frame for the dial gages was totallyindependent of the test specimens andthe loading system, thus providing afixed reference line for deflection data.

In order to monitor the loading ofthe specimens, a 100-kip load cell wasplaced under the column stub. Loadcells were also located under two of theloading rams and at two load points onthe top of the specimen. When loadwas applied by the hydraulic rams, allload cells were monitored to check theefficiency of the loading system in dis-tributing the desired line loading.

Test procedureThe applied load was primarily regu-

lated by reading the 100-kip load cellunder the column with a calibratedpressure gage on the pump used as acheck. Load was applied in 2000-lb in-crements up to the first cracking loadfor all three specimens and then un-loaded in 4000-lb increments. Speci-men S-1 was then loaded in 4000-lbincrements to failure and Specimens S-2 and S-3 were loaded in increments of2000 and 1000 lb to failure.

At every loading stage all instrumen-tation data were recorded. All crackswere marked and when cracks began toenlarge significantly, crack widths atseveral locations were measured with asmall hand microscope. No additionalload was applied after failure had oc-curred.

RESULTS

GeneralAll three specimens behaved almost

identically in the elastic range up tofirst cracking loads recovering practi-cally all deflection when unloaded, asshown in the deflection curves in Fig.10. As the specimens were reloaded

4N

IL[FORMED BARS.8 ksi

(3' - 6' LONG)4ot4"

Fig. 4. Bonded reinforcement layoutand strain gage location for Specimen

S-2.

7 a^

Fig. 5. Bonded reinforcement layoutand strain gage location for Specimen

S-3.

Fig. 6. Test frame with specimen inposition.

4-N-

-M

ti

JO.2 DEFORMEDBARS

Fy = 55.8 ksi

(3'-fi"LONG)


Fig. 7. Load cells at holding end oftendons.

past the first cracking load, the extrastiffness and ductility provided by thesupplementary bonded reinforcementin Specimens S-2 and S-3 began to benoticeable when compared with Speci-men S-1.

The cracking patterns and failure

shear perimeters are shown in Figs. 11and 12. All specimens failed in a combi-nation of flexure and shear with a finalfailure mode of "punching shear." In alltests, the shear perimeter that appearedat failure outlined the base of a truncat-ed pyramid. The surface of failure ex-

0

-N1-O

LOADING LINE

I o I1 11 p ,(^ CONCRETE GAGE

i o /

0 0 o 0 o to o 0

i IO 112' I

I O 1

®,-DEFLECTION DIAL GAGE

215 09" 10^ 14" 9" 15" 16"

Fig. 8. Location of dial gages and concrete gages.

IVj

I0

a)

N

(N

0)

82

Fig. 9 . Dial gage ar- rangement (Specimen

S-2 after failure).

tended conically from the perimeter of the column at the bottom surface of the slab to the perimeter on the top surface of the slab, shown as a dotted line in the above-referenced figures.

After failure, all three specimens remained supported by the four tendons passing through the column and aided

by the bonded reinforcement in Speci- mens S-2 and S-3. Each specimen remained intact after failure and did not completely collapse.

The first cracking loads and failure loads are given in Table 1, as compared to the design service load of 13.6 kips. All specimens exhibited similar cracking

STATION

DEFLECTION, in.

Fig. 10. Load-deflection curve in east-west direction for specimens.

PC1 Journal/May- June 1974 83

behavior throughout the test. Firstcracks occurred at the column cornersand at the column-slab interface.

In Specimen S-1, very large flexuralcracks opened suddenly under an ap-plied load of 21 kips, propagatingacross the entire width of the slab inthe north-south direction and greatlyreducing the stiffness of the slab.

Specimens S-2 and S-3 exhibited sim-ilar behavior, except that when thelarge flexural cracks opened at 20 and21 kips, respectively, the decrease instiffness was not as pronounced. Also,the overall crack distribution was muchimproved in Specimens S-2 and S-3.

Ductility improved with the increaseof bonded reinforcement as both Speci-mens S-2 and S-3 maintained appliedload even when severely cracked.

Specimen S-1 did not behave wellafter severe cracking, and was unable tosustain the applied load without severedeformation, and failed soon afterward.

Crack width measurements showedthat for a given loading, crack widthsfor Specimen S-1 were higher than foreither Specimens S-2 or S-3.

Tendon stressLoad cell measurements located on

the tendons indicated only very small

k21 -N-

IL/ LOADING LINE

20r 2121

19 9 21

18 D IB_ __ 117

15 C 13

\\ 20 13 E21

\ 18 14 183

19 18 18 Ir

A 'I 2017 15 ----..-

B 218 21

t 19 212

I

20 - • -- ° ---•--I

21

Fig. Ila. Cracking pattern for Specimen S-1.

84

I20 -N-

2320 {

22 LOADING LINE

•20

t 22

T 2 ^2 17 2

e-

I ^I la n2Ii 2022 19 5

I 1 2^^ 12 I /

`` I 5 13 ^It 18 q I ` 1

I I eYYY///

I

l

I IB I 18 t

22

20 I

23

2120

Fig. 11b. Cracking pattern for Specimen S-2.

tendon stress increases near the failureload for all three specimens. Through-out the elastic range and past the firstcracking load there was no tendonstress increase observed at the holdingend in any tendon for all three speci-mens. This trend continued until severeflexural cracking and deformation oc-curred at the higher loads near failure.

Measurements for Specimen S-1were last taken at 18 kips applied loadand indicated no stress increases. InSpecimen S-2 the tendon stress increaseat an applied load of 23 kips (last mea-sured load increment) for one interiortendon was 13.82 ksi with the greatest

increases occurring in the east-westtendons due to the smaller effectivedepth.

Specimen S-3 exhibited similar be-havior with little increase in tendonstress until large flexural cracks anddeformation had occurred near the fail-ure load.

Stress increases occurred in all ten-dons with maximum increases of 13.82and 14.16 ksi in the two east-west col-umn tendons. Only small increases wereobserved in all other tendons.

Concrete strainsConcrete strain measurements on the

top and bottom slab surfaces near the


23 26 21

t-N-26

26 24 25 0II

2 i

8 25,.I 3,

23 2125 I

121,Q ^

21:231^ O

27 24 -^22ft 23 25 0I--ro---- - ---2fi

1zJ o _ i

LOADING LINE

24

2222

Fig. llc. Cracking pattern for Specimen S-3.

column indicated elastic behavior of allthree specimens up to first cracking.When the higher loads were reached,erratic measurements were observedfrom the top gages, indicating crackingnear the gages.

Marked strain increases were ob-served from the compression gages onthe bottom surface as the applied loadpassed the cracking load and flexuralcracking became prevalent. Compres-sive strains were quite high for Speci-mens S-2 and S-3, indicating compres-sive stresses near f',. However, nocrushing of the concrete was observedprior to failure.

Supplementary bondedreinforcement

Strains in the supplementary bondedreinforcement in Specimens S-2 and S-3showed that the behavior for both spec-imens was virtually identical. All load-

strain curves for the bars exhibit a lin-ear elastic region up to first crackingwith almost complete recovery on un-loading (see Fig. 13).

In both specimens the bars passingthrough the columns yielded first withall bars reaching yield prior to failureand exhibiting some strain hardening.In the north-south direction only thecolumn bars reached yield prior to fail-ure. The east-west bars would be ex-pected to yield first since they werelocated under the north-south bars andthus had a smaller effective depth.

DISCUSSION

GeneralDeflection characteristics for all three

specimens were virtually identical forthe range of loading up to 18 kips ap-

86

plied load. The deflection readingshave little quantitative significance forcontinuous slabs, since no edge restraintwas present that could provide thestiffness due to continuity.

However, the deflections do showthe elastic recovery of the specimens inthe early load stages and exhibit thereduced stiffness due to cracking athigher applied loads. They also showthe increase in ductility and stiffnessdue to the supplementary bonded rein-forcement in Specimens S-2 and S-3over the behavior of Specimen S-1 atthe higher applied loads near failure.

Although the first cracking load wastaken at the first appearance of hairlinecracks occurring at the slab-column in-terface and at the column corners, thesecracks have little practical significance,since they were localized at points ofpeak moment and were due largely tostress concentrations. Significant crack-ing did not occur in any specimen untilapplied loads of 16 to 18 kips werereached.

Crack distribution and the number ofcracks observed varied considerably be-tween each specimen. Up until failure,relatively few cracks appeared in Spec-imen S-1. As the load was increased,the first cracks simply propagated andwidened instead of new cracks forming.This is typical of structures reinforcedwith unbonded tendons only, as thetendons tend to slip causing large con-crete strains to be accumulated at afew locations.

However, for Specimens S-2 and S-3,the crack distribution was good with agreater number of cracks observed andsmaller crack widths due to the effectsof the supplementary reinforcement(see crack patterns in Fig. 11a, lib, andlie).

Tendon stressMaximum stress increase was ob-

served in the east-west tendons passingthrough the column for all three speci-mens. This was due primarily because

the east-west tendons had a smaller ef-fective depth (see Fig. 3b) than thenorth-south tendons. The maximum in-creases occurred in Specimen S-3 withSpecimen S-2 tendon stresses very nearthe same, Specimen S-1 indicated littleor no increase prior to failure.

The maximum stress increases ob-served at the holding end were only 59percent of that predicted by the ACIEq. (18-4). However, this type testspecimen, due to its lack of continuityand actual plate behavior, may not yieldquantitatively significant results.

Bonded reinforcementThe presence of bonded reinforce-

ment in both Specimens S-2 and S-3greatly improved their performance un-der load. The crack width reductionsand greatly increased number of cracksobserved for Specimens S-2 and S-3were significant improvements over ob-served behavior of Specimen S-1.Strain measurements on the reinforcingbars proved their effectiveness, as allthe east-west bars in both SpecimensS-2 and S-3 reached the yield point.

As the amount of bonded reinforce-ment was increased, the ductility andload-carrying capacity of the specimenwas increased. The bonded reinforce-ment also helped to hold the specimenintact after failure.

Flexural strength

ACI 318-71 specifies in Section18.8.3 that the total amount of pre-stressed and nonprestressed reinforce-ment shall be adequate to develop adesign load in flexure at least 1.2 timesthe cracking load as calculated on thebasis of a modulus of rupture, f,. =7.5 Vf'0.

Since this requirement refers to acracking moment based on beam ac-tion, there is some question as to itsvalidity when applied to two-way pre-stressed flat plates. For the designloads used, 1.2 times the cracking


Fig. 12. Specimen S-2after failure.

moment is substantially in excess of thedesign ultimate moment.

Cracking loads from the test,Pcr(test), were all found to be belowthe cracking loads based on ACI 318-71, Per(calc), with Pcr(test)/Pcr(catc)ratios ranging from 0.82 for Speci-men S-1 to 0.95 for Specimen S-3.This discrepancy is due to the factthat first cracking was taken at theappearance of small hairline cracks atthe column corners, as described ear-lier. The applied loads, when signifi-cant cracking occurred, would haveyielded ratios exceeding unity.

Table 1 shows a comparison ofactual ultimate loads, Pu(test), and theultimate loads, based on thepredicted ultimate moment capacityaccording to ACI 318-71. The pre-dicted ultimate moment capacity wascalculated assuming that the tendonstress would increase according to Eq.(18-4) in the Code.

According to the values shown inTable 1, Pu(t,st) was higher thanPu(ea zc) for all three specimens andthey all should have failed in flexure.The large flexural cracks and deforma-tion of each specimen near the failureload indicate that flexural failure wasimminent and probably contributed tothe punching shear failure.

Since the predicted values assumedhigher tendon stress increases at ulti-mate than were actually observed, thePu (test)I pu ( eaac) ratios should actuallyke even larger if based on the actualtendon stress at failure.

Although the tendons did not con-tribute the full tensile force predictedby ACI 318-71, the extra moment ca-pacity can be attributed to the extraflexural stiffness provided by the two-way deformation of the slab. Also, inSpecimens S-2 and S-3, the strain read-ings for the bonded reinforcement indi-cated that most bars, especially thosepassing through the column, had yield-ed and had begun to strain harden,thus contributing a higher tensile forcethan the yield strength used for the pre-dicted ultimate moment values.

Shear strengthThe shear strength of post-tensioned

flat plates is governed by two-way ac-tion with potential diagonal crackingalong the surface of a truncated pyra-mid around the column, as defined inSection 11.10.2 of ACI 318-71. Allthree specimens failed in a combinationof flexure and shear, with the actualfailure mode occurring as punchingshear.

Shear strength design methods and

88

No. 3N o. 2

No.l

((11,000IL - in.

• 1,000µ - in. l STRAIN, fc - in/in

Fig. 13. Load-strain curve for reinforcing bars in east-west direction (SpecimenS-3).

30

25

20

015

UiwJ

I0

requirements for prestressed flat platesare not defined in ACI 318-71 or localbuilding codes. Section 11.10 of theCode allows an ultimate shear stressof 4 Vf' C used in conjunction with thecb factor for shear of 0.85.

This allowable shear stress is to beused for two-way slabs at a critical arealocated a distance d/2 from the faceof the support. However, no differentia-tion is made between prestressed andnormal reinforced concrete construc-tion.

Another ACI 318-71 equation hasbeen recommended by ACI-ASCECommittee 423. 2 The same critical sec-tion is used and the nominal ultimateshear stress is calculated by Eq. (11 -12)of ACI 318-71:

VPv. = 3.5 i./f', + 0.3f f, + bwd

The last term contributes only a verysmall portion to the shear strength and

is usually conservatively neglected.Although this equation takes into con-sideration the prestress force, it hasthe disadvantage of being based onbeam action instead of two-way platebehavior.

Other empirical data and shearstrength equations have been pub-lished by Lin, Scordelis, and May,3Grow and Vanderbilt, 4 and Gerber andBurns. 5 All of these test programsinvolved the testing of specimens verysimilar to Specimens S-1, S-2, and S -3reported herein and therefore signifi-cant comparisons can be made be-tween them. The shear strength ofSpecimens S-1, S-2, and S -3 are com-pared with the two ACI 318-71 pre-dicted values and the empirical valuesin Table 1.

The v.(test)/vu(calc) ratios are greaterthan unity and therefore conservativeusing the ACI 318-71 provisions for


two-way slabs. However, ACI CodeEq. (11-12) provides ratios for Speci-mens S-1 and S-2 which are less than1, while the Specimen S-3 ratio was1.05, showing very close correlation tothis equation. When the 4) factor of0.85 is introduced, both ACI equationsyield conservative ratios above unity.

Comparison with the empirical equa-tions based on prestressed slab be-havior indicates unconservative ratiosfor v, (test)/vu(CQie). The Lin-Scordelis-May equation, "Vu(test)/vu(LSM), pre-dicts the behavior of Specimens S-1and S-2 quite closely. As the amountof bonded reinforcement increases theratios increase with the most conserva-tive ratio for Specimen S-3. All ratiosusing the Grow-Vanderbilt equation,Vu(test) /vu(GV), are less than 1.0 andunconservative, but do indicate thesame upward trend as the amount ofbonded reinforcement is increased.

The low Vu(test)/Vu(eaie) ratios canbe attributed to the severe flexuralcracking present at loads near failure.After the applied load reached thetheoretical ultimate moment capacityof the specimens, severe flexural dis-tress was observed in the form of wide-spread cracking in the critical areaaround the column. This cracking re-duced the shear strength of the speci-mens in this region, resulting in earlyshear failure.

The higher ratios exhibited by Speci-mens S-2 and S-3 can be attributed tothe presence of the bonded reinforce-ment which increased the moment ca-pacity and was effective in distributingcracks and controlling crack width inthe column area.

CONCLUSIONS

From the results of this test programthe following conclusions can bedrawn:

1. In the elastic range the behavior

observed for all three specimens was al-most identical. All specimens showedelastic response up to the first crackingload, unloading, and reloading to thepoint of first cracking.

2. The first cracking load has littlepractical significance since first crackingis localized at points of peak momentand is due largely to stress concentra-tions. The slab is capable of sustainingmuch larger loads before severe crack-ing occurs.

3. The bonded reinforcement im-proved the flexural behavior, improvedthe crack distribution, reduced themaximum crack widths, and increasedthe load carrying capacity of SpecimensS-2 and S-3. As the amount of bondedreinforcement increased, the behaviorof the specimens improved.

4. The stress in the unbonded ten-dons remained virtually constant untilsevere flexural cracking and deforma-tion had occurred near the failure load.The ACI 318-71 equation for estimat-ing tendon stress at ultimate, Eq. (18-4), proved unconservative for all threespecimens.

5. All three specimens failed bypunching shear, which followed the de-velopment of severe flexural cracks.

6. The observed flexural strengthwas greater than the strength predictedby ACI 318-71 for all three specimens.The flexural capacity was improved bybending in two-way action, while theACI Code equation is based on beamtheory. The bonded reinforcement in-creased the capacity of Specimen S-2by 8 percent and Specimen S-3 by 20percent compared to the companionSpecimen S-1 with no bonded rein-forcement.

7. Shear strengths from these testspecimens (4.45 ^/f' 4.93 -/ and5.16 i/f', for Specimens S-1, S-2, andS-3, respectively) were slightly abovethe values predicted by Eq. (11-12)based on ACI-ASCE Committee 423

P

90

3.ACKNOWLEDGMENT

recommendations. The nominal per-missible shear strength based on two-way action according to Section 11.10of ACI 318-71 was conservative forall three specimens. The shear strengthfrom these test specimens gave resultssimilar to those results predicted fromthe empirical equation by Lin-Scordel-is-May. The equation developed byGrow and Vanderbilt was unconserva-tive for all three specimens althoughthe shear strength of Specimen S-3was very closely estimated.

This study is the first phase of a con-tinuing study on post-tensioned flatslabs (two later nine-panel models weretested) at The University of Texas atAustin sponsored by the Post-Tension-ing Division of the Prestressed Con-crete Institute and the Reinforced Con-crete Research Council.

Theses by Stephen W. Smith andChaiwat Changwatchai in partial ful-fillment of their MS degrees served asthe basis for this paper. The work wassupervised by Dr. Ned Burns and anoral presentation by Dr. Burns to thePCI at its Annual Convention in Atlan-ta, Georgia, October 11, 1972, coveredpreliminary results from this study.

REFERENCES

1. ACI Committee 318, "BuildingCode Requirements for ReinforcedConcrete (ACI 318-71)," AmericanConcrete Institute, Detroit, 1971, 78pp.

2. ACI-ASCE Committee 423, "Tenta-tive Recommendations for Pre-stressed Concrete Flat Plates," ACIJournal, Proceedings Vol. 67, No. 2,February 1974, pp. 61-71.

Lin, T. Y., Scordelis, A. C., andMay, H. R., "Shearing Strength ofPrestressed Concrete Lift Slabs,"Document Section, State of Califor-nia, Sacramento, October 1957. Al-so, summarized in "ShearingStrength of Prestressed Lift Slabs,"ACI Journal, Proceedings Vol. 55,No. 4, October 1958, pp. 485-506.

4. Grow, J. B., and Vanderbilt, M. D.,"Shear Strength of PrestressedLightweight Aggregate ConcreteFlat Plates,' PCI JOURNAL, Vol.12, No. 4, August 1967, pp. 18-28.

5. Gerber, Loris L., and Bums,Ned H., "Ultimate Strength Tests ofPost-Tensioned Flat Plates," PCIJOURNAL, Vol. 16, No. 6, Novem-ber-December 1971, pp. 40-58.

Discussion of this paper in invited.Please forward your discussion to PCI Headquartersby October 1, 1974, to permit publication in theNovember-December 1974 PCI JOURNAL.


Documents

POST-TENSIONED FLAT PLATE TO COLUMN CONNECTION … Journal/1974/May-June... · square) with a single column stub in the center and reinforced with various amounts of bonded reinforcement