4e Confrence spcialise en gnie des structures de la Socit canadienne de gnie civil 4th Structural Specialty Conference of the Canadian

Society for Civil Engineering Montral, Qubec, Canada 5-8 juin 2002 / June 5-8, 2002

ANALYTICAL STUDY OF THE EFFECTS OF TENDON LAYOUT ON THE PERFORMANCE OF POST-TENSIONED TWO-WAY SLAB SYSTEMS
A.J. Schokker, S.C. Lee, and A. Scanlon Department of Civil and Environmental Engineering, The Pennsylvania State University, United States ABSTRACT: This paper presents the results of an analytical study based on finite element analysis to evaluate the effects of different arrangements of tendons on the performance of post-tensioned cast-in-place concrete two-way slab systems. Various arrangements from uniform cable layouts to closely banded arrangements adjacent to column centerlines are considered. The slab is modeled using plate-bending elements in the software package SAP 2000. Equivalent loads based on cable profiles are applied to the slab according to the tendon layout. Moment fields and deflected shapes are compared for the various tendon patterns. Gravity loads are applied in the usual way and superimposed on the results from the equivalent tendon loads to determine net moments and deflections at service load levels. In addition to a study of tendon layouts, the finite-element model is compared with experimental results. 1. INTRODUCTION Post-tensioned flat plate slabs provide an economical alternative to traditional reinforced concrete slabs due to their relatively reduced thickness. The load-balancing technique is typically used in the design of these sections. A portion of the service load is balanced by the prestressing; typically all of the dead load and sometimes a portion of the live load will be balanced. Various tendon arrangements may be utilized to provide this balancing. The two extremes of tendon spacing vary from uniformly spaced to a fully banded arrangement where tendons are concentrated along column lines. A slab containing uniform tendons in one direction and banded tendons in the perpendicular direction would behave similarly to a one-way slab system supported on beams. A pattern between the two extremes is often used, providing tendons across the entire slab with a concentration of tendons nearer to the column lines. There does not appear to have been a systematic analytical study of the effects of various tendon arrangements on the distribution of moments and deflections in two-way post-tensioned slabs. This paper presents an analytical study of the effects of tendon layouts including comparisons with experimental results.

2. TEST PROGRAM FOR TENDON LAYOUT STUDY 1.1. Analysis

The structural analysis software, SAP 2000, was used for finite element analysis of the slabs. The prestressing force in the slabs was modeled using the load-balancing concept as described in the PostTensioning Manual published by the Post-Tensioning Institute (2001). For the tendon layout study, a typical interior panel of a two-way flat plate system with a thickness of 165 mm is modeled. Dimensions and boundary conditions of the panel are shown in Figure 1.

=0 y

Columns 360mm x 510mm 6.1 m Thickness = 165mm Y

=0 x

7.6 m

=0
x = y

=0

Figure 1. Slab panel dimensions and boundary conditions.

The slab panels were analyzed with their self-weight and equivalent prestressing load from the loadbalancing method. The two load cases were run separately and then superimposed to obtain net moment and deflection values. The tendon force for each direction is kept at a consistent value to produce a uniform compression in the slab of approximately 175 psi. The average equivalent load is approximately 1.25 times the self-weight in the long direction and approximately 0.86 times the self-weight in the short direction. The average equivalent load is 1.06 times the self-weight. 1.2. Tendon Layouts

Four tendon layouts were considered in the study. Layouts fully uniform in both directions and fully banded in both directions were considered as extremes. A panel with one direction banded and one direction uniform was considered as a case approaching one-way system behavior. A layout of 75% tendons within the column strip in both directions with the remaining 25% in the middle sections was also included in the study. The four tendon layouts are summarized in Table 1 and illustrated in Figure 2. Case 100C-100C is essentially unreinforced over most of the slab and is included to demonstrate the effect of the extreme case producing line loads in each direction. Case Uni-Uni provides the opposite extreme. Case 100C-Uni and 75C-75C are more traditional layouts.

Table 1. Tendon Layouts Considered. Case 100C-100C 75C-75C 100C-Uni Uni-Uni Tendon Layout 100% of tendons banded along the column line in each direction 75% of tendons within the column strip and 25% within the middle strip in each direction 100% of tendons banded along the column line in one direction and uniformly distributed across panel in perpendicular direction Uniform distribution of tendons in each direction

tendons

100%C-100%C

75%C-75%C

Uni-100%C Figure 2. Tendon layouts. 1.3. Results

Uni-Uni

Deflection results for each case are summarized in Figure 3. Results from prestress forces alone, selfweight alone, and the combined effect are shown. As shown in the figure legend, Point A represents the deflection at mid-span along the column line in the long direction, point B represents the deflection at midspan along the column line in the short direction, and point C is the mid-panel deflection. The most effect layout for producing upward camber is the 100C-100C case, while the least effective is the Uni-Uni case. The least variation in the deflection values is seen in the mid panel values with a maximum difference of only 0.7 mm. Larger variations are seen in the deflections at the mid-span along the column lines. A slight upward camber is produced at each of the three points in the 100C-100C case and a slight downward deflection is produced in the Uni-Uni case.

Pt.C
5
Prestress

Pt. B

Pt. A

Pt. C
Deflection (mm) 3

Pt.A
1

Pt. B Pt. A Pt. B

Pt. C

-1

Pt. B
Combination

-3

Pt. A Pt. C
-5
100C-100C 100C-Uni 75C-75C Uni-Uni

Figure 3. Deflection comparison for different tendon layouts. Figures 4 and 5 show the effect of tendon layout on bending moment distribution. Bending moments are plotted for each case along the column face line in the long direction. The largest moment intensities due to prestressing are seen in the 100C-100C case near the columns. When these positive prestress moments are combined with the negative self-weight moments, a positive moment results at the column. In all other cases, the combined effect produces a negative moment near the columns. Overall, the least effective layout for balancing the self-weight with the prestress is the Uni-Uni case. Moments in the short direction are similar to those shown for the long direction, except in the 100C-Uni case which shows behavior similar to the 100C-100C case in the short direction.

100C-100C 25 20 15 Moment (kN-m) 10 5 0 -4 -3 -2 -1 -5 0 -10 -15 -20 -25 Global X-Direction (m) Prestress Self-Weight Combined 1 2 3 4

Figure 4. Bending moment distribution long direction, 100C-100C layout

75C-75C 25 20 15 Moment (kN-m) 10 5 0 -4 -3 -2 -1 -5 0 -10 -15 -20 -25 Global X-Direction (m) Prestress Self-Weight Combined 1 2 3 4

Figure 5. Bending moment distribution long direction, 75C-75C layout

100C-Uni 25 20 15 Moment (kN-m) 10 5 0 -4 -3 -2 -1 -5 0 -10 -15 -20 -25 Global X-Direction (m) Prestress Self-Weight Combined 1 2 3 4

Figure 6. Bending moment distribution long direction, 100C-Uni layout

Uni-Uni 25 20 15 Moment (kN-m) 10 5 0 -4 -3 -2 -1 -5 0 -10 -15 -20 -25 Global X-Direction (m) Prestress Self-Weight Combined 1 2 3 4

Figure 7. Bending moment distribution long direction, Uni-Uni layout 3. COMPARISON WITH EXPERIMENTAL RESULTS The analysis procedure used in the tendon layout study was also compared to experimental test results by Burns and Hemakom at the University of Texas (Hemakom, 1975; Burns and Hemakom, 1985). Figure 12 the tendon layout and points where deflections were monitored during testing. Figure 13 shows the experimental test slab dimensions and layout. The finite element model included the slab stubs and roller supports at the base of the columns that was used in the experimental test. A comparison of the deflections between the test slab and the modeled slab is shown in Figure 14. Deflection measurements were not reported for Panels A and B. Results show good agreement for most of the slab points as summarized in Table 2. The maximum difference of less than 0.6 mm is seen at point 7 located in the middle of the center panel.

10

11

12

13

14

15

16

17

Figure 12. Deflection monitoring locations

100 mm

3m

3m

0.8 m

3m

100 mm 0.8 m 3m 3m 3m

70 mm 25 mm ball bearing 536 mm

Figure 13. Experimental test slab dimensions Table 2. Summary of experimental and analytical results Point 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Measured (mm) -1.68 -2.01 -1.02 -1.17 -1.52 -0.94 -1.09 -1.12 -1.78 -0.81 -1.35 -1.60 -1.32 -1.55 -1.63 -2.31 -1.30 SAP2000 (mm) -1.40 -2.06 -1.02 -1.07 -1.83 -1.17 -1.68 -1.22 -2.03 -0.76 -1.25 -1.70 -1.37 -1.68 -1.42 -1.91 -1.09 Average SAP2000/Measured 1.20 0.98 1.00 1.10 0.83 0.80 0.65 0.92 0.88 1.07 1.08 0.94 0.96 0.92 1.14 1.21 1.19 1.02

Panel C

Panels D, E, F

Panels G, H, I

Panel C Deflection
0

Test 203 (Measured)

SAP2000

Column

Midspan

Column

Panel DEF Deflection

Test 203 (Measured)

SAP2000

10

Column Panel D

Column Panel E

Column Panel F

Column

Panel GHI Deflection

Test 203 (Measured)

SAP2000

15

16

17

Column Panel G

Column Panel H

Column Panel I

Column

Figure 14. Experimental versus analytical deflection comparison

4. CONCLUSIONS This analytical study of tendon layouts demonstrated the following in uncracked slabs: ! A finite element model with prestressing forces applied as loads using the load-balancing technique can accurately predict deflection behavior. ! Mid panel deflection values are relatively insensitive to tendon layout for the cases considered. ! Moment distributions across the slab panels may vary significantly between different tendon layouts, sometimes resulting in different signs at critical sections. ! Both of the banded tendon arrangements, 100C-100C and 100C-Uni, produced a prestress moment at mid panel with the same sign as the self-weight moment resulting in a larger combined moment at mid panel. Further studies will be performed to investigate behavior at ultimate. 5. REFERENCES Post-Tensioning Institute (2001) Guide Specification for Grouting of Post-Tensioned Structures, First Edition, PTI Committee on Grouting Specifications, Post-Tensioning Institute, Phoenix, Arizona, USA. Hemakon, R. (1975) Strength and Behavior of Post-Tensioned Flat Plates with Unbonded Tendons, Ph.D. Dissertation, the University of Texas, Austin, Texas, USA. Burns, N.H. and Hemakom, R. (1985) Tests of Post-Tensioned Flat Plate with Banded Tendon, ASCE Structural Journal, 3:1899-1915.