Research & Development Information

PCA R&D Serial No. 2622

High Strength Concrete Columns with Intervening Normal Strength One-Way Slab and Beam Floors
by Ehab Abdel Wahab and Scott Alexander

PCA Education Foundation Fellowship

Portland Cement Association 2005 All rights reserved

KEYWORDS
columns, floors, high-strength concrete, reinforced concrete

ABSTRACT
This paper presents an experimental study of seven 2 / 3-scale specimens tested to investigate the behavior of joints between normal strength concrete floors intersecting high strength concrete columns. Two different loading scenarios were examined. Two new construction techniques were presented and their effects on the column overall capacity studied. The results show that casting a high strength core in the joint can restore virtually all the capacity of the high strength concrete column. Debonding the top flexural reinforcement of the floor where it passes through the joint reduces the impact of floor loading on the strength of the joint. Where the joint is made of normal strength concrete, the vertical strain in the joint at maximum concrete stress would be sufficient to develop high strength steel provided there is sufficient development of this steel in the column outside of the joint.

REFERENCE
Ehab Abdel Wahab, and Scott Alexander, High Strength Concrete Columns with Intervening Normal Strength One-Way Slab and Beam Floors, SN2622, Portland Cement Association, Skokie, Illinois, USA, 2005.

High Strength Concrete Columns with Intervening Normal Strength One-Way Slab and Beam Floors
by Ehab Abdel Wahab1 and Scott Alexander2

INTRODUCTION
One of the best applications to exploit the advantages of high-strength concrete is in the columns of buildings. When the design objective is to carry an axial compression load, the most economical solution will use the highest strength concrete available to produce the smallest cross-section possible. Using high strength concrete (HSC) not only reduces the capital cost of the structure, it improves utility by increasing useable space. A problem arises at the joints between the high strength columns and intervening floors. As a rule, the joint region is cast with the same concrete as the floor; however, there are limited advantages to using high strength concrete in the floor system. As a result, the designer may wish to use in the floor a concrete that has a much lower design strength than the supporting column. The effect of this is to place a layer of lower strength concrete in the column, reducing the effective strength of the column. The magnitude of the strength loss depends on the strength differential between the column and floor concrete, the level of loading on the floor system and the geometry of the joint. Even in the best circumstances, the strength of the joint can significantly reduce the useable strength of the column. The primary objective of this investigation is to examine two methods intended to reduce the impact of an intervening floor on the strength of a column. The first is the casting of an HSC core within the joint. The second is debonding, in the joint, of the top flexural reinforcement in the floor. The secondary objective is to determine whether extreme loading on a floor might initiate failure in a column at service load. The tertiary objective is to obtain data for a type of floor-column joint not previously tested by any researcher.

BACKGROUND
Existing research (Bianchini et al, 1960; Gamble and Klinar, 1991; Ospina and Alexander, 1997; Ospina and Alexander, 1998) shows that the effective concrete strength of an HSC column with an intervening normal strength concrete (NSC) floor is generally greater than the concrete strength of the floor but less than that of the column. For a given combination of column and floor concrete, the effective strength of the joint depends on the degree to which the surrounding

1 2

Ph.D. Candidate, Dept. of Civil Eng., University of Alberta, Edmonton, Alberta, Canada, T6G 2G7. Senior Structural/Bridge Engineer, UMA Engineering Ltd., 17007 107 Ave. Edmonton, Alberta, Canada, T5S 1G3.

floor confines the joint. Joints at edge and corner column locations are less confined than those at interior columns. Loading of the floor system reduces the effective strength of the joint. Above the neutral bending axis, flexural tension in the floor acts to pull the joint apart. The effect of floor loading is most pronounced for interior column locations where the joint is confined on all sides. Floor loading appears to be less important at edge and corner column locations. The effective strength of the joint is inversely proportional to its aspect ratio, h / c, where c is the least column dimension and h is the maximum depth of floor concrete within the joint (Ospina and Alexander, 1998). The most severe aspect ratios are likely to occur at joints between edge or corner columns and floors with perimeter beams. ACI 318 02 (ACI Committee 318, 2002) deals with two types of joints, labeled here as either unconfined or confined. In most cases, joints are unconfined. Unless special steps are taken to strengthen the joint, the effective strength of the column with an unconfined floorcolumn joint is limited to 1.4 times the strength of the floor concrete. The joint is confined only if it is surrounded on all sides by beams or a slab of approximately equal depth. The strength of the column with a confined joint is higher than the unconfined case but it will still fall short of the strength based on the column concrete properties. Section 10.15 of ACI 318 02 directly addresses two strategies to strengthen a floor column joint. The designer may add vertical reinforcement (dowels) to the joint to reinforce as required. This option leads to impractical levels of reinforcement in cases where the required strength increase is large. Alternatively, the designer may call for high strength concrete to be placed in the vicinity of the floor-column joint, a procedure called puddling or mushrooming. This option can be a construction and management problem, as the floor must be cast using two grades of concrete. An implicit third option is to avoid the problem altogether by increasing the strength of the entire floor system so that the design strength of the column concrete does not exceed that of the floor by more than forty percent. All of these options can result in substantial added cost, reducing the advantage of HSC columns.

EXPERIMENTAL PROGRAM Overview

The research program involved seven 2 / 3-scale test specimens, labeled SP1 through SP7, tested to failure. Six specimens modeled the joint between an interior HSC column and a normal strength one-way slab and beam floor system. The remaining specimen, SP7, had normal strength concrete throughout. Figure 1 shows the test set-up. Two loading regimes were considered. Under Type I loading the column load was increased to failure while maintaining a service level of load on the floor. This method of testing is similar to that reported in Ospina and Alexander (1998) on flat plate-column joints. Type II loading involved increasing the floor load while maintaining a service level of load on the column. The research objective here was to see if the flexural tension in the floor was sufficient to weaken the joint and cause a premature failure of the column. In an effort to restore the strength of the HSC column, two specimens, SP3 and SP4, were cast with high strength concrete in the core of the column-floor joint. In specimens SP5 and SP6 the top reinforcement in the beam was debonded through the joint region in an effort to reduce the effect that floor loading has on the strength of the joint.

Materials
Concrete. High strength concrete for the columns was produced in a 0.25 cubic meter (9 cubic foot) laboratory mixer. The typical mix design is given in Table 1. The target strength for the high strength concrete was 80 MPa (11.6 ksi). A local ready-mix concrete producer supplied concrete for the floors. The target strength for the floor concrete was 20 MPa (2.9 ksi). Column and floor concretes were consolidated using a pencil vibrator and cured in conformance with ASTM C 31. Eight 152 x 304mm (6 x 12in.) concrete cylinders were cast for each batch of normal strength concrete and eight 100 x 200mm (3.9 x 7.9in.) concrete cylinders were cast for each batch of high strength concrete. Half of these were moist-cured and tested at 28 days. The remaining cylinders were field-cured alongside the specimen and tested at the same time as the main specimen, at ages ranging from 34 to 171 days. ASTM C 39, ASTM C 192, and ASTM C 511 were followed for testing and curing of concrete cylinders. While sulfur capping compound was used for the NSC cylinders, the faces of the HSC cylinders were ground. Table 2 shows results of all concrete cylinder tests for all specimens. Reinforcement. Four different bar sizes were used. The characteristics of all reinforcement were established by tension tests according to ASTM A 370 92. Table 3 reports average material properties for each size of reinforcing bar.

Specimen Design and Fabrication

All seven specimens had the same dimensions and reinforcement details, as shown in Figure 2 through Figure 4. Table 2 lists details of the test specimens. A 100-mm (3.9-in.) slab measuring 1.4 m x 3.2 m in plan was supported on 250 x 360mm (9.8 x 14.2in.) beam. The beam framed into a 250 x 350mm (9.8 x 13.8in.) column. The column sections above and below the joint were made long so that there would be a segment of column above and below the joint that was free of boundary effects. Three factors led to this choice of specimen geometry. First, the joint has a relatively high aspect ratio and is essentially unconfined perpendicular to the axis of the beam. These conditions present a severe test of any method to restore the full strength of the HSC column. From the design perspective, such a joint is also has the most to gain from any strengthening technique. Second, an interior column-floor joint is convenient from a testing point of view since the floor loading can be applied symmetrically with respect to the column. Floor loading is essential to the proper modeling of the stress and strain conditions within the joint. While an edge or corner column presents a similar lack of confinement to the joint, applying a floor load to such a specimen introduces lateral loads to the column resulting in complications to the testing frame. Finally, there are no other tests of joints between an HSC column and a one-way slab and beam floor system. The reinforcement for each specimen was assembled in two cages, one for the column and the other for the floor (slab and beam). Cages were held in position within the forms using plastic chairs, placed outside the test region. The plywood forms were oiled before each casting.

Each slab had flexural and shear strengths sufficient to prevent premature failure of the slab. The beams were designed to be governed by flexure at about 75% of balanced conditions. The columns ends were enclosed in steel shoes to prevent specimen failure resulting from local crushing at the cut off points of the column reinforcement. The dimensions of the HSC cores in specimens SP3 and SP4 are shown in Figure 2. The core concrete was cast as part of the upper column. Partial debonding in SP5 and SP6 involved encasing the top beam reinforcement within the joint in PVC sleeves. As-built dimensions and column misalignments were measured for use in detailed analysis. The verticality of the column forms was verified by using bulb-level while the alignment of the top column forms with the bottom column was verified using a plumb lines and a laser-oriented leveler.

Loading Apparatus and Instrumentation

Each specimen was subjected to a primary load Pc applied to the column by a 6600 kN (1500 kip) capacity universal testing machine and secondary load Pf applied to the floor using two 30ton and two 60-ton hydraulic jacks, as shown in Figure 5. The four hydraulic jacks were connected through a common manifold and controlled by a single hand pump. Five load cells monitored all five applied loads. The displacements of the floor and column were monitored by an array of 15 linear variable differential transformers. Concrete strains in each column were measured using nine embedded strain gauges at three different sections. A total of 50 foil resistance gauges were attached to the steel reinforcement in the slab, beam and column.

TESTING PROCEDURE
The floor and column loading records are presented graphically in Appendix A. Figures A1 through A7 show the floor and column loads for each test plotted against the stroke of the universal test machine. Because it includes deflection of the loading frame, the stroke of the test machine does not exactly match the total deformation of the test specimen; nevertheless, it provides a useful metric of the testing record. Under Type I loading (SP2, SP4, SP5, and SP7), column and floor loading was applied in two main stages. The first stage simulated sequences of construction and service loading on the floor and column. At the end of the first stage, the floor load was increased to approximately 160 kN (36 kip) simulating full service load of 9 kN/m2 (1.31 psi) on the floor. In the second stage, this floor load was maintained as the column load was increased to failure. With Type II loading (SP1, SP3, and SP6) column and floor loading was applied in three stages. As was the case for Type I loading, the first stage simulated sequences of construction and service loading on the floor and column. In the second stage, a service load was maintained on the column while the floor load was increased. In all specimens tested here, the floor beams reached their flexural capacity. Crushing of the flexural compression block and a drop in the floor load marked the end of the second stage. In the third stage, the column load was increased to failure.

OBSERVATIONS AND TEST RESULTS

4

In presenting the test results it is necessary to distinguish two terms, effective stress and effective strength. From a plane section analysis of any test, the effective stress in the concrete is given by:

fc =

Ptest Ast f st Ag Ast

[1]

where fc is the effective stress in the concrete, Ptest is the column load, Ag is the gross area of the column, Ast is the area of longitudinal column reinforcement, and fst is the stress in the longitudinal reinforcement estimated on the basis of strain gage readings. In presenting stress-strain results for tests, the reported effective joint stresses are those obtained from [1] with the gross area of the joint assumed equal to that of the column, Ag. To account for the effect of floor load on the average vertical load in the joint, the value of Ptest is taken as Pc + 2 3 Pf . The effective stress is a reasonable estimate of the true stress in the concrete at every stage of loading. The effective strength of the joint concrete is derived from the design equation given in ACI 318-02 for the column crushing strength under zero eccentricity.

P0 = 0.85 f c ' Ag Ast + Ast f y

[2]

With the measured failure load, Pult, substituted for P0, rearranging [2] produces an expression for the effective strength of the joint concrete, f ce ' . f ce ' = Pult Ast f st 0.85 Ag Ast

[3]

The effective strength is notionally the cylinder strength of some hypothetical concrete that combines the properties of the floor and column concretes.

General Observations
Specimens SP2 and SP5, both with Type I loading and NSC joints, failed by crushing of the joint concrete. Figure 6 shows a photograph of SP2 after testing and with loose concrete removed. It can be seen that the beam is relatively intact. Buckling of the longitudinal column reinforcement and spalling of the side cover concrete accompanied crushing of the joint concrete. The spalling extended into the high strength concrete of the lower column. The behavior of SP5 was similar to that of SP2. Specimens SP1 and SP6 had similar behavior. Both had NSC joints and were subjected to Type II loading. Both failed first by yielding of the tension reinforcement followed by crushing of the flexural compression block of the beam. At ultimate load in the column, the joint concrete crushed with spalling of the side cover and buckling of the longitudinal column reinforcement. Figure 7 shows SP1 after testing. Compared to SP2 in Figure 6, there is substantial loss of section in the beam and buckling of the beam compression reinforcement. SP3, with Type II loading and a high strength concrete joint, failed first by rupture of the flexural compression block of the beam. While it appeared that failure of the beam did some damage to the joint concrete, it did not seriously compromise the strength of the joint under column load. Failure of the column was by rupture of the upper column, as shown in Figure 8.

SP4, with Type I loading and a high strength joint, failed in both the joint and in the lower column. Figure 9 shows SP4 after testing. Buckling of the column reinforcement can be seen in both the joint region and the lower column. SP7, with Type I loading and normal strength concrete throughout, failed by crushing in the lower column, the weaker of the two columns. Figure 10 shows SP7 after testing.

Joint Strain
Figures 11 and 12 show the effective stress as calculated by [3] plotted against vertical strain in joint as measured by embedded strain gauges in the concrete. In all cases combining a normal strength joint with high strength columns (SP1, SP2, SP5, SP6), the peak column stress occurs at a strain in excess of 1.5 percent and there is a significant descending (softening) branch to each stress-strain curve, indicating that most of the axial deformation of the column was localized in the joint. Because neither specimen SP3 nor SP7 failed in the joint, the stress-strain curves show no softening branch. The results for SP4 show a flattening of the stress-strain curve for the joint, suggesting that the ultimate strength of the joint had been reached; however, rupture of the lower column resulted in unloading of the column prior to softening within the joint. Figures 13 and 14 show the effective joint stress plotted against the average transverse joint strain as measured by embedded strain gauges. For those specimens failing in the joint (SP1, SP2, SP5, SP6), crushing of the joint concrete is marked by a significant increase in transverse strain, resulting in a long plateau in the stress-strain curve. Transverse strains in specimens with high strength joints, SP3 and SP4, are substantially smaller than those for specimens with normal strength joints; nevertheless, both curves show signs of flattening, suggesting that the concrete in the joint was at or near ultimate.

Moment Curvature of Beams

Figures 15 through 17 show typical bending moment vs. curvature results for the beams. The curvature is based on strain measurements on the top and bottom beam reinforcement at the face of the column. The bending moment is calculated at the face of the column and is normalized with the nominal flexural strength of the beam, given by:
As f y M n = As f y d 1 .7 f c ' b w

[4]

For specimens with Type II loading, illustrated in Figure 15, the beam reached its nominal capacity as predicted by standard analysis. There was no apparent effect of beam load on the response of the column to axial load. For specimens with Type I loading, the beam was loaded to approximately 50% of its nominal capacity. As shown in Figs. 16 and 17, increasing the column load had the effect of increasing the measured curvature at the face of the column. The curvature increase is much larger in the case of SP2 with a normal strength joint, Figure 16, than it is for SP4 with a high strength joint. By comparison to those measured under Type II loading, the beam curvatures under Type I loading are small.

Effective Strength
Figure 18 shows the relationship between the effective stress ratio at failure, f 'ce f 'cs , and the columns strength ratio, f 'cc f 'cs . Also shown are curves representing the nominal column strength and joint strength according to ACI 318 02. The column strength ratio is calculated using the lower of the two column strengths. In all tests, the current provisions of ACI 318 02, indicated by the bold line in Figure 18, underestimate the nominal strength of the joint. The lowest effective strength ratios were for the two specimens with normal strength joints and Type II loading (SP1 and SP6). It is felt that some damage was done to the joint concrete when the flexural compression block ruptured, resulting in a slight reduction in effective strength. Both specimens with high strength joints (SP3 and SP4) developed essentially the full strength of the weaker of the two columns framing into the joint. The effect of debonding reinforcement is at most slight. The effective stress ratios of the debonded specimens, SP5 and SP6, are slightly higher than those of their companion specimens, SP2 and SP1, respectively.

DISCUSSION
The results presented here suggest that the existing design provisions in ACI 318 02 are safe for joints between high strength interior columns and normal strength one-way slab and beam floor systems, even with significant load on the floor. Similar results would be expected for joints at edge column locations as the confinement conditions at this location are like those tested here. The ACI limit of 1.4 times the floor concrete strength falls substantially below all of the results from specimens with high strength columns. With regard to strengthening the joint by adding reinforcement (referred to as dowels in ACI 318), basing the joint strength on the lower strength concrete rather than 1.4 times this value seems overly conservative and a little illogical. There is no justification for ignoring the enhancement of the joint concrete strength simply because the designer has decided to add within the joint more longitudinal reinforcement. The very large longitudinal strains for specimens SP1, SP2, SP5 and SP6 suggest that there may be benefit in considering high strength steel for strengthening joints between HSC columns and NSC floors. For these specimens, the maximum concrete stress was reached at a strain of 1.5%, more than enough to yield even prestressing steel. Using high strength steel for dowels would reduce reinforcement congestion in the joint and make this design option a little more practical. The results clearly show that providing a high strength concrete core in the joint can restore essentially the full strength of the column. In these tests, the area of the HSC core accounted for 74% of the column cross section. How large this percentage has to be for this strengthening technique to be effective is a question for further study. It should be easier to achieve a higher percentage with a larger scale joint. There is also the possibility of using high strength precast cores or steel embedments, such as a hollow section filled with high strength concrete, that can be placed when casting the lower column. The loading regime had little effect on the strength of the joints under column load. Under Type I loading the strength of the column was tested while the floor was under a generous service load. Nevertheless, all specimens easily surpassed their design strength based on ACI

318 02. With Type II loading no column failed as a result of extreme loading on the floor. All floor beams reached their full flexural capacity before rupture of the compression block. The above comments notwithstanding, partial debonding of the beam top reinforcement appears to have some potential to improve the capacity of the column joint. It may be that the beneficial effect observed here was not dramatic because the effect of floor loading was not severe. This might not be the case for situations such as joints between high strength interior columns and two-way flat plate floors, in which the effect of floor loading is pronounced.

SUMMARY AND CONCLUSIONS

Seven 2 / 3-scale specimens were tested under two different loading regimes to investigate the behavior of joints between interior high strength concrete columns and one-way slab and beam floors made of normal strength concrete. Two construction techniques were presented and their effects on the overall column capacity were studied. The major findings are as follows. Casting an HSC joint with the top column can restore the full strength of a high strength column, making the procedure a viable alternative to puddling two different concretes or casting the entire floor with high strength concrete, especially for applications where there is high strength differential between the column and floor concretes. More work is needed to determine size limits for the high strength core to be effective. Partial debonding of the beam top RFT has some potential to improve the capacity of the column joint. The strains at maximum concrete stress in joints between HSC columns and NSC floors were in excess of 1.5%. Such high strain values would be sufficient to develop high strength steel reinforcement (dowels) to increase the capacity of the joint provided such reinforcement can be effectively developed outside the joint. The combination of ultimate load on the floor and service load on the column did not cause failure of the joint; however, flexural failure of the floor caused a reduction in the cross-sectional area of the joint and consequent reduction in the column capacity.

ACKNOWLEDGEMENTS
The research reported in this paper (PCA R&D Serial No. 2622) was conducted in Structural Laboratory of University of Alberta, Canada with the sponsorship of the Portland Cement Association Education Foundation (PCA Project Index No. F09-02). The contents of this report reflect the views of the authors, who are responsible for the facts and accuracy of the data presented. The contents do not necessarily reflect the views of the Portland Cement Association or the PCA Education Foundation. The authors would like to thank Larry Burden and Richard Helfrich for their valuable technical assistance at I.F. Morrison Structural Engineering Laboratory.

REFERENCES
ACI Committee 318 (2002), Building Code Requirements for Reinforced Concrete (ACI318-02) and Commentary (ACI 318R-02), American Concrete Institute, Farmington Hills, Michigan, USA, 391 pages.

Bianchini, A.C.; Woods, R.E.and Kesler, C.E., Effect of Floor Concrete Strength on Column Strength, ACI Journal, Vol.31, No. 11, 1960, pages 1149-1169. Gamble, W.L. and Klinar, J.D., Tests on High Strength Concrete Columns with Intervening Floor Slabs, Journal of Structural Engineering, ASCE, Vol. 117, No. 5, pages 1462-1476. Ospina, C.E.and Alexander, S.D.B., Transmission of High Strength Concrete Column Loads through Concrete Slabs, Structural Engineering Report 214, Department of Civil Engineering, University of Alberta, Edmonton, Canada, 1997. Ospina, C.E. and Alexander, S.D.B., Transmission of interior concrete column loads through floors, Journal of Structural Engineering, ASCE, 124(6), 1998, pages 602610.

NOTATION
Ag: As: Ast: bw: c: d: fc: fc: fcc: fce: fcs: fy: h: Mn: Pc : Pf : Ptest: Pult: P0:
Gross area of column Beam flexural reinforcement Area of longitudinal reinforcement in column Beam width Short dimension of column Flexural depth Effective stress in joint concrete Specified compressive strength of concrete Compressive cylinder strength of column concrete Effective compressive strength of joint concrete Compressive cylinder strength of floor concrete Yield strength of beam reinforcement Overall floor thickness Nominal flexural strength Load applied to upper column Total load applied to floor Average column load in joint assumed equal to Pc + 2/3 Pf Value of Ptest at failure Nominal compression strength of column under zero eccentricity

Table 1. Typical Mix Design for High Strength Concrete (per cubic meter)

Cement (Type III), kg Silica fume, kg Coarse aggregate (SSD), kg Fine aggregate (SSD), kg Superplasticizer, litres Total water, kg Water / Cement Water / Total Cementitious
Note: SSD = saturated surface dry

491 27 1059 611 16 128 0.26 0.25

Table 2. Measured Compressive Strength of Concrete (MPa) at 28 days at time of test Specimen Lower Floor Joint Upper Lower Floor Joint Column Column Column SP1 79.1 13.3 13.3 69.8 84.6 18.0 18.0 SP2 82.2 17.9 17.9 77.1 86.6 20.4 20.4 SP3 82.0 27.8 81.2 81.2 89.2 31.8 90.7 SP4 81.2 15.7 63.5 63.5 90.7 17.6 66.7 SP5 63.5 18.5 18.5 91.3 66.7 19.8 19.8 SP6 89.6 18.9 18.9 91.4 93.7 20.6 20.6 SP7 17.8 17.8 17.8 28.2 18.7 18.7 18.7

Upper Column 72.8 93.0 90.7 66.7 95.5 94.9 28.2

Table 3. Properties of Steel Reinforcement Bar diameter Yield strength Ultimate strength (mm) (MPa) (MPa) 16 409 613 16 442 700 11 458 618 9 500 700 6 418 538

Remarks Used in SP1 through SP6 Used in SP7 Used in all specimens Used in all specimens Used in all specimens

10

LIST OF FIGURES
Figure 1. Overall Test Set-up Figure 2. Dimensions of Test Specimen Figure 3. Floor Reinforcement Details Figure 4. Column Reinforcement Details Figure 5. Applied Loads Figure 6. Failure of Joint of SP2 after Test Figure 7. Failure of Joint of SP1 after Test Figure 8. Failure of Top Column of SP3 after Test Figure 9. Failure of Bottom Column of SP4 after Test Figure 10. Failure of Bottom Column of SP7 after Test Figure 11. Type I Specimens: Effective Stress vs.Vertical Strain Figure 12. Type II Specimens: Effective Stress vs. Vertical Strain Figure 13. Type I Specimens: Effective Stress vs. Average Lateral Strain Figure 14. Type II Specimens: Effective Stress vs. Average Lateral Strain Figure 15. Effect of Column Load and Floor Moment on Floor Curvature for SP6 Figure 16. Effect of Column Load and Floor Moment on Floor Curvature for SP2 Figure 17. Effect of Column Load and Floor Moment on Floor Curvature for SP4 Figure 18. Effective Strength Ratio Figure A1. Column and Floor Loads vs. Stroke for SP1 Figure A2. Column and Floor Loads vs. Stroke for SP2 Figure A3. Column and Floor Loads vs. Stroke for SP3 Figure A4. Column and Floor Loads vs. Stroke for SP4 Figure A5. Column and Floor Loads vs. Stroke for SP5 Figure A6. Column and Floor Loads vs. Stroke for SP6 Figure A7. Column and Floor Loads vs. Stroke for SP

11

Figure 1. Overall Test Set-up

Figure 6. Failure of Joint of SP2 after Test

Figure 7. Failure of Joint of SP1 after Test

Figure 8. Failure of Top Column of SP3 after Test

Figure 9. Failure of Bottom Column of SP4 after Test

Figure 10. Failure of Bottom Column of SP7 after Test

80

70

60
SP4 Effective Stress (MPa)

50

SP2

40

SP5

30

Gage Locations

20
SP7 failed in bottom column

10

0 0 5 10 15 20 25 30 35
Thousands

Vertical Strain (micro)

Figure 11. Type I Specimens: Effective Stress vs. Vertical Strain

80
SP3 failed in top column

70

Gage Locations

60

Effective Stress (MPa)

50
SP6

40
SP1

30

20

10

0 0 5 10 15 20 25 30 35
Thousands

Vertical Strain (micro)

Figure 12. Type II Specimens: Effective Stress vs. Vertical Strain

70

60

SP4 SP2

50
* Effective Stress (MPa)

40
SP5

30
Gage Locations

* Gage was lost

20

10

0 0 2 4 6
Average Lateral Strain (micro)

10

12
Thousands

Figure 13. Type I Specimens: Effective Stress vs. Average Lateral Strain

80
SP3
Gage Locations

70

60

Effective Stress (MPa)

50
SP6 *

40

30

SP1 * Gage was lost

20

10

0 0 2 4 6
Average Lateral Strain (micro)

10

12
Thousands

Figure 14. Type II Specimens: Effective Stress vs. Average Lateral Strain

1.20

1.00

0.80
M/Mu & P/Pu

0.60 M P 0.40
Gage Locations M: moment P: Column Load

0.20

0.00 0.00

2.00

4.00

6.00
Curvature (Degree/m)

8.00

10.00

12.00

Figure 15. Effect of Column Load and Floor Moment on Floor Curvature for SP6

1.20
Gage Locations

1.00

0.80
M/Mu & P/Pu

0.60

0.40 M P 0.20
M: moment P: Column Load

0.00 0.00

2.00

4.00

6.00
Curvature (Degree/m)

8.00

10.00

12.00

Figurte 16. Effect of Column Load and Floor Moment on Floor Curvature for SP2

1.20
Gage Locations

1.00

0.80
M/Mu & P/Pu

0.60 M P 0.40
M: moment P: Column Load

0.20

0.00 0.00

2.00

4.00

6.00
Curvature (Degree/m)

8.00

10.00

12.00

Figure 17. Effect of Column Load and Floor Moment on Floor Curvature for SP4

5.0
Column Equation, ACI 318-95

4.0 SP4 failed in joint and in bottom column

SP3 failed in top column 3.0 f'ce / f'cs

SP2 SP7 failed in bottom column 2.0 SP5 SP1 SP6

1.0
Joint Equation, ACI 318-95

0.0 0.0 1.0 2.0 3.0 f'cc / f'cs 4.0 5.0 6.0 7.0

Figure 18. Effective Strength Ratio

APPENDIX A

700

600
Column Load (10 kN), Floor Load (kN)

Column Floor

500

400

300

200

100

0 0 5 10
Stroke (mm)

15

20

25

Figure A1. Column and Floor Loads vs. Stroke for SP1

700

600
Column Load (10 kN), Floor Load (kN)

Column Floor

500

400

300

200

100

0 0 5 10
Stroke (mm)

15

20

25

Figure A2. Column and Floor Loads vs. Stroke for SP2

700

600 Column Load (10 kN), Floor Load (kN)

Column Floor

500
MTS unloaded automaticlly during night

400

300

200

100

0 0.0 5.0 10.0 Stroke (mm) Figure A3. Column and Floor Loads vs. Stroke for SP3 15.0 20.0 25.0

700

600
Column Load (10 kN), Floor Load (kN)

Column Floor

500

400

300

200

100

0 0 5 10
Stroke (mm)

15

20

25

Figure A4. Column and Floor Loads vs. Stroke for SP4

700

600

Column Floor

Column Load (10 kN), Floor Load (kN)

500

400

300

200

100

0 0 5 10
Stroke (mm)

15

20

25

Figure A5. Column and Floor Loads vs. Stroke for SP5

700

600
Column Load (10 kN), Floor Load (kN)

Column Floor

500

400

300

200

100

0 0 5 10
Stroke (mm)

15

20

25

Figure A6. Column and Floor Loads vs. Stroke for SP6

700

600

Column Floor

Column Load (10 kN), Floor Load (kN)

500

400

300

Failure happened at bottom column

200

100

0 0 5 10
Stroke (mm)

15

20

25

Figure A7. Column and Floor Loads vs. Stroke for SP7