of this test are illustrated in figure 9.12. This test is also plotted in figure 9.13 along with Test 2 which had 0.9D and 25 psf wind load. The agreement between these two tests is good. The structure experienced a considerably larger lateral drift under the application of the larger vertical load acting alone than it did under the smaller vertical load (0.024 in versus 0.007 in). Under the subsequent wind-load application the structure with the larger vertical load exhibited greater stiffness than it did when more lightly loaded. At the point where the 15 psf wind load was reached, the two deflections were approximately equal (0.050 in).

9.3.1.3. Frame Action Versus Wall Action

The frame was racked after removal of the walls in Test No. 14. Results are illustrated in figure 9.14. This test is compared with an identical racking test performed before removal of the walls in figure 9.10. This comparison clearly indicates that a major portion of the lateral stiffness is provided by the walls rather than by the frame.

Figure 9.15 shows the results obtained in a later racking test (Test 18) on the structure with walls removed. This test was carried to the point where the frame no longer developed increasing resistance to load. In this test the structure had a vertical load on the columns of 0.9D. In interpreting these results it must be remembered that the wind load reported here is in pounds per square foot of total vertical surface area of the structure. If none of the walls are present, then the surface area upon which the wind forces act is also not present. However, it is also conceivable that a situation could develop in which walls in one direction are present while walls in the opposite direction are absent. In such a case these results would have relevance. Figure 9.15 indicates that the frame acting alone in the north direction cannot be expected to withstand a wind force in excess of 10 psf on the gross area of the structure. By the time this test was performed the structure had been carried through a number of earlier tests which might have somewhat weakened the frame. However, the structure at this point exhibited no obvious signs of distress attributable to earlier testing. It is recognized that the simulation of the column foundation which was used in the test structure was extremely conservative compared to that used in the real structure, particularly with respect to tests without walls. Thus the results of this test possibly fall well below the results which would be obtained from the test of a real structure.

9.3.2. Horizontal Loads in the East Direction

West wind forces in the test structure were resisted by one interior "structicore" wall and one exterior wall. These walls would not normally be expected to be as strong as the fire walls; however, their rigidity in the lateral load tests appeared to be comparable to that of the fire walls. Both of the walls in this direction had openings; however, these walls also had a greater overall length resisting load.

9.3.2.1. Racking Tests With Minimum Vertical Loads

These racking tests were conducted in the same manner as in the south direction (sec. 9.3.1.1). Figure 9.16 illustrates Test 3, which subjected the structure to 0.9D plus 25 psf wind load from the west. In this test, measured deformations were extremely small (0.012 in). Recovery characteristics were similar to those observed for the fire walls. There appear to be two breaks in the load-deflection curve (fig. 9.16), one at 10 psf and the other at 24 psf. Neither of these was accompanied by any visual signs of distress in either the concrete frame or in the gypsum walls. Figure 9.17 shows the results obtained from a racking test (Test 11) carried to a wind load in excess of 70 psf. A portion of this test, along with the results of Test 3, is shown in figure 9.18. It is interesting to note that a definite break developed in the load-deflection curve of Test 11 at 6 psf; again, this break was not associated with any visual signs of distress. These breaks in the load deflection curve are not considered to be particularly significant since, for example in Test 11, even at a wind load of 25 psf the lateral drift of the structure is still less than 0.04 in. Figure 9.19 shows a plot of load versus wall diagonal compressive deformation for a wall resisting west wind load during Test 11, which was carried to 74 psf. As was the case for the walls resisting south wind load (fig. 9.10) deformations and correspondingly average strains were extremely small and recovery was good. In this test, distress in the wall was not noted until the very upper range of the loading sequence was reached. At these loads distress was observed in the interior (north) wall in the form of shear cracks (fig. 9.20). However, no noticeable distress was observed in the wall-frame connections. Some progressive opening of the joint between the wall planels and the columns was observed at loads in excess of 35 psf but the separation was not particularly pronounced and was similar to that discussed in section 9.3.1.1.

9

The erratic behavior noted in the first unload-reload cycle was due to a stuck instrument which was subsequently freed.

9.3.2.2. Racking Test at High Vertical Loads

A racking test was performed in the east direction in Test 8 with 1.3D + 1.7L and 15 psf wind load. The results of this test are illustrated in figure 9.21. It should be noted that the application of the vertical loads caused a horizontal deflection in the opposite direction to that in which the wind loads were subsequently applied. This deflection, resulting from the vertical loads, was only partially reversed by the application of the 15 psf wind load. Test 8 (1.3D 1.7L) is plotted along with Test 3 (0.9D) in figure 9.22. It can be seen that in either case the deformations due to lateral loads are so small as to make questionable any conclusions concerning the effect of the magnitude of vertical load on lateral stiffness.

9.3.2.3. Frame Action Versus Wall Action

Figure 9.16 (Test 3) shows the response of the structure with walls to racking in the east direction, while figure 9.23 (Test 15) shows the response of the structure after removal of the walls. Both of these curves are plotted together in figure 9.18 10. As was the case in the north direction, it is evident that in the east direction the walls provide most of the stiffness against lateral loads.

In designing the test sequences to which the structure was subjected, it was felt that only one meaningful racking test could be carried through to the point at which the structure was approaching collapse. The north direction, which is the narrow direction for this system, was chosen for this test and its results were reported earlier (Test 18) in section 9.3.1.3. In the east direction the maximum load to which the frame without walls was subjected was applied in Test 15, which is shown in figure 9.23. In this test the frame resisted a wind load in excess of 15 psf without collapse, although some minor flexural cracks were observed in the columns at maximum load. This load level was substantially higher than that sustained in the racking test without walls in the north direction. The conservative nature of the foundation simulation in the test structure, which provided a hinge at the lower column connection, undoubtedly affected the results obtained. in this test in an adverse manner.

9.3.3. Compliance With Performance Criteria, Horizontal Loads

9.3.3.1. Performance Criterion 4.4.1, Horizontal Deflection Under Dead and Wind Loads

At a load level of 0.9D + 1.1H the horizontal deflection due to the superimposed load of 1.1H shall not

10 Only a portion of Test 15 is shown on this figure because of the scale.

Dh = horizontal gross deflection h = height above grade.

In Test 2 (fig. 9.8) under 0.9D and a wind load of 22 psf (1.1H) from the south, the maximum lateral drift was approximately 0.073 in, and in Test 3 (fig. 9.16) with a west wind load. the maximum lateral drift was approximately 0.007 in, while the allowable drift under this criterion is 0.19 in.

Criterion 4.4.1 is therefore satisfied.

9.3.3.2. Performance Criterion 4.4.3, Horizontal Deflection Under Dead, Live, and Wind Load

At a load level of 1.3D + 1.7L + 0.8H the horizontal deflection due to the superimposed load of 0.3D + 1.7L +0.8H shall not exceed the following:

Dh≤ 0.002h = 0.19 in

In Test 7 (fig. 9.24) under 1.3D + 1.7L and a wind load of 15 psf (0.8H) from the south, the maximum lateral drift was 0.045 in and in Test 8 (fig. 9.20) with a west wind load the maximum lateral drift was 0.032 in", while the allowable drift under this criterion is 0.19 in. Criterion 4.4.2 is therefore satisfied.

9.3.3.3. Performance Criterion 4.4.5(b), Ultimate Strength The structure or any portion thereof shall not fail at a load smaller than the following:

(b) 0.9D + 1.4H, (28 psf wind load).

The structure was tested under these loading conditions in the north direction and in the east direction in Tests 10 and 11 respectively (see figs. 9.9 and 9.17). No distress was experienced in either test at that load level.

Criterion 4.4.4.(b) is therefore satisfied.

9.4. Summary

(1) All conclusions pertaining to the structural performance of the system in question are based on the structure as built in the laboratory and on the erection methods and materials used therein. Variations in materials and erection methods may influence performance.

(2) The building system satisfied the performance criteria which were set for its evaluation with substantial margin. As a system, it exhibited strength and stiffness in excess of service and ultimate load requirements.

11 In Test 8 the maximum drift was measured upon the application of the vertical load and took place in the opposite direction to the wind-induced deflection. If these two had been in the same direction rather than opposite, the maximum deflection would have been approximately 0.04 in.

(3) The walls of the system behaved as an integral part of the structure. They provided most of the stiffness of the system with respect to lateral loads, and provided a significant portion of the stiffness against vertical loads.

(4) The building system with its walls removed had considerable reserve strength above the required vertical load bearing capacity; however, without the aid of its walls it was not capable of resisting the required service wind. loads.

10. Component Tests

10.1. Introduction

Load tests were conducted on the three principal precast load-bearing components of the structure. The components tested were columns, main beams with an appropriate portion of the topping slab connected to them, and a floorchannel slab. These tests were performed to determine the behavior and ultimate strength of the components. Included were tests to determine the effects of creep on the columns and repeated loading on the beams.

10.2. Column Tests

The column specimens tested were typical "short" columns as shown in figure 5.6. Reinforcement was 60 ksi steel. Actual outside dimensions and concrete cover of individual test specimens are shown in table 10.1. Values of concrete strength of the various specimens are reported in section 10.2.1.1. The tests consisted of the following:

(1) Short-term destructive loads were applied parallel to the column axis. Four columns were tested with an eccentric load on the major axis and three with an eccentric load on the minor axis.

(2) Two sustained-load tests were carried out: one with an eccentric load on the major axis, the other with an eccentric load on the minor axis.

The method of applying the eccentric loads to the columns is shown schematically in figures 10.1 and 10.2. The same method was used for both the short-term and the sustained-load tests.

12 The term "short" column is applied in the plans to a column with end fixtures at both ends. This column in the structure has the same slenderness ratio as all other columns.

10.2.1. Short-Term Destructive Tests

10.2.1.1. Specimens

Seven columns were tested to destruction, three with a load eccentricity of 0.5 in on the minor axis (e/t= 0.1 for columns 1, 5, and 8), three with a load eccentricity of 2.0 in on the major axis (e/t= 0.33 for columns 2, 6, and 7), and one with a load eccentricity of 1.5 in on the major axis (e/t= 0.25 for column 9).

Columns 1 and 2 were cast at the same time as the test structure components (April 16 and 17, 1968) and from the same concrete, with concrete compressive strength (f'c) ranging from 4900 psi to 7200 psi (see table 11.1). These specimens were approximately 20 days old when tested. Columns 5, 6, 7, 8, and 9 were cast from similar concrete at a later date (May 15). Columns 5, 6, 7, and 8 were approximately 30 days old when tested and the concrete had a compressive strength of approximately 5400 psi. Column 9 was 60 days old when tested and the concrete compressive strength was 7000 psi. The longitudinal reinforcing bars (No. 6 deformed bars) were approximately 3 in shorter than their full required length in Columns 1 and 2, leaving a distance of about 111⁄2 in between reinforcing bars and end fixtures, but in columns 5, 6, 7, 8, and 9 these bars were only 1/4 in shorter.

10.2.1.2. Loading

The loads were applied continuously until failure, through a knife-edge loading plate (figs. 10.1 and 10.2) by a 600,000-lb hydraulic testing machine at a rate of 8,000 lb per minute. Deflections were measured with longthrow mechanical dial gages at mid-height. Figure 10.3 illustrates a typical test setup.

10.2.1.3. Results

Test results are shown in figures 10.4 through 10.10 as load-deflection curves. Ultimate loads are tabulated in table 10.2. The average maximum load for columns 1, 5, and 8 (minor axis bending, e = 0.5 in) was 78.9 kips. Figure 10.11 shows these columns after testing. Column 1 failed near its end connection, and its mode of failure appeared to be partially due to the short reinforcement used. Columns 5 and 8 (which had longer reinforcement) experienced compression failures in the concrete at mid-height at about 12 percent higher loads than did column 1.

The average maximum load for columns 2, 6, and 7 (major axis bending, e = 2.0 in) was 51.2 kip. All three specimens failed in a similar manner, as illustrated in figure 10.12, with excessive bending of the channel-shaped,

top-fixture and some spalling of the concrete near this fixture.

The maximum load for column 9 (major axis bending, e = 1.5 in) was 89.5 kips. This column failed by concrete compression at mid-height.

10.2.2. Sustained Loading (Creep) Tests

10.2.2.1. Specimens

Two columns (columns 3 and 4) were tested under a 25 kip sustained load. Both columns were cast with the test-structure components and were about 20 days old when placed under load. Concrete compressive strength ranged from 4900 psi to 7200 psi (see table 11.1).

10.2.1.2. Loading

The loading frames used in these tests are shown in figures 10.13 and 10.14. Column 3 had a load eccentricity of 0.5 in on the minor axis (e/t = 0.1), and column had a load eccentricity of 2.0 in (e/t = 0.33) on the major axis. A detail of the bottom of the column 4 loading frame is shown as figure 10.15. This figure also shows the heavy spring used to sustain the load on the specimen.

The 25 kip load (1D+ 1L) was applied by means of a 30-ton hydraulic ram and a load cell inserted between the top two plates of the test frame. The ram load was applied through the column to the spring, causing the spring to compress. Once the required load was applied, nuts on the 3-in tie-bars were tightened against the top knife-edge plate. The deflection of the springs was about 112 in at the 25 kip load. The loads were checked and adjusted periodically.

Mid-height deflections were measured by means of a taut wire and a mirrored scale. The progression of the deflections with time was measured periodically.

10.2.2.3. Results

Results of a 170-day observation period are presented in figures 10.16 and 10.17 as timedeflection curves. The initial deflections are included in the total deflection for information and comparison purposes.

10.2.3. Interpretation of Column Test Results

10.2.3.1. Short-Term Destructive Tests

Figure 10.18 shows a plot of test results for columns with major axis load eccentricity, together with computed interaction curves for both the column cross section and the overall column slenderness effects. This figure also shows interaction curves derived from performance criteria for lower-story columns.

Curve C is the locus of extreme values of combined axial loads and end moments which the columns must be able to resist in the direction of their major axis. Critical loading conditions for columns were found to be 1.5D + 1.8L and 1.25 (D + L + H). The actual points plotted for these component requirements correspond to 1/0.8 times the critical loading, where 0.8 represents an understrength factor. The requirement here is similar to that explained in the commentary to section 4.4.5 of this report; namely, that in the absence of a laboratory sample of large size, individual columns are required to exhibit a strength in excess of their ultimate loading requirement. The ultimate moment imposed on the column by wind load was assumed to be 1/7 of the computed ultimate moment imposed on the frame in the absence of walls. This assumption is based on the test results illustrated in figure 9.10, which compares horizontal wind deflections of the total system to these of the system with walls removed. Curve A of figure 10.18 is a theoretical interaction diagram for crosssectional capacity, computed for the combinations of vertical load and moment which would cause failure in columns with a concrete compressive strength of 7,000 psi, which was the concrete compressive strength of column 9. Curve B is a similar interaction curve for the specified concrete compressive strength of 4,500 psi.

The total maximum moment acting on a column is the end moment plus an additional moment which equals the product of the applied vertical load times the maximum deflection of the column. To determine the combination of maximum axial load and maximum end moment that can be imposed on a column at its supports, the maximum column moment in interaction curves A and B, which represent the total cross-sectional capacity of the columns, must be reduced by the value of P X dh, which is the product of axial load and maximum column net deflection at column failure. Curve B' has been plotted to account for this moment reduction at the specified concrete compressive strength of 4,500 psi. Curve B' is therefore the interaction curve of ultimate loads and ultimate end moments which a column, constructed in accordance with the plans and specifications for this system, should be theoretically expected to resist. It may be noted by comparing Curves B' and C' that the theoretical column capacity exceeds the required critical loading by a considerable margin.

Values of P X dh were computed in accordance with "Proposals for Revision to Sections 915 and 916 of ACI 318-63" by MacGregor, Breen, and Pfrang (unpublished). These computations accounted for concrete cracking.

The actual column tests for loads with major axis eccentricity are also plotted in figure 10.18. It will be noted that each specimen test is plotted twice. The triangular points represent a plot of the axial load at failure against the end moment caused by the axial load times its eccentricity.

The square points represent a plot of the axial load at failure against the maximum moment that actually existed in the specimen at failure. This actual maximum moment is the product of the axial load times the sum of its end eccentricity and the maximum center line deflection at failure. It should be noted that in a slender column such as the specimens tested the maximum center line deflection is relatively large (Refer to figures 10.4 through 10.10).

Only column 9 failed by compression at midheight. Column 9 had a concrete strength of 7,000 psi, and it can be seen that this column developed strength slightly in excess of the strength predicted by interaction curve A. Columns 2, 6, and 7 failed at their top fixture and therefore did not develop their theoretical ultimate strength. In column 2 the reinforcement was 3 in short. This was corrected in columns 6, 7, and 9, without appreciable effect on columns 6 and 7. To evaluate column strength in terms of component requirements the triangular plots of the test results should be compared with curve C. It may be noted that all the tested columns had considerable excess strength over the component requirements.

Results for columns tested with minor axis. eccentricity are plotted in figure 10.19. Interaction curves A, B, and B' are plotted as in figure 10.18, except that in this figure they represent relationships for loads with minor axis eccentricity, and curve A was computed for a concrete compressive strength of 5,400 psi. Columns 5 and 8 experienced a compression failure at mid-height, and show strengths close to the theoretical strength predicted by interaction curve A, which was computed for 5,400 psi concrete (the actual strength of the concrete in these columns). Column 1 failed near its top fixture, and this mode of failure may have been caused by the fact that the reinforcement was 3 in short. In the case of minor axis eccentricity, no sizable end moments are expected to act on the columns, since the tie beams do not participate in the support of vertical loads to an appreciable extent. All columns tested at a minor axis eccentricity of 0.5 in were able to sustain vertical loads in excess of the 51 kip needed to satisfy the component requirements.

In summary, all the tested columns were able to withstand axial loads and moments in excess of required performance. Some of the columns did not develop their full theoretical ultimate strength because of weakness at the end fixture.

10.2.3.2. Sustained Loading (Creep) Tests

Figures 10.16 and 10.17 show the results of creep tests conducted on columns 3 and 4.

Column 4 was loaded with an axial load of 25 kip at an eccentricity of 2 in on its major axis. The test results are illustrated in figure 10.16. This figure also shows a computed value of instantaneous deflection. It should be noted that the vertical load is applied outside the kern of the section. The instantaneous deflection was therefore computed on the basis of a cracked section neglecting concrete tension. This will tend to overestimate the computed deflection, since not every section along the length of the column is cracked. In this case the computed instantaneous deflection is 0.32 in while the measured instantaneous deflection was only 0.22 in. However this measured value is low compared with values measured in tests on columns 2, 6, and 7 which were subjected to similar loading conditions. For the latter three tests the instantaneous deflection at a 25 kip load averaged 0.29 in (see figures 10.7, 10.8, and 10.9).

Figure 10.16 also shows an upper deflection limit for the given conditions, obtained by considering only the steel reinforcement and neglecting the concrete. This limit was computed by assuming that all the load is carried by the reinforcement. The deflection limit thus computed under the conditions of this test is 0.54 in and the steel stress at this deflection limit would be 38.2 ksi, which is well below the specified 60 ksi yield stress of the column reinforcement. Creep buckling under the conditions of this test therefore cannot occur. All computed deflections referred to in this section accounted for an added moment equal to the axial load multiplied by the deflection at each point along the column.

Column 3 was loaded with an axial load of 25 kip at an eccentricity of 0.5 in on the minor axis. This test is illustrated in figure 10.17, together with the computed instantaneous deflection of 0.106 in and the computed deflection limit assuming that no stress is carried by the concrete, which is 0.565 in. In this case the computed instantaneous deflection is based on an uncracked section and is in good agreement with the measured instantaneous deflection of 0.12 in as well as with instantaneous deflections measured in the tests of columns 1, 5, and 8 (figures 10.4, 10.5, and 10.6). The deflection limit under these loading conditions is 0.565 in and the computed steel stress at the deflection limit is 21 ksi, which precludes the possibility of creep buckling under the conditions of this test.

The creep tests on columns 4 and 3 respectively are also plotted in figures 10.20 and 10.21. Curves D in these figures are the interaction curves for maximum axial load and maximum