Page images
PDF
EPUB

total moment at which the steel carries the entire load without concrete participation. Curves D' show the same interaction curves for the reduced moments when deflections are taken into account. Thus the curves marked D' represent the combination of axial loads and applied end moments which can be supported by the column reinforcement without concrete participation as a limit condition. It is significant to note that curve D' in figure 10.20 when compared with curve C indicates that creep buckling can not occur in this structure even under the assumed ultimate loading conditions. The creep tests of columns 3 and 5 simulate sustained loading of 1 live + 1 dead load in the structure. It can be seen from the plot of these tests in figures 10.20 and 10.21 that there is a considerable margin of safety against creep buckling in the direction of both the major and the minor column axes.

10.3. Channel Slab Test

One of the channel slabs was picked at random and tested to destruction under centerpoint loading.

The slab was supported at each end and loaded through a 4-in-wide loading beam at midspan. Deflection of the slab was measured at midspan by two 2-in-throw mechanical dial gages.

The test results are shown in figure 10.22. The load at the yield point (2.5 kip) was higher than the predicted load at the yield point of the reinforcement (2.25 kip) using the nominal specified reinforcement yield strength, (40 ksi) and a 4.500 psi concrete strength. The tests on the laboratory structure also indicated satisfactory performance of these components. No material specimens were tested to determine the actual steel and concrete strengths of the floor channels.

10.4. Beam Tests-Repeated Loading

10.4.1. Test Specimens

The test specimens were typical main beam. components with a 22-in-wide and 2-in-thick topping slab cast on each beam. Results are presented for seven beams. Preliminary tests on three other beams are not reported because the test conditions (quarter-point loading) were found to be far too severe in relationship to service conditions.

Beams Nos. 6 through 11 were prepared with column stubs passing through the topping slab and with column connection fixtures in place simulating conditions in the structure except

that tie beams were not attached and grouted to these connections (a block of wood was used as a spacer to fill the void caused by omission of the tie beams). The connections were not grouted. Beam No. 5 did not have the column stub or column fixtures.

In all of the beams tested, the top surface had not been roughened as required by the plans and specifications of the Neal Mitchell Housing System. All specimen preparation, including the placing of the topping slab, was performed by Neal Mitchell Associates. The top surfaces of these beams had been cast against steel forms and were very smooth.

Two types of shear connectors were used in the seven beams. These shear connectors are illustrated in figure 10.23. Beams Nos. 5, 7, and 9 used Star+ inserts spaced 19 in. on centers similar to the shear connectors used in the test structure. Beams Nos. 8 and 10 used Richmond (Kohler) inserts spaced 19 in. on centers. Beams Nos. 6 and 11 used the Richmond1+ inserts spaced 91⁄2 in. on centers. All data on shear connector type and spacing in individual specimens are summarized in table 10.3.

14

10.4.2. Beam Loading

Figure 10.24 is a general view of the test setup. Two 10 kip servo-controlled hydraulic rams applied the load by reacting against a frame bolted to the laboratory tie-down floor. Loading beams under the two rams distributed the test loads. All beams were tested by applying the loads in accordance with the sketch shown in Figure 10.25 and where simply supported by rollers on a clear span of 12.5 ft.

The beams were subjected to 1000 cycles of stress, alternating between intensities corresponding to 1D and 1D + 1L, (for each ram 1D = 2.5 kip, and 1L 2.5 kip; see Appendix C). Subsequently, the beams were tested to failure by 1000-cycle increments with the upper load level being increased at each increment by 0.5L (1.25 kip).

The rate of cyclic loading was 1 cycle per second except for a few cycles at the beginning and end of each increment. During these periods when the rate was 0.01 cycle per second, mechanical dial gage readings were made. During these few cycles, center-span deflection measurements were made using a 5-in-throw mechanical dial gage. In addition, continuous center-span deflection measurements were recorded on a strip chart recorder by using a 3-in linear variable differential transducer (LVDT). Both measuring systems can be seen in figure

14 Proprietary commercial product.

10.24. In an effort to measure the relative horizontal slip between the beam and its topping slab, 0.001-in dial gages were mounted on the beam ends. One of these gages can also be seen in figure 10.24.

10.4.3. Test Results

Graphs reproduced from strip-chart recordings of the midspan deflections are presented as figures 10.25 through 10.28. Table 10.4 shows the midspan deflections measured at the beginning of each increment of loading for each of the beams. This table also indicates the point at which noticeable slippage between the topping slab and the beam occurred, as measured by the slip dial gages installed at the ends of the beam. After testing, the topping slab was removed from each beam, and spacing and condition of anchorage inserts were determined. The data relative to these tests are presented in table 10.5.

10.4.4. Interpretation of Results

A study of table 10.5 and figures 10.25 through 10.28 indicates that all specimens tested showed a similar pattern of failure First a slip occurred between the precast beam and the topping slab. After this initial slip the beams no longer acted monolithically with the slab, and as a consequence the deflections caused by applied load increased. Deflections also increased moderately with the number of load cycles applied during the application of 1D + 1L, 1D + 1.5L and 1D + 2L. During the repeated application of the load of 1D + 2.5L deflections of all specimens tested increased rapidly and some of the specimens failed. All the remaining specimens failed during the first few cycles of application of the load of 1D + 3L.

The initial slip that occurs between the precast beam and the topping slab is caused by horizontal shear. In the structure this shear is resisted by the shear connectors (inserts), the column-beam connection (the column base plate and part of the upper story column bear against the topping slab), and friction between. the precast beam, the topping slab, and the floor channels. In the separate components that were tested not all these elements were present. Shear resisting devices were varied in the tests to determine their effectiveness in preventing slippage.

Beam 5, which had no column stub or columnbeam connection fixtures and had Star inserts at 19 inches on center, experienced slip between the beam and the topping slab during the first cycle of application of the 1D + 1L load (figure 10.25). In terms of ultimate

strength it performed considerably better, resisting 1000 cycles of 1D + 2.5L without failure.

The results of the test on Beam 9 are shown in figure 10.26. This beam, which was similar to Beam 5 except that it did have a partial column-beam connection, performed approximately equal with Beam 5. Beam 9 exhibited signs of first slip during the first cycle of loading to 1D 1L and failed at the 820th cycle of 1D + 2.5L, while Beam 5 failed during the first cycle of 1D + 3L.

In comparison, Beam 7, the companion specimen Beam 9, performed considerably better than either Beam 5 or Beam 9. Beam 7 was able to sustain 1000 cycles of 1D + 1L without any signs of slip. First indications of slip for this beam were observed after the first few cycles of 1D+ 1.5L. Failure occurred at approximately the same point as that of Beam

5.

The results of the test on Beam 8 are shown in figure 10.27. Note that this beam, which had partial column-beam connections and Richmond inserts at 19 inches on center, had about the same initial slip behavior as did Beam 7, which was similar except for type of insert. Beam 8 experienced ultimate failure somewhat earlier in the loading sequence than did Beam 7. Beam 10 (the companion to Beam 8) slipped at about the same point in the loading sequence as did Beam 8, but its ultimate failure took place during the first cycle of 1D + 3L.

The results of the test of Beam 6 are shown in figure 10.28. Beam 6 had a partial columnbeam connection and Richmond inserts spaced at 912 in on center. This beam was able to sustain the full 1000 cycles of loading from 1D to 1D 1L without slippage and went on to sustain about 500 cycles of loading from 1D to 1D 1.5L before slip developed. Beam 6 was able to sustain 1000 cycles of 1D to 1D + 2.5L without failure and finally failed during the seventeenth cycle of 1D+ 3L. Its companion, Beam 11, showed first signs of slip at 300 cycles of 1D to 2L and ultimately failed at 1230 cycles of 1D + 3L.

When the repeated load tests were conceived, it was felt that from the standpoint of slip behavior, the beams should be capable of sustaining 1000 cycles of loading from 1D to 1D +1L, and should be capable of sustaining a loading of at least 1D to 1D + 2L before ultimate failure. This component requirement was set for this particular test, even though it was realized that the performance of the main beam as a separate component does not necessarily simulate the behavior of the complete system. All of the beams tested which had partial column connections, except for Beam 9, satisfied this requirement. The reason that the presence

of the partial connections had no effect on Beam 9 is not clear.

Beam strength was substantially improved at an insert spacing of 912 in, as in Beams 6 and 11.

The beams tested as isolated components experienced considerably larger deflections at 1D

1L than did the center main beam of the test structure (figure 9.2), indicating that the component test was conservative in comparison with the system test.

11. Material Tests

11.1. Introduction

Tests were conducted on the concretes used in the various parts of the structure as well as on the reinforcing steel used in the precast components. The objective of these tests was to determine the relationship between minimum specified properties of materials and the properties of the materials used in the test structure, and to determine material properties which might be useful in analyzing the tests on the main structure and structural components.

11.2. Concrete Tests

The concretes tested were: (1) concrete used in the precast components, except the channel slabs, (2) concrete used in the on-grade floor slab, and (3) concrete used for the topping slab. Concrete specimens were tested for the following: (1) compressive strength, (2) tensile splitting strength, (3) unit weight, (4) air content, and (5) modulus of elasticity.

11.2.1. Precast Component Concrete

The precast components were cast in two days (April 16 and 17, 1968) from five batches of lightweight aggregate concrete. This concrete was made from a g-in maximum size expanded shale aggregate, with preformed foam added at the time of mixing. A rather high cement content (about 9 U.S. bags per cubic yard) was used, and water was added to produce a workable mix. The amount of the preformed foam used was adjusted to provide a concrete with a fresh weight of about 96 lb/ft3 at the mixer. The slump was judged to be about 2 in although it was not measured. The workability of the concrete was excellent with no indication of either segregation or bleeding.

Test specimens (6 x 12 in cylinders) were cast in cardboard molds from four of the five batches. These specimens were shipped in the molds to the test site with the structural com

ponents and were removed from the molds when about 8 days old. They were then stored in the laboratory air until tested.

The components were cast under commercial conditions and no records were available which would permit the association of individual components with particular batches of concrete.

11.2.2. Floor and Topping Slab Concrete

The on-grade floor slab was cast from a 1-in maximum size crushed-stone concrete delivered by a ready-mix truck. The mix proportions and slump are not known. The compressive test specimens were molded in 6 x 12 in cast iron molds which were removed when the concrete was 3 days old. The specimens were then airdried until tested.

The topping slab was cast from a standard 6-bag, 3000 psi lightweight mix delivered by a ready-mix truck in two batches. The first batch was placed in the west section of the topping slab. The coarse aggregate was a 34-in maximum size expanded shale and the fine aggregate was a natural sand. The compressive test specimens were molded in 6 x 12 in cast iron molds which were removed at 2 days of age. The specimens were then air-dried until tested.

11.2.3. Concrete Test Results

The results from the strength tests are shown in table 11.1. The unit weight and air-content determinations are presented in table 11.2. Aircontent determinations were made by ASTM Method C-457 [12]. The values of the modulus of elasticity are shown in table 11.3. By way of comparison, values for an average lightweight aggregate concrete are included in these tables. These values are averages from a total of 46 batches of concrete made from 21 different expanded shale, lightweight aggregates [13]. The average cement content for these concretes was 6.5 bags/yd and the average wet density was 100.3 lb/ft3.

The results indicate that: (1) the compressive strength of the lightweight concrete used in the precast components was well above the design strength of 3500 psi; (2) there was considerable variation in the strengths from batch to batch of the lightweight concrete; (3) there was considerable variation in the unit weights from batch to batch of the lightweight concrete; and (4) there was an apparent increase in the unit weight of the concrete as placed in the precast components when compared to the fresh unit weight at the mixer (about 96 lb/ft3).

These indications justify three conclusions: (1) The concrete strengths in the test structure were significantly higher than the

strengths called for in the plans and specifications of the Neal Mitchell System.

(2) Handling and placing techniques of the fresh concrete can affect the unit weight (and therefore the strength) of the concrete. This is especially true in the case of the high-aircontent concrete used in the precast components. The unit weight of the concrete at the mixer may not necessarily be equal to the unit weight of the concrete in the form.

(3) When working with lightweight concretes, quality control tests on the fresh concrete should be made at the point of placing in such a manner that handling and placing effects can be evaluated.

11.3. Reinforcing Steel

Specimens of the reinforcing steel used in the precast components were tested to determine their yield and ultimate strengths. The results are presented in table 11.4.

12. Summary and Conclusions

12.1. Summary

A full-scale, first-story portion of a building system was tested in the laboratory in a manner that simulated the structural behavior of a three-story building under both service and potential ultimate loading conditions. Additional tests were carried out on components of this building system to determine their behavior and capacity and to provide data needed for the evaluation of the system. Performance criteria for the evaluation of the structural safety and adequacy of certain building systems were developed.

12.2. Conclusions

This series of tests demonstrated that it is feasible and practical to use structural performance tests as a basis for the evaluation of innovative building systems.

All conclusions pertaining to the structural performance of the system in question are based on the test structure as built in the laboratory and on the erection methods and materials used therein. Variation in materials and erection methods may influence perform

ance.

The following significant deviations of the test structure from the plans and specifications of the Neal Mitchell System have been determined:

(1) The test structure had higher than specified concrete strength.

(2) Topping slab thickness exceeded that shown in the plans.

(3) Floor channel reinforcement size was increased.

(4) Gypsum wallboard thickness was less than that shown in the plans.

(5) Greater than ordinary variations in concrete strength and in the dimensions of precast members were observed.

The following conclusions relative to the performance of the building system have been reached:

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

(2) 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.

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

(4) All columns tested as separate components satisfied satisfied component requirements. Column creep tests indicate that the application of service loads over a long period of time is not likely to result in creep-buckling of columns.

(5) Five of six subassemblages consisting of a precast main beam, a section of the topping slab and a partial column-beam connection were able to resist 1000 cycles of repeated loading from dead to dead-plus-live load without exhibiting signs of deterioration. Two such subassemblages with reduced shear connector spacing satisfied this performance requirement by a considerable margin. The column-beam connections appeared to play a major role in shear transfer between the precast beam and the cast-in-place topping slab.

13. Acknowledgment

The contribution of the following persons is acknowledged:

FRANK A. RANKIN and JAMES W. RAINES, Engineering Technicians, were respectively in charge of erection and electronic instrumentation of the test structure.

EARLE F. CARPENTER was the Research Engineer in charge of electronic data processing and participated in the preparation of this report.

THOMAS W. REICHARD, Research Physicist, was in charge of specimen and material testing and participated in the preparation of Chapters 10 and 11 of the report.

ROBERT G. MATHEY, Assistant Chief of the Structures Section, supervised and coordinated the testing effort.

JOHN E. BREEN, Professor of Civil Engineering at the University of Texas critically reviewed the report and made many helpful suggestions.

The team effort of these persons made the successful completion of this project possible.

14. References

[1] Ackleston, A. Y., "Load Tests on a 3 Story Reinforced Concrete Building in Johannisburg, South Africa," Structural Engr. 33, 304-322 [2] American Concrete Institute, Committee 318, Proposed Revision to (ACI 318-63) Building Code Requirements for Reinforced Concrete (Unpublished, Oct. 1955).

[3] State of New York, Building Codes Bureau, State Building Construction Code (Dec. 1964).

[4] Federal Housing Administration Research Paper No. 30 (1954).

[5] American Society for Testing Materials, Philadelphia, Pennsylvania. ASTM Designation: A61-66 "Deformed Rail Steel Bars for Concrete Reinforcement with 60,000 psi Minimum Field Strength" (1966).

[6] American Society for Testing Materials, Philadelphia, Pennsylvania. ASTM Designation: A15–66 "Billet-Steel Bars for Concrete Reinforcement" (1966).

[7] American Society for Testing Materials, Philadelphia, Pennsylvania. ASTM Designation: A185-64 "Welded Steel Fabric for Concrete Reinforcement" (1964).

[8] American Society for Testing Materials, Philadelphia, Pennsylvania. ASTM Designation: C442-67 "Specifications for Gypsum Wallboard" (1967).

[9] American Society for Testing Materials, Philadelphia, Pennsylvania. ASTM Designation: (36-67) "Specifications for Gypsum Wallboard" (1967).

[10] American Society of Testing Materials, Philadelphia, Pennsylvania. ASTM Designation: C39-66 "Test for Compressive Strength of Molded Concrete Cylinders" (1966). [11] American Standards Association Minimum Design Loads in Buildings and Other Structures (Sept. 1955). [12] American Society for Testing Materials, Philadelphia, Pennsylvania. ASTM Designation C457-67T, "Recommended Practice for Microscopic Determination of Void Content, Specific Surface, and Spacing Factor of the Air-Void System in Hardened Concrete "Tentative)" (1967).

[blocks in formation]
« PreviousContinue »