UNIVERSITY OF HAWAII
DEPARTMENT OF CIVIL AND ENVIRONMENTAL ENGINEERING
Seismic
Resistance of Slab-Column Connections in
Existing Non-Ductile Flat-Plate Buildings
BACKGROUND INFORMATION:
Name Yong
Du
Phone
N/A
Email
N/A
Name Ahmad
Durrani
Rice University
Phone (713)
348-5603 Civil
Engineering Department
(713)
348-4949
6100 Main Street
Email durrani@rice.edu Houston,
TX 77005-1892
SUMMARY
General
An experimental investigation was conducted to evaluate the seismic resistance of slab-column connections in existing non-ductile flat-plate buildings. The test subassemblies were designed and detailed in accordance with the building codes of the late forties and fifties. Each subassembly consisted of two exterior and one interior connection and was subjected to several cycles of increasing lateral displacements. The test variables included the slab reinforcing detail, intensity of the gravity load applied to the slab, and the presence of a spandrel beam.
General Testing Program
Prototype Building
The prototype structure for this investigation was a five story flat-plate building with three bays in the short direction and four bays in the long direction. This configuration of the prototype building was chosen to facilitate comparison of results with small-scale shake-table tests on slab-column frame systems in other National Center for Earthquake Engineering Research (NCEER) supported projects. The columns were typically 20 in. x 20 in. in cross section and were spaced 20 ft. apart with each story of 10 ft. height. The compression strength of concrete was chosen as 3000 psi with reinforcement of Grade 40 steel. These material properties are typical of older buildings.
Specimen Design
The test specimens were half-scale slab-column connection subassemblies, each consisting of two exterior connections and one interior connection. Since the study focused on the slab response in the connection region, the columns were designed to remain elastic with reinforcement consisting of six No. 7 Grade 60 bars. The gravity load applied to the subassemblies was adjusted such that the resulting shear in the connection region was approximately the same as in the prototype building. The reinforcement ratio of the slab top reinforcement in the column strip for both exterior and interior connection was 0.59%. The reinforcement ratio for the slab bottom reinforcement in the column strip of both interior and exterior connections was 0.22% for bent-up detail and 0.3% for the straight bar detail. Furthermore, the slab bottom reinforcement in the column strip was 37.6% of the slab top reinforcement for the bent-up bar configuration and 63% of the slab top reinforcement in the straight bar detail. The specimens were constructed with ready-mixed concrete of 3000 psi specified compressive strength.
For a true half-scale modeling of the chosen prototype, the specimens would have a span of 10 ft. and slab width of 9 ft. Due to constrains of the testing frame, these dimensions were reduced to 9.5 ft. and 6.5 ft., respectively. The columns in the test subassemblies were terminated at inflection points which were assumed stationary at mid-height of the story above and below the slab. Based on the intended scope of the study and typical construction details of the existing flat-slab buildings, three variables were studied during this investigation. These included configuration of the slab reinforcement in the connection region, presence of the spandrel beams at exterior connections, and intensity of the gravity load on the slab at the time of lateral cyclic loading. Further details of the variables and material properties are give in table 1.
|
Specimen |
Reinf. Detail |
Gravity Load |
Concrete Strength (psi) |
Steel Strength (ksi) |
|
Bent-Up |
DL+0.3LL |
5115 |
54 |
|
|
Bent-Up |
DL+LL |
3731 |
54 |
|
|
Straight |
DL+0.3LL |
3566 |
54 |
|
|
Edge Beam |
DL+0.3LL |
2772 |
54 |
|
Prototype Design and Subassembly Design
The design of test specimens was based on a five story prototype flat-plate building shown in figure (A-1). It has four bays of 20 ft. each in the long direction and three bays in the short direction. The story height is typically 10 ft. except for the first story which is 21.5 ft. high. The design of this building is based on Building Code Requirements for Reinforced Concrete (ACI 318-47) and it represents a typical gravity load design for flat-plate buildings constructed during forties and fifties in the Eastern United States.
The prototype building was assumed as an office building with a live load of 50 lbs/ft2. The total gravity load on the floors, including the slab weight, was estimated as 145.5 lbs/ft2. The slab thickness in this case was governed by shear considerations which was chosen as 9 inches. Moments in the frame under gravity loads were determined using the direct design method and the slab reinforcement in the column and middle strips was selected based on the working stress design procedure.
Material Properties
Prototype Building
Concrete compressive strength,
fc': 3000 psi
Reinforcing Steel, Grade 40, fy: 40,000 psi
Test Setup
The slab-column connection subassemblies were tested in a steel reaction frame. The top of columns were all connected to a rigid beam through load cells with the bottom of each column attached to the reaction frame with a hinge connection. The lateral cyclic load was applied to the specimen through the distribution beam with a servo-controlled closed-loop hydraulic actuator. The shear in each column was independently measured along with the vertical reaction at the center column which rendered the subassembly statically determinate.
In three of the test specimens, the gravity load applied to the slab represented service load condition and it consisted of design dead load plus thirty percent of the live load distributed uniformly over the slab surface. Full dead and live load was added in one specimen to study the effect of high gravity shear on the drift response of connections. A part of the gravity load, corresponding to the service load condition, was applied to the slab by hanging dead weights from cables anchored on the top of the slab surface. Additional load required for full dead plus live load was provided by stacking lead weights on the slab.
Each specimen was subjected to approximately twenty reversed displacement cycles. The displacement cycles at drifts of 1.5%, 2%, and 4.5% were repeated to estimate the loss of strength and energy dissipation capacity. A number of small amplitude cycles of 1.0% drift were introduced in the routine to estimate degradation of low amplitude stiffness. Furthermore, the specimens were subjected to at least four cycles of drift less than one percent at the beginning of the test to study their elastic response under lateral loading.
Instrumentation
Each specimen was extensively instrumented to measure various response parameters of the subassemblies. Load cells at top and bottom of the columns enabled determination of all reactions which provided all moments and shears in the connection regions. LVDT's (Linear Variable Differential Transducers) were attached to the top and bottom of the slab adjacent to the columns to measure the slab rotation relative to the column along the loading direction. Yielding of the slab reinforcement and the loss of anchorage of reinforcement was determined from strain gages attached to slab bars at suitable locations. Targets were attached to the slab surface on a grid pattern which could be read using a level for determining deformed shape of the slab at various stages of the test. All of these instruments were continuously monitored using an automated data acquisition and control system. The formation and development of cracks and deflection measurements were taken while the test was briefly paused at regular intervals for visual observations.
General Results
Rapid stiffness degradation, significant reduction in deformation capacity under increased gravity load, and limited moment-transfer capacity of connections were observed to be the main response characteristics of the non-ductile slab-column connections subjected to earthquake-type loading. The connections were able to sustain full design dead and live loads through at least 2% drift. However, 70% of the initial lateral stiffness of the connections was lost by this drift level. Under normal service loads, the connections were able to maintain at least 80% of their strength through approximately 4% lateral drift and the transfer of unbalanced moment occurred mainly on the column face with slab under negative bending.
The mode of failure and deformation capacity of the connections was observed to depend greatly on intensity of the gravity load on the slab. At lower gravity loads, the response of connections was dominated by flexural yielding of the slab. As the gravity load was increased, the interior connections punched at significantly smaller lateral drift. Provided the gravity shear is kept within a certain maximum limit, the test results suggest that the rapid degradation of the lateral stiffness and the lack of protection against progressive collapse resulting from punching of connections may be the two most critical factors affecting the response of non-ductile flat-plate buildings during moderate earthquakes.
For details on the properties and test results for each specimen, click on the specimen below.
Stiffness degradation plots were generated along with the hysteretic plots. Please refer to the web pages of the individual specimens for the downloadable EXCEL 97 version files.
|
Designation |
Description of Reinforcement and Gravity Load |
|
Bent-Up Reinforcement/ (DL + 0.3LL) |
|
|
Bent-Up Reinforcement/ (DL + LL) |
|
|
Straight Reinforcement/ (DL + 0.3LL) |
|
|
Edge Beam Reinforcement/ (DL + 0.3LL) |
