Precast Bent System for High Seismic Regions: Laboratory Tests of Column-to-Footing Socket Connections
CHAPTER 3. EXPERIMENTAL PROGRAM
The three specimens were tested using the self-reacting loading assembly shown in figure 11. The specimen was placed in the loading assembly, leveled with shims, and attached to the underlying footing with Hydrostone. To prevent possible overturning, four threaded 1.25-inch-diameter Williams bars were placed in plastic tubes through the footing and threaded into nuts cast into the underlying concrete block. The top nuts on these bolts were left with a clearance of 1/16 inches to prevent them from providing resisting forces until the base started to overturn. This gap was introduced because the specimen had been designed to simulate a spread footing in the field, in which no such holding down force could exist. The bolts were used only as a backup system to prevent overturning in the event that the column proved unexpectedly strong, and the lateral strength was controlled by overturning rather than the strength in one of the structural components.
The horizontal loads were applied to the column with a servo-controlled actuator with a capacity of 220 kips. The other end of the actuator was attached to a steel frame with two vertical W24x94 beams, which were stiffened on each side with a diagonal HSS 6x6x3/8. Reactions were provided by a large concrete base on which the footing rested. The axial force on the column was provided by the laboratory's 2.4-million-lb Baldwin Universal Testing Machine. The force was transferred through a spherical bearing that slid against the Baldwin head, guided by a channel. To minimize the friction across this interface, stainless steel plates were placed in the channel and sheets of greased polytetrafluoroethylene (PTFE) were glued to the top and the sides of the bearing.
The response of the specimen was monitored with load and displacement transducers on the exterior of the column and the footing, and by internal strain gauges. The strain gauges were applied to selected vertical bars and spiral turns in the column, and to flexural bars, diagonal shear friction bars, and vertical stirrups in the footing.
Load cells in the Baldwin Universal Testing Machine and the MTS actuator captured the vertical and horizontal loads applied to the specimen. Load cells were also placed between the top surface of the footing and the upper nuts on the Williams bars to detect any forces that restrained overturning.
As shown in figure 12, a total of 25 linear potentiometers were used to capture deformations of the specimens. Potentiometers No. 1 to 8 were used to determine relative rotations of the column at various heights by measuring relative displacements of aluminum flat-bars that were mounted on threaded rods protruding from the columns. These rods were located 1.75, 6.75, 11.75, and 18 inches above the footing surface.
A linear variable differential transformer (LVDT) built into the actuator monitored its displacements, but these displacements are not reported here, because those measurements included both the column displacement and the displacement of the steel reaction frame. Instead, five string potentiometers (No. 9 to 13) measured horizontal displacements of the column relative to an independent, unloaded, instrumentation reference frame. The first four of them were attached to the threaded curvature rods, and the fifth one was attached at the point where the horizontal load was applied.
Two potentiometers (No. 14 and 15) were to measure potential opening of the column splice. Eight potentiometers (No. 16 to 23) were used to measure any uplift and slip of the specimen and the testing rig. One LVDT (No. 24) was used to measure any vertical slip between column and footing, and another (No. 25) measured deflections of the steel reaction beam where the actuator was attached. Four inclinometers (No. 26 to 29) were placed at 10, 18, 30, and 40 inches up the east side of the column, and they measured the angle of the column at those heights.
As shown in figures 13 and 14, key reinforcing bars were strain gauged. All gauges were supplied by Texas Instruments. YFLA-5-5L strain gauges were used with some exceptions because of stock availability, in which case FLA-5-3LT gauges were used. Table 1 summarizes the quantity of strain gauges, the types used in the specimens, and their location.
(a) Location of gauges in specimens SF-1 and SF-2
(b) Location of gauges in Specimen SF-3
(a) Location of gauges in specimens SF-1 and SF-2
(b) Location of gauges in Specimens SF-3
|Location||Type||Quantity of each in SF-1||Quantity of each in SF-2||Quantity of each in SF-3|
The two column bars (N-NE and S-SW, see the nomenclature for the bars in figure 13) that were expected to experience the largest strains were equipped with strain gauges. Specimens SF-1 and SF-2 each bar had 10 gauges, placed in pairs at 5 locations. Three pairs of gauges were placed below the top surface of the footing to check the development of the bars and the effectiveness of the bar terminators. The other two were located at the footing surface to determine the peak bar strain and 18 inches above the surface to check development near the precast column splice. Specimen SF-3 had a shallower foundation and no column splice; therefore, each bar had only four gauges placed in pairs at the footing surface and near the bar terminators.
Selected bars in the footing were strain gauged as well. Because the footing was symmetric about two axes, gauges were placed only on half of the north side bottom bars, half of the diagonal bars, and selected footing ties.
Data were recorded with sampling rate of 0.2 seconds using LabVIEW from National Instruments.
All three specimens were tested following the same testing protocol. In the first stage of testing, the column was subjected to pure axial load. The target load of 240 kips consisted of the scaled-down factored dead and live loads, according to the AASHTO LRFD. The specified load rate was 8 kips per second. Once the target load was reached, it was held for 5 minutes, and the connection was inspected visually.
The axial load was then reduced to an unfactored dead load value of 159 kips, and the specimen was subjected to displacement-controlled cycles. The same displacement history was applied as in previous research on rapid construction.(8,11) This load history is a modification of the loading history for precast structural walls recommended in the National Earthquake Hazards Reduction Program (NEHRP).(20)
The target displacement history is provided in figure 15 and table 2. The history consisted of sets of four cycles in which the peak amplitude was 1.2A, 1.4A, 1.4A, and 0.33A respectively, where A is the peak amplitude from the previous cycle set. The small amplitude cycle was intended to evaluate the residual small-displacement stiffness in the column after the set's peak amplitude. In all cycles, the actuator moved from zero displacement to the peak cycle displacement in 20 seconds.
In this report, the positive and negative peak displacements within a cycle are referred to as "peak" and "valley." Positive displacements occurred when the actuator was in tension, pulling towards the south in the laboratory.
After cyclic testing of SF-1 and SF-2, the columns were loaded again with pure axial load until the column failed. This post-test axial loading was not performed for SF-3, because it was damaged too much during the lateral-load testing.
|Set||Cycle||Drift Ratio (%)||Displ. (in.)||Set||Cycle||Drift Ratio (%)||Displ. (in.)|