Behavior and Displacement Capacity of Pile and Pile-to-Cap Connections
Yan Xiao and Geoffrey Martin, University of Southern California
Task E1-1 is a joint effort between the University of Southern California (Principal Investigators: Prof. Yan Xiao and Prof. G. Martin) and the University at Buffalo (Principal Investigator: Prof. John B. Mander). The research objective is to investigate the soil-pile interaction effects on the ductility capability of common pile types and the strength and ductility of pile-to-cap connections for short-to-medium span bridges as well as long span bridges. This report summarizes the first year results and describes the second year research plan.
Significant research on seismic behavior and design of bridge structures has been conducted in various research programs, such as the retrofit program sponsored by California Transportation Department (Caltrans) and the Federal Highway Administration (FHWA) Highway Project at the Multidisciplinary Center of Earthquake Engineering Research (MCEER). The primary research focus has been in two areas: pile foundation and bridge substructures. Research has shown that considerable loads from the superstructure must be transferred down though the substructure, the pile cap into the piles, and then finally the soils. This requires that considerable moment and shear be transferred through the pile cap. Existing pile-to-pile cap connections generally have inadequate details to develop their full moment capacity at the connection. It is unclear as to how moment and shear transfer and strength deterioration under cyclic loading will take place during large ground motion. Potential problems which were also noticed (Xiao et al. 1994-1999) but not yet addressed include the anchorage of pile in the pile cap which may not be adequate to develop the full design axial load of the pile.
Under a previous funding sponsored by the FHWA-MCEER Highway Project, a joint research program has been conducted at the State University of New York at Buffalo (SUNY) and the University of Southern California (USC) to study the seismic behavior of bridge pile-to-pile cap with typical details representing current design standards. The tests conducted at SUNY were on pile-to-pile cap connections that is representative of construction in the eastern and central US, whereas the study at USC is on testing the pile-to-pile cap connections representative of construction in the western of US, including California, the major seismic area in the US. At USC, 5 full-scale steel H-pile-to-pile cap connection subassemblies were tested and the test results revealed the following problems: i. Despite a pin connection design assumption, the actual connection can sustain a substantial moment; ii. however, the capacity is unreliable due to the lack of a proper flexure design; iii. the current design details used for the steel HP-pile connection anchor can not develop the ultimate design tensile capacity.
Current research Task E1-1 is an extension of the previous work to investigate the tensile and moment capacities of the connections with other pile types, such as prestressed, precast piles.
Full-scale Pile-to-Pile Cap Subassembly Specimens - Four square PC pile-to-pile cap connection subassemblies are being tested in this program. The specimens were designed to simulate a corner portion of a typical pile footing of the actual prototype bridge in the western US. The piles had a square section of 350mm (14in) by 350mm (14in), as shown in Fig.1. The main experimental parameter is the loading condition. Table 1 summarizes the testing matrix of the four pile-to-pile cap connection subassemblies.
Figure 1. Precast Prestressed Pile Details
Cyclic axial loading
Cyclic lateral loading under constant axial load at 100ton
Cyclic lateral loading under constant axial load at 100ton
Cyclic lateral loading under variable axial loads
As shown in Table 1, the precast prestressed concrete pile footing specimen PCPF-1 is tested under cyclic axial loading only, to investigate the pull-out capacity and behavior of the PC piles. Specimens PCPF-2 and PCPF-3 are subjected to cyclic lateral loading under two different constant axial loads, respectively, to study the seismic behavior of pile-to-pile cap connections. Specimen PCPF-4 is tested under cyclic lateral loading under variable axial loads due to overturning.
Construction of Specimens - The precast presstressed piles were manufactured on a production line of a PC pile manufacturer in Riverside County, California, through a donation arranged by PCMAC (Precast/Prestressed Concrete Manufacturers Association of California). Fig.2 shows the construction process of the specimen piles.
(a) (b) (b) (d)
Figure 2. Construction of Precast Piles
Test Setup - The test setup used for cyclic vertical loading tests and cyclic horizontal loading tests under constant vertical load is shown in Fig.3. The specimen was tested in an upside down position with the pile cap portion post-tensioned to the strong reaction floor beam. The horizontal force was applied to the pile using an actuator with 1,334 kN (300 kip) capacity and push/pull stroke of 229mm (9in). The horizontal actuator was positioned at a distance of 1,067mm away from the pile cap surface, constituting a shear aspect ratio of 3.0 in the pile. The cyclic vertical loading system consisted a vertical actuator with 1,334 kN (300kip) capacity and lever arm system which can amplify the actuator force by 6 times. The vertical loading system was originally developed by the first author for testing of full-scale high-strength concrete columns (Xiao et al., 1998). In the case cyclic vertical loading test, the horizontal actuator was used as the lateral support to the free end of the HP pile. Similar test setup except configuring one of the actuator in an inclined position was used for the test of the pile-to-pile cap connection subassembly specimen under proportionally varied cyclic vertical and horizontal forces.
Figure 3. Test Setup
Instrumentation - Calibrated load cells were used to monitor the applied forces. The horizontal and vertical displacements at the application points of the forces were measured by 500mm stroke linear potentiometers. Four liner potentiometers were mounted to monitor the cap displacements in vertical and horizontal directions. Electrical-resistance strain gauges were mounted on the surfaces of the reinforcing bars at selected positions in pile cap. Strain gauges were also applied on the surface of the HP pile to monitor its strains near the connection.
Loading Program - For the vertical loading tests, the cyclic vertical force was applied following a sequence based on a peak force increment of 222kN. However, after the cap started to rock, the loading for tensile force was based on the uplift displacement control. For the other three tests where a horizontal load was involved, the loading cycle was applied corresponding to a displacement control sequence with a horizontal displacement increment of 0.25% drift ratio, ?/H, until ?/H=1% was reached. Two loading cycles were attempted for each of the drift ratios, ?/H=1%, 1.5%, 2%, 3%, 4%, and 6%.
The current research is planned for a two-year period. During the first year, due to the extra efforts spent for analysis and construction of the PC piles, the actual testing could not finished. Based on current schedule, the planned experimental work and analysis should be completed in the first half of 2001.