Chapter I. Introduction

Background

Lateral loading of piles, pile groups, and drilled shafts can be caused by earthquakes, scour, ship impact, and wind, and usually is the driving factor in the design of deep foundations in areas where these hazards might exist. Methods for predicting the capacity of a single pile subjected to lateral loads are quite reliable, and many full-scale pile tests have been conducted in the field to verify and refine the methods. Usually, however, piles are driven and connected in groups for which the single pile design methods are inadequate. Due to great expense, few full-scale lateral load tests have been conducted on pile groups to confirm design parameters and gather data on how pile proximity may affect the load-bearing capacity of the group. This lack of data and of confirmed design methods usually leads to overly conservative designs.

The expense of full-scale tests has given rise to other methods of testing pile groups. Scaled miniature piles have been tested in centrifuges where the effects of overburden can be simulated. Computer software employing finite element or finite difference methods is another tool used to analyze pile group design.

Florida Pier is a 3-D, nonlinear, finite element analysis program developed at the University of Florida under endorsement of the Federal Highway Administration for use in designing piles, pile groups, and drilled shafts subjected to lateral loads. Verification of this software through pile group lateral load tests will give greater confidence to people who currently use this software for pile group design and to those who might use it in the future. This project is part of an ongoing effort to validate the Florida Pier program through testing of model piles.

In Phase 1 of this project, the instrumentation configuration for model piles was developed and tested, and clay soil was consolidated in a vessel where the model piles were tested, and in which further testing could be done. Three model piles were made from 1.52 m (5.0 ft.) lengths of aluminum pipe. Twenty-eight foil strain gages were mounted in 14 diametrically opposed pairs spaced at regular intervals on the inside surface of the pipes. Two static lateral load tests were performed on one of the model piles after it had been calibrated. The loads were applied by hanging weights from a wire rope attached to the top of the pile. Data was gathered by a mechanical multiplexer that switched between all gage channels during sampling, after which the data were converted to digital output by a datalogger and stored in a personal computer. Bending moment and lateral pile displacement versus applied load data gathered during the tests was compared with predictions made using the Florida Pier and COM624P programs, the results of which were quite favorable.

Objectives

Phase 2 built on the accomplishments of Phase 1, namely, the strain gage configuration of the model piles and the soil test vessel, and expanded the scope of the project to include a cyclic lateral load test on a model pile group. The results of this test were then compared to predictions made for group deflection and bending moment using Florida Pier.

The accomplishment of a number of tasks was required in moving from the statically-loaded single pile test of Phase 1 to the cyclically-loaded pile group test of Phase 2. They include the following:

1. Develop test hardware.
2. Design and construct a data acquisition system.
3. Calibrate all instrumentation.
4. Perform a pile group test.
5. Analyze data.

The model pile group tested consisted of five piles arranged in linear configuration and loaded on the group's long axis. Since only three piles had been made for Phase 1, and since they were used mainly as learning tools for instrumentation and calibration procedures, it was decided that six new piles should be constructed, thereby avoiding any compatibility problems between old and new piles. They were constructed based on the pattern set in Phase 1. If the project continues to progress in complexity, then a point might one day be reached where all piles could be implemented in a nine-pile group.

Linking piles together in a group necessitated design of a pile cap. It was decided that pinned connections would reduce the complexity of the whole project, and so a pinned pile cap/load rod was designed and fabricated to act, in conjunction with the piles, as a load cell between each pile while transferring the load through the group. Master load cells also were fabricated for measuring force delivered to the pile group.

Linear Variable Differential Transformers (LVDTs) were used to measure deflection at the pile cap elevation and at a distance of 76.2 mm (3.0 in.) above the pile cap. The higher deflection measurement was needed to continuously calculate the slope of the pile top. Hardware for mounting the LVDTs was specifically constructed for these purposes.

Cyclic loading of the pile group required a more sophisticated loading system than was used in Phase 1. On-hand hydraulic cylinders were central components in the new system, which used regulated air pressure to control load magnitude and a computer controlled solenoid valve to control cylinder actuation. Considerable effort went into developing this system.

The data acquisition system needed for Phase 2 far exceeded capabilities of the Phase 1 system. A new system was designed and constructed by Utah State University students specifically for this project. Emphasis was placed on allowing for a high number of instrumentation channels to be sampled at a rate approaching 10 Hertz. LabVIEW™ software and an analog-to-digital converter circuit board, both products of National Instruments Inc., were key elements in the data acquisition system.

All instrumentation, from piles to the LVDTs, was calibrated using the new data acquisitions system. Calibration factors were calculated for all instruments by comparing measured stresses with theoretical stress values. These calibration factors were then used in post test data analysis.

The pile group test involved a great deal of effort to solve glitches in the loading and data acquisition systems. The testing was successful after ironing out the problems and developing a good test procedure.

MATLAB® software was used for the data reduction process. Output included moment distribution for individual piles and for the pile group, load distribution among the piles, and load versus deflection. This output was expressed in graphical form. Test results for load versus deflection and moment distribution were compared with predictions by Florida Pier as the final step in the project.