Similitude theory is used to develop a scale model for determining the dynamic response properties of a single pile embedded in over-consolidated clay. The basis for the design is a full-scale pile embedded in natural soil, for which dynamic response measurements had been made in previous work. Correlation of the model and prototype results constitutes a major difference in this work over previous efforts using scale models. The model pile material is selected to provide the correctly scaled stiffness and mass properties. The required model soil properties are achieved by developing a mixture of bentonite, aerosil, and veegum. Elastic properties of the model soil are compared with those of the prototype by standard monotonic stress and cyclic stress soil tests. Nonlinearity of the soil stiffness is included in the modeling. Dynamic response of the pile is monitored while excited by impact and swept sine forces at the pile top cap. The results are obtained in terms of time histories for excitation and response at various locations, frequency response functions, natural frequency and mode shapes, and modal damping. Validity of the model is established by comparing the appropriately scaled responses with those of the prototype under similar excitation conditions. It is concluded that the approach should be suitable for measurement of pile/soil dynamic interaction behavior in other types of material and excitation conditions, providing that suitable soil and pile material properties can be selected to allow testing in a one-g environment. Therefore, the scale model approach can be used to verify predictions made by analytical design methods or to provide input parameters for those methods.

A series of lateral dynamic loading tests were conducted on a rigid concrete cap supported by nine 273-mm-diameter steel pipe piles in a square group. The piles were driven 13.7 m into a layered deposit of overconsolidated clay. A linear inertial mass vibrator applied horizontal constant-rate frequency sweep and steady state loadings from frequencies above the first mode resonance to frequencies below the resonance frequency of the pile-soil-cap system at several amplitudes of dynamic load. Sixty channels of system response data were recorded for each test. Frequency response functions between the applied load and points on the pile cap, points on the embedded piles, and locations in the surrounding soil were computed by standard digital signal processing techniques. The average measured pile cap frequency response function peak amplitude was 1.03 × 10−4 mm/N, or about 10 times the static flexibility, and the horizontal resonance frequency was about 7.5 Hz. The measured performance of the group was interpreted in terms of simple modeling techniques.

The design of pile foundations for dynamic lateral loads is significant in the design of both offshore and onshore structures subject to vibratory loads and earthquake motion. Sophisticated numerical models are available for predicting the response of these foundations; however, simple and cost-effective methods are needed to confirm the analysis results and improve the understanding of the overall system behavior. A field testing procedure consisting of slow cyclic load tests and plucking tests is suggested and integrated with simplified single-degree-of-freedom and modal analyses of the test results to produce response functions for dynamic design. The technique is particularly attractive because the field tests may be performed quickly and economically on a test pile or piles installed at the site to determine bearing capacity and/or drivability. The proposed test and analyses procedures were applied to a single pile and nine-pile group installed in stiff clay and tested at the University of Houston pile test facility, and the results are compared favorably with more detailed response information published previously. A good approximation of the pile group dynamic behavior was obtained from the simplified single pile response analysis.

A synthetic clay soil mixture was designed and fabricated to comply with similitude requirements set forth in a dynamic scale-model pile group foundation test. The prototype soil was a stiff, fissured, overconsolidated clay soil at the University of Houston Pile Test Facility. The dynamic response of the model soil was determined through testing by resonant column/torsional simple shear methods and by seismic wave velocity tests in the model test bed. The synthetic clay soil was more nearly linear elastic to higher strain levels with less damping than natural clays. The ability of the soil to regain its strength after disturbance was beneficial to the scale-model tests. The design, fabrication, and testing procedures utilized to obtain an appropriate geotechnical material for scale-model soil/structure interaction tests is detailed in this article.