A Caltech Library Service

Analysis of Centrifuge Pile Tests; Simulation of Pile-Driving

Scott, Ronald F. (1980) Analysis of Centrifuge Pile Tests; Simulation of Pile-Driving. California Institute of Technology , Pasadena, CA.

PDF - Submitted Version
See Usage Policy.


Use this Persistent URL to link to this item:


Previous studies on the centrifuge have been directed towards simulating the behavior of a laterally-loaded pile in fine, dry and saturated sand. After data had been obtained on the model pile, attention was turned to modelling the soil-pile interaction behavior. Since the Winkler [continuous reaction elements (springs) distributed along the pile length] foundation representation is the simplest that can be adduced, and, moreover, has been found to give adequate results for design in a variety of foundation problems, attempts were made to extract a Winkler type of function from the model pile test results. The pile response is obtained from the output of a series of strain gauges attached to the pile. In effect these indicate the bending moment in the pile as a function of length along it. As a consequence, to obtain the pile-soil interaction behavior at various locations along the pile, it is necessary to integrate the bending moment function twice for each level of applied load to obtain pile displacements (the top displacement is measured and known), and to differentiate it twice, to get the soil interaction pressure. Then, at a given point on the pile, the pile-soil interaction behavior is given by plotting the pressure versus the displacement at various load levels. A series of such functions at different depths gives the information required for subsequent analyses. The troubles with this procedure are well-known. Double integration is satisfactory and gives a good indication of pile deflections, since the smoothing process eliminates the effect of random errors in the measurement of pile strains. However, double differentiation exaggerates the same errors, and the resulting pressure function can be quite erratic. It is necessary to smooth the strain gauge data first before processing it; various smoothing techniques are available and have been tried. The results of preliminary attempts at obtaining a smoothing function are described in a previous report (3). In the stage of the work reported here, a revised method of analysis was developed, and applied both to calibration tests of the pile, and to the tests carried out in dry and saturated sand. The results of these applications are described below.

Item Type:Report or Paper (Technical Report)
Other Contributors:
ContributionOther Contributors NameIdentifierPersonID (may be blank)
Author of afterword, colophon, etc.Bardet, Jean-PierreBardet-J-PUNSPECIFIED
Additional Information:Report Through September 30, 1979 Research Program for American Petroleum Institute OSAPR Project 13. February 20, 1980, Revised June 20, 1980. The tests described herein were performed by John Lee and the author. John Christenson participated in some of the work. Computer programming assistance was supplied by Albert Chang. Sharon Vedrode typed the report.
Group:Soil Mechanics Laboratory
Funding AgencyGrant Number
American Petroleum InstituteUNSPECIFIED
Record Number:CaltechAUTHORS:20140529-134059571
Persistent URL:
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:45976
Deposited On:29 May 2014 21:59
Last Modified:03 Oct 2019 06:38

Repository Staff Only: item control page