Seismic Vulnerability of the Highway System

Task C2-2: Pile Group Effects

Subject Area: Special Bridges - Ground Motion & Geotechnical Studies
Research Year 2

Principal Investigator(s) and Institution(s)

Po Lam, Earth Mechanics, Inc.
Geoffrey R. Martin, University of Southern California


Evaluations of the earthquake response of bridges normally divide the bridge system into two equivalent substructure types, a bridge substructure and a foundation substructure, usually separated at a pile cap or at mudline pile nodes. the foundation substructure requires the development of an appropriate foundation stiffness matrix and foundation forces (reflecting excitation due to ground motion) for use in the bridge substructure dynamic analysis. As described in a prior FHWA MCEER task (Task 106-E-4.10 of Project DTFH61-92-C-00106), substructuring methods may be deployed in the time domain or the frequency domain, or possible via a hybrid frequency-time domain. Each of these methods have unique merits and degrees of difficulty.

Whereas engineers have a good understanding of the pros and cons of these methods and the associated questions of the relative importance of inertial and kinematic interaction in the case of smaller (say 3x3) pile groups, the appropriate application of these methods in the case of very large groups involving hundreds of piles normally associated with long span bridges, remains to be resolved. For this case, ignoring group effects can lead to serious errors, and the use of an elasticity approach is known to be deficient.

By incorporating advances in constitutive modeling and the application of 3-D finite element analysis, this task will establish a practical approach to determine stiffness matrices appropriate for very large pile groups, and examples where the use of the developed approach will be illustrated by application to pile groups representative of long span bridges in both the Western and the Eastern U.S.


In Research Year 1, the following subtasks were accomplished:

  1. To establish a database on representative configurations of large pile groups, a Survey Document was complied in conjunction with a broader long span bridge survey compiled as part of Task C1-1. This survey document has been circulated to selected consultants and State DOT engineers for response.

  2. For the purpose of initial calibration of analytical models, a database was compiled on available static and dynamic lateral loaded pile group tests conducted in the field or in a centrifuge. Typically these tests have been conducted on up to 4x4 pile groups at various pile spacings, to investigate p-y multipliers to be applied to single pile test data.

  3. The development of an analysis methodology initially focused on the use of the finite element computer program DYNAFLOW developed at Princeton University. Preliminary analytical studies investigated the use of a periodic boundary to simulate large pile groups and evaluated the concept of p-y multipliers for establishing stiffness matrices. However, a decision was made to adopt the computer program FLAC as the primary analysis tool due to its better pile and interface modeling capabilities and its widespread commercial use. A copy of the 3-D FLAC code was purchased for use in the project, and preliminary studies have been conducted on its capabilities.

The scope for Research Year 2 is as follows:

  1. Initially, selected test data from the smaller pile group tests, will be used to check and evaluate the non-linear soil constitutive models adopted for FLAC analytical modeling. The calibrated constitutive model will then be used to predict group effects on static stiffness of much large groups using periodic boundary conditions, and to evaluate the concept of the use of p-y multipliers.

  2. The influence of a large buried pile cap on the pile group stiffness using coupled and uncoupled approaches will also be studied using FLAC, as the cap contribution can be significant to the overall group stiffness.

  3. The results of the above analyses will be used to establish design guidelines for the determination of the static stiffness characteristics of large pile groups and an associated equivalent linear stiffness matrix for dynamic response analyses.

  4. In establishing the guidelines, the results of the survey will be used to establish representative configurations for the analyses conducted. It is envisaged that both vertical and battered pile groups with pile caps (representative of older bridges) and pile or shaft extension groups (often typical of new bridges) will be addressed.

The extension of the above research program to address dynamic response issues by applying equivalent linear soil properties in FLAC to evaluate the importance of dynamic group effects such as radiation damping and kinematic interaction, will be deferred until Year 3. The concept of using hybrid models to study these effects where non-linear interface models are combined with elastic for field models would also be part of a Year 3 scope. The consideration of design guidelines for caissons will also be deferred to Year 3.


08/01 Report describing analysis results and design guidelines for determining the stiffness characteristics of large pile groups.

Technical Challenges



return to Task Statement List