Behavior characterization and development of LRFD resistance factors for axially-loaded steel piles in bridge foundations

Abdelsalam, Sherif
Major Professor
Sri Sritharan
Committee Member
Journal Title
Journal ISSN
Volume Title
Research Projects
Organizational Units
Journal Issue
Civil, Construction, and Environmental Engineering

The Federal Highway Administration (FHWA) mandated utilizing the Load and Resistance Factor Design (LRFD) approach for all new bridges initiated in the United States after October 1, 2007. Consequently, significant efforts have been directed by Departments of Transportation (DOTs) in different states towards the development and implementation of the LRFD approach for the design of bridge's deep foundations. The research presented in this thesis is aimed at establishing the LRFD resistance factors for the design of driven pile foundations by accounting for local soil and pile construction practices. Accordingly, regional LRFD resistance factors have been developed for different static analysis methods, incorporating more efficient in-house and combinations of suitable pile design methods, following the AASHTO LRFD calibration framework. Typical calibration framework was advanced in the research presented in this thesis to incorporate the effects of layered soil systems and to reduce the uncertainties associated with soil variation along pile embedment.

To achieve the calibration process successfully, the following three major tasks were accomplished as part of the research presented here: (1) completion of nationwide and statewide surveys of different state DOTs and Iowa county engineers, respectively, to obtain necessary information regarding current pile design and construction practices, the extent of LRFD implementation and regional calibration, as well as to learn of existing local practices; (2) calibration of the LRFD resistance factors for bridge deep foundations, based on the local database (PIle LOad Tests in Iowa [PILOT-IA]), was developed as part of the project and contained data from 82 load-tested steel H-piles, as well as adequate soil profile information; and (3) conduction often full-scale instrumented pile static load tests that cover different local soil regions, accompanied by various soil in-situ tests, including standard penetration test (SPT), cone penetration test (CPT), borehole shear test (BST), and push-in-pressure-cells, in addition to soil laboratory tests with soil classification, 1-D consolidation, CU-Triaxial tests, and direct shear test (DST).

In addition, the AASHTO LRFD calibration framework only addresses pile design at the strength limit state; however, more comprehensive and practical design recommendations should account for the strength and serviceability limit states, simultaneously. For this purpose, two different levels of advanced analysis to characterize the load-displacement response of piles subjected to axial compressive loads were used. The first level of analysis was based on an improved load-transfer method (or t-z model), attained as follows: (a) establishing a modification to the Borehole Shear Test equipment (mBST), that, for the first time, allows for a direct field measurement of the soil-pile interface properties for clay soils; (b) establishing a modification to the Direct Shear Test (mDST), that allows for an accurate and simple laboratory measurement of the soil-pile interface for sands; and (c) adapting a new Pile Tip Resistance (PTR) laboratory test that can measure practically the pile end-bearing properties. The improved t-z analysis uses the measured soil-pile interface properties from the mBST and/or the mDST for different soil layers, and also uses the end-bearing properties of the soil under the pile tip from the PTR laboratory measurements. The t-z analysis showed significantly improved characterization for the pile load-displacement behavior and load distribution along the pile length, compared to field test results. The second level of analysis was based on finite elements (FE), where the Mohr-Coulomb soil constitutive properties were adjusted, using a sensitivity analysis based on various laboratory soil tests, such as the mDST and CU-Triaxial tests. After improving the reliability of the different analytical models in characterizing the behavior of axially-loaded steel piles, a new LRFD displacement-based pile design approach was provided in this thesis, utilizing the improved analytical models.