Toronto Metropolitan University
Browse

Improving the Prediction for the Ultimate Axial Capacity of Driven Piles in Ontario Soils with a Genetic Algorithm

Download (3.19 MB)
thesis
posted on 2023-07-04, 14:16 authored by Markus Jesswein
A genetic algorithm (GA) was developed to improve predictions for the ultimate axial capacity of driven piles in Ontario soils. Challenges arise to accurately predict the ultimate capacity due to many influential factors, such as the ground conditions, installation method, and pile geometry. A total of 43 piles (H or pipe piles) were collected from the Ministry of Transportation of Ontario. Side and tip resistances were extracted from piles subjected to extension and compression load tests. The soil measurements and pile resistances were regressed with a statistical analysis and GA, and the developed relationships were compared to existing design methods. On average, existing design methods overestimated the capacity by a factor of 1.16 to 3.00. The proposed correlations were slightly conservative with the capacity but provided errors within ± 30 % of the measured side resistance. The new design methods from the GA offer substantial improvements for pile design

History

Language

English

Degree

  • Master of Applied Science

Program

  • Civil Engineering

Granting Institution

Ryerson University

LAC Thesis Type

  • Thesis

Year

2018