- Title
- Probabilistic calibration of resistance factors for piling designs
- Creator
- Zhang, Yuting
- Relation
- University of Newcastle Research Higher Degree Thesis
- Resource Type
- thesis
- Date
- 2024
- Description
- Research Doctorate - Doctor of Philosophy (PhD)
- Description
- The traditional Allowable Stress Design (ASD) combines all uncertainties from various sources of load and resistance into a single global factor of safety (FS), leading to significant variations in safety levels across similar structures. In contrast, the Load and Resistance Factor Design (LRFD) overcomes this limitation by employing separate resistance and load factors to manage uncertainties in resistance and load respectively. Nonetheless, resistance factors in current LRFD design codes are primarily determined through engineering judgments. Therefore, the primary objective of this thesis is to develop rigorous probabilistic approaches to calibrate resistance factors for piling designs, enhancing the specification of resistance factors in LRFD design codes. The first part of this thesis develops a robust calibration approach to calibrate shaft and base resistance factors utilizing existing databases. Traditional calibration approaches typically overlook the cross-site variability in the statistics of resistance bias factors, potentially leading to the designs based on the calibrated resistance factors violating safety requirements. The proposed approach explicitly considers the cross-site variability, and the calibration process is implemented through a multi-objective optimization, which leads to a Pareto front that describes the trade-off relationship between shaft and base resistance factors and feasible robustness. The optimal shaft and base resistance factors are determined using the minimum distance approach. The proposed approach is applied to calibrate shaft and base resistance factors for three design methods, namely the Vesic, Meyerhof, and Nordlund methods. Recommendations for shaft and base resistance factors are provided as a function of design methods, the ratio of shaft and base resistances, and levels of robustness. The second part of the thesis proposes a Bayesian calibration approach for recalibrating resistance factors of single piles when site-specific load tests are performed. Initially, prior information on resistance bias factors is derived from existing databases. Subsequently, Bayes’ theorem is utilized to update the resistance bias factors based on load test results. Reliability methods like the First Order Reliability Method (FORM) and Monte Carlo simulation (MCS) are then employed to recalibrate resistance factors. The proposed approach is applied to load test data collected from the literature. Recommendations for resistance factors are provided as a function of the number of load tests and corresponding test results for various ground conditions, design methods, target reliability indices and within-site variabilities. The third part of this thesis develops two approaches to directly calibrate resistance factors of pile groups based on individual pile load tests. These approaches enable the consideration of complex pile-soil-pile interaction and the correlations among individual piles, which are inherent influenced by the spatial variability of soils. The first approach employs Bayes' theorem and random finite difference (RFD) analysis, where the RFD analysis is employed to evaluate the correlation between resistance bias factors of individual piles in spatially variable soils. The resultant correlation matrix is subsequentially employed in Bayes’ theorem to update resistance bias factors using individual pile load test results and their corresponding test locations. The updated resistance bias factors are then used for the direct calibration of resistance factors for pile groups within the framework of LRFD. The second approach introduces a surrogate model based on convolutional neural network (CNN) to replace traditional RFD analyses. Subsequently, the CNN model is used to derive pile resistances in spatially variable soils. The resistance factors are then calibrated by counting and conditional probability based on the outcomes of load test results. This thesis systematically addresses the calibration of resistance factors for single piles and pile groups, with and without site-specific load tests, providing substantial contributions to the development of LRFD design codes.
- Subject
- piling designs; Allowable Stress Design (ASD); calibration of resistance factors; LRFD design codes; thesis by publication
- Identifier
- http://hdl.handle.net/1959.13/1514829
- Identifier
- uon:56858
- Rights
- This thesis is currently under embargo and will be available from 28.11.2025, Copyright 2024 Yuting Zhang
- Language
- eng
- Hits: 169
- Visitors: 161
- Downloads: 3
| Thumbnail | File | Description | Size | Format |
|---|