Development of a New Load Analysis Tool and a Unified p-y Method for Lateral Analysis of Large-Diameter Drilled Shafts Validated Using Load Tests
Loading...
Authors
Bhuiyan, Fahim Mashroor
Issue Date
2022
Type
Dissertation
Language
Keywords
Cemented Soil , Deep Foundation , Large Diameter , Load Test , p-y Analysis , t-z Analysis
Alternative Title
Abstract
The drilled shaft foundation is often considered a reliable cost-effective solution in major engineering projects around the globe. The construction of a single drilled shaft can often replace several monopiles with smaller diameters while providing high load-bearing capacity and resistance to seismic loading. Recent literature indicates the use of larger diameter drilled shafts, particularly to support offshore wind turbines, bridge piers, abutments, and other heavy infrastructures. Performing field load tests to validate the design of drilled shaft is a common engineering practice. The improvement in computational technology and high-quality soil investigation technique often allows more accuracy in predicted responses by adopting numerical simulations of field load tests. The beam on nonlinear Winkler foundation (BNWF) model, also known as the p-y method, is a widely accepted tool to perform numerical lateral load analysis. The p-y method is known for its simplicity and reliability due to its earlier application to perform preliminary design calculations. The major goal of this research is the improvement of conventional p-y analysis in the context of a larger diameter shaft. Based on several past research, it is evident that the conventional p-y method implemented in commercial programs neglects the contribution from lateral resistance components more prominent in the case of larger diameter shafts. The conservatism due to the larger diameter of the drilled shaft in p-y analysis has been termed as ‘diameter effect’ in past research.In this study, a unified p-y analysis is proposed, by including the resisting moment due to side shear, tip shear, and tip moment resistances in the conventionalBNWF model. A simplified tip moment resistance model applicable for any type of soil material, and to be used as part of the unified p-y spring model is also proposed in this study. To apply and evaluate the proposed unified p-y analysis, a MATLAB-based, finite-difference program, NVShaft, has been developed as an integral part of this research. During the initial development stage, NVShaft was verified by comparing the predicted responses with classical p-y solutions and responses obtained from available commercial programs. To validate the new program, available literature on lateral load tests on larger diameter shafts was explored. Evaluation of the proposed unified p-y analysis has been carried out for different subsurface conditions. The study puts more emphasis on investigating the diameter effect in the local cemented soil condition of Nevada. This was done by simulating field lateral load tests on 8 ft and 2 ft diameter drilled shafts from the I-15/US 95 load test program in Las Vegas, Nevada. The engineering challenges in site investigation and proper characterization of caliche material in numerical modeling have been addressed. For the 8 ft diameter shaft, the addition of resisting moment due to side shear resulted in a maximum of 28.2% reduction in shaft head deflection compared to conventional p-y predicted response. As both shafts exhibited flexible lateral response, mobilization of tip resistance was negligible. In another example, evaluation of the unified p-y analysis was conducted by simulating two, 2 m diameter rigid piles in sandy, and clayey site conditions from the PISA load test program. Performing unified p-y analysis in NVShaft resulted in 41.5% and 5.1% reduction in ground-level deflection in clayey and sandy sites, respectively, in this case.NVShaft is also capable of performing numerical axial load (t-z) analysis, along with a unique feature to simulate bi-directional axial load test. The validation example of the t-z capability of NVShaft is presented by simulating axial load tests from I-15/US 95 and Las Vegas City Center projects. The use of t-z models for soft rock and Florida limestone to model axial resistance of caliche material is also presented in this context. It was observed that the t-z model for Florida limestone yielded higher capacity, and stiffer response, compared to the t-z model for soft rock material. To understand the effect of soil strength parameters, embedment depths, and axial load on different additional lateral resistance components with increasingshaft diameter, further investigation was done in the form of a parametric study. The numerical models constructed from original test conditions in the sandy and clayey sites from the PISA load test program were used as reference models in this context. Similar to the original lateral load test simulations, the diameter effect seems to be more apparent in the clayey soil, compared to the sandy soil from the PISA load test program. It was found that shorter embedment depth and larger axial load significantly increase the tip resistances in the p-y model. In an attempt to improve lateral load analysis in the local soil condition of Las Vegas, a semi-empirical p-y model for caliche is proposed. No site-specific p-y mode for caliche has been developed at the time of this study. The available laboratory stress-strain data of caliche has been converted to the proposed caliche p-y model by implementing a simple scaling formula. The applicability of the proposed p-y model has been evaluated by simulating four lateral load tests from the Raiders Stadium project in Las Vegas. The NVShaft predicted response by using the proposed p-y model was compared with the measured data, as well as predictions obtained from using similar rock p-y models. The proposed p-y model is expected to serve as baseline for more sophisticated p-y model for caliche in near future.
Description
Citation
Publisher
License
Creative Commons Attribution 4.0 United States