Date of Award

12-2015

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Legacy Department

Civil Engineering

Advisor

Juang, Charng Hsein

Committee Member

Atamturktur, Sez

Committee Member

Chen, Qiushi

Committee Member

Pang, Weichiang

Abstract

This dissertation advances the robust geotechnical design methodology by offering improvements, which makes it more efficient and practical for the design of geotechnical systems. Robust geotechnical design (RGD) methodology seeks an optimal design, which is insensitive to, or robust against, the variation in the uncertain input parameters (called “noise factors”) by only adjusting the easy to control parameters (called “design parameters”). The main goal of robust design is to consider safety, cost and robustness simultaneously. Because the cost and the robustness are conflicting objectives, the multi-objective optimization that considers these two objectives while enforcing the safety constraint yields not a single best design but a set of non-dominated designs, which are neither superior nor inferior to one another. These non-dominated solutions form a Pareto front. All these non-dominated designs on the Pareto front are equally optimal in the sense that no improvement can be achieved in one objective without worsening in the other objective. To locate the best compromise between the objectives, knee point concept is often adopted. In this dissertation, the existing RGD methodology, different robustness measures, and different methods for locating the knee point are examined, followed by the development of a new simplified procedure for determination of knee point. The reliability-based RGD approach is also improved in efficiency by coupling the reliability analysis of the system performance and design robustness evaluation. A simplified and efficient procedure is also proposed to implement the RGD optimization procedure in the Microsoft Excel spreadsheet. With all improvements made in this research, the RGD approach can still be computationally challenging for the practicing engineer. In this regards, an efficient and practical RGD procedure using a Microsoft Excel spreadsheet is developed. Because the numerical software programs are often used to evaluate the system response, a response surface model is proposed to approximate the performance functions and integrated into the simplified RGD approach. The significance and practicality of the proposed simplified RGD methodology is illustrated with multiple geotechnical applications, including the design of shallow foundations, rock slopes, drilled shafts, and supported excavations.

Share

COinS