Reliability-Based Optimization for Multidisciplinary System Design

ETD
Add to Collection

Abstract

Reliability-Based Optimization (RBO) for engineering design deals mainly with twodesign attributes, namely the merit, for example cost, and the reliability of the design. Inthis work the class of design problems which are considered, are designs characterizedby a minimum merit function and that satisfy certain reliability constraints. The reliabilityconstraints are typically constraints on the probabilities of failure due to componentfailure events or a system failure event. These are obtained using standard reliability analysistechniques such as First Order Reliability Method (FORM), Second Order ReliabilityMethods (SORM) and Monte Carlo Simulation (MCS) techniques.The reliability analysis and RBO are very expensive for multidisciplinary systemsconsisting of various disciplines that are dependent on each other or coupled, for example,an aeroelastic structure. Hence, the primary goal of the research is to developefficient methodologies that perform RBO for multidisciplinary systems. The methodologiesconsidered incorporates a Concurrent Subspace Optimization technique that allowsconcurrent design optimization in each discipline. The methodologies also incorporateapproximation concepts to reduce the computational costs. There are essentially twomethodologies, one that uses a traditional reliability analysis method and the other thatuses a new reliability analysis method geared towards reduction of computational expensesfor coupled multidisciplinary problems. A new reliability analysis tool based on Trust Region methods was developed for the latter case. Both methodologies were appliedto multidisciplinary test problems and about 20%-30% computational savings wereobserved.A second goal of the research was to investigate the use of Monte Carlo Simulation(MCS) techniques for reliability analysis in RBO, that are more accurate but more expensivethan FORM or SORM. In this work, conditional expectation MCS was selectedover indicator-based MCS techniques based on smoothness criteria and the availabilityof analytic sensitivities. A MCS-based RBO methodology was developed and successfullyimplemented to problems with both component and series failure events. It wasobserved that designs with significantly lower merit functions were obtained for the applicationproblems considered, compared to a FORM-based RBO approach. It was alsoobserved that the computational costs were extremely high for one of the application problems.Some suggestions for future research are made regarding development of efficientmethodologies for the MCS-based RBO.

Attributes

Attribute NameValues
URN
  • etd-07022003-191128

Author Dhanesh Padmanabhan
Advisor Michael M. Stanisic
Contributor Michael M. Stanisic, Committee Member
Contributor Timothy C. Ovaert, Committee Member
Contributor John E. Renaud, Committee Member
Contributor Stephen M. Batill, Committee Member
Degree Level 2
Degree Discipline Aerospace and Mechanical Engineering
Degree Name Doctor of Philosophy
Defense Date
  • 2003-06-18

Submission Date 2003-07-02
Country
  • United States of America

Subject
  • Multidisciplinary Design Optimization

  • Structural Reliability

  • Reliability-Based Optimization

Publisher
  • University of Notre Dame

Language
  • English

Access Rights Open Access
Content License
  • All rights reserved

Files