Abstract
The aim of this paper is to demonstrate that stochastic analyses can be performed on large and complex models within affordable costs. Stochastic analyses offer a much more realistic approach for analysis and design of components and systems although generally computationally demanding. Hence, resorting to efficient approaches and high performance computing is required in order to reduce the execution time. A general purpose software that provides an integration between deterministic solvers (i.e. finite element solvers), efficient algorithms for uncertainty management and high performance computing is presented. The software is intended for a wide range of applications, which includes optimization analysis, life-cycle management, reliability and risk analysis, fatigue and fractures simulation, robust design. The applicability of the proposed tools for practical applications is demonstrated by means of a number of case studies of industrial interest involving detailed models.
| Original language | English |
|---|---|
| Pages (from-to) | 31-48 |
| Number of pages | 18 |
| Journal | Finite Elements in Analysis and Design |
| Volume | 51 |
| DOIs | |
| Publication status | Published - 1 Apr 2012 |
Funding
The financial support of the Translational Research Project L269-N13 of the Austrian Science Foundation (FWF) is deeply appreciated by the authors. The fourth author is a recipient of a DOC-fForte fellowship of the Austrian academy of Science. We gratefully acknowledge M.F. Pellissetti, L. Pichler, M.A. Valdebenito, P. Furegato, G. Schulze and H. Marth for their contributions during the development of the software as described here.
Keywords
- finite element analysis
- general purpose software
- optimization
- reliability
- sensitivity analysis
- stochastic finite element methods
- surrogate model
- uncertainty quantification