KLFS : KarhunenLoeveFieldSensitivity

This repository contains codes used for sensitity analysis of models where input uncertainties are modeled by stochastic processes and scalar distributions. These codes are meant to be used in parallel with the openTURNS library. They were done in the frame of an internship at PHIMECA in Clermont-Ferrand.

Aim of the project

Sensitivity analysis is the study of how the uncertainty in the output of a mathematical model or system (numerical or otherwise) can be divided and allocated to different sources of uncertainty in its inputs. A related practice is uncertainty analysis, which has a greater focus on uncertainty quantification and propagation of uncertainty; ideally, uncertainty and sensitivity analysis should be run in tandem.

Sensitivity analysis itself is, although being almost 30 years old, a rather new discipline with it's limitations. In fact, sensitivity analysis is usually carried out on scalar uncertainties and not so often on more complex random structures, as random fields. The aim of this project is to develop a set of usable codes, that will allow to carry out sensitivity analysis on models where the input uncertainty is not only scalar, but can also occur under the form of random fields.

The methodology was based on different research papers, with the principal idea gotten from a 2017 paper called "Time-variant global reliability sensitivity analysis of structures with both input random variables and stochastic processes" from P. Wei, Y. Wang & C. Tang. Paper can be found here.

Getting Started

To use the codes, the only pre-requisite is openTURNS, and python 3.x.

Written with :

• openTURNS - An Open source initiative for the Treatment of Uncertainties, Risks'N Statistics.
• anaStruct - Analyse 2D Frames and trusses for slender structures. Determine the bending moments, shear forces, axial forces and displacements.

Authors

• Kristof S. - Initial work

Acknowledgments

• A lot of thanks to Ritchie Vink and the superb anastruct library : https://www.ritchievink.com/
• Also a lot of thanks to PHIMECA and the team working on openTURNS, for their really efficient sensitivity analysis library