Probabilistic sensitivity analysis to understand the influence of micromechanical properties of wood on its macroscopic response



Hristov, PO ORCID: 0000-0002-3302-686X, Diaz De La O, FA, Saavedra Flores, EI, Guzmán, CF and Farooq, U
(2017) Probabilistic sensitivity analysis to understand the influence of micromechanical properties of wood on its macroscopic response. Composite Structures, 181. pp. 229-239.

This is the latest version of this item.

[img] Text
SA_of_wood_FFS_black.pdf - Author Accepted Manuscript

Download (939kB)

Abstract

This paper investigates the influence of the uncertainty in different micromechanical properties on the variability of the macroscopic response of cross-laminated timber plates, by means of a probabilistic sensitivity analysis. Cross-laminated timber plates can be modelled using a multiscale finite element approach which although suitable, suffers from high computational cost. Investigating parametric importance can incur considerable time penalty since conventional sensitivity analysis relies on a large number of code evaluations to produce accurate results. In order to address this issue, we build a statistical approximation to the code output and use it to perform sensitivity analysis. We investigate the effect of a collection of parameters on the density and Young’s moduli of wood. Additionally, the influence on the response of cross-laminated timber plates subject to bending, in-plane shear and compression loads is investigated due to its relevance within the engineering community. The presented results provide a practical insight into the importance of each micromechanical parameter, which allows research effort to be focused on important wood properties.

Item Type: Article
Uncontrolled Keywords: Probabilistic sensitivity analysis, Gaussian process emulation, Cross-laminated timber, Multi-scale analysis, Finite elements
Depositing User: Symplectic Admin
Date Deposited: 05 Dec 2017 08:56
Last Modified: 19 Jan 2023 06:48
DOI: 10.1016/j.compstruct.2017.08.105
Related URLs:
URI: https://livrepository.liverpool.ac.uk/id/eprint/3013514

Available Versions of this Item