A finite element simulation approach for glued-laminated timber beams using continuum-damage model and sequentially linear analysis

Joonas Jaaranen*, Gerhard Fink

*Corresponding author for this work

Research output: Contribution to journalArticleScientificpeer-review

2 Citations (Scopus)
105 Downloads (Pure)

Abstract

Glued-laminated timber (GLT) is an efficient and renewable material with potential to enhance sustainability in construction, but its potential has been partly hindered by relatively large uncertainty related to its strength. Previous research indicate that by linking information from the grading process to the GLT fabrication more reliable GLT beams could be made. To this end, this paper represents a finite element based modelling approach combining a continuum damage model with sequentially linear analysis that is suitable for simulation of the bending stiffness and strength of GLT beams. The model has been implemented in Matlab and can be easily integrated into different parametric simulation workflows. Based on validation against 36 experimentally tested GLT beams, the model can predict bending stiffness with high accuracy and provides reasonably accurate predictions for bending strength. The model is also capable of producing qualitatively realistic fracture patterns of GLT beams. The mesh-size study suggests that the best agreement for bending strength with test data is obtained with a sparse mesh, which is discussed in detail. The model is considered suitable for follow-up studies involving the optimization of GLT beam layups.

Original languageEnglish
Article number117679
Number of pages10
JournalEngineering Structures
Volume304
DOIs
Publication statusPublished - 1 Apr 2024
MoE publication typeA1 Journal article-refereed

Keywords

  • Finite element simulation
  • Glued-laminated beams
  • Progressive failure
  • Sequentially linear analysis

Fingerprint

Dive into the research topics of 'A finite element simulation approach for glued-laminated timber beams using continuum-damage model and sequentially linear analysis'. Together they form a unique fingerprint.

Cite this