Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Experimental and Computational Analysis of Mode I Fracture in Wood : Likelihood-based inference and finite element model updating

Jonasson, Johannes LU (2025)
Abstract
This thesis covers the estimation of the fracture behaviour of the European wood
species Norway spruce and birch. The fracture behavior of wood can be quantified
through three material parameters: the stiffness, the strength, and the specific fracture
energy. In the present work, these parameters are characterized through two methods.
The first method is purely experimental and only estimates one parameter, by use
of the so-called work-of-fracture method. In this procedure, a single-edge-notched
beam is loaded in three-point-bending under stable crack propagation until complete
failure. The specific fracture energy, i.e. the energy dissipated during the creation
of a unit crack surface, can then be evaluated... (More)
This thesis covers the estimation of the fracture behaviour of the European wood
species Norway spruce and birch. The fracture behavior of wood can be quantified
through three material parameters: the stiffness, the strength, and the specific fracture
energy. In the present work, these parameters are characterized through two methods.
The first method is purely experimental and only estimates one parameter, by use
of the so-called work-of-fracture method. In this procedure, a single-edge-notched
beam is loaded in three-point-bending under stable crack propagation until complete
failure. The specific fracture energy, i.e. the energy dissipated during the creation
of a unit crack surface, can then be evaluated as the work carried out to complete
fracture, divided by the fractured area. This method has previously been used in
a large number of studies to establish the specific fracture energy of various wood
species. In the present work, this experimental procedure is used to test 80 specimens,
divided into four different series.
The second method employed is based on the same experimental tests, but in combination with numerical models. The difference between the numerically and experimentally obtained force-displacement responses is quantified in a so-called cost function,
which further is minimized through a finite element model updating (FEMU) procedure using a Bayesian optimization framework, to establish the optimal set of material
parameters. In addition, two commonly employed cost functions employed in FEMU
are reformulated in the context of likelihood-based inference to estimate the uncertainty in the optimal set of parameters.
It is shown that neither of the two commonly employed cost functions can recover the
variance in the experimental load-displacement curves. However, the mean behaviour
is captured fairly well for both cost functions. (Less)
Abstract (Swedish)
This thesis covers the estimation of the fracture behaviour of the European wood species Norway spruce and birch. The fracture behavior of wood can be quantified through three material parameters: the stiffness, the strength, and the specific fracture energy. In the present work, these parameters are characterized through two methods.

The first method is purely experimental and only estimates one parameter, by use of the so-called work-of-fracture method. In this procedure, a single-edge-notched beam is loaded in three-point-bending under stable crack propagation until complete failure. The specific fracture energy, i.e. the energy dissipated during the creation of a unit crack surface, can then be evaluated as the work carried... (More)
This thesis covers the estimation of the fracture behaviour of the European wood species Norway spruce and birch. The fracture behavior of wood can be quantified through three material parameters: the stiffness, the strength, and the specific fracture energy. In the present work, these parameters are characterized through two methods.

The first method is purely experimental and only estimates one parameter, by use of the so-called work-of-fracture method. In this procedure, a single-edge-notched beam is loaded in three-point-bending under stable crack propagation until complete failure. The specific fracture energy, i.e. the energy dissipated during the creation of a unit crack surface, can then be evaluated as the work carried out to complete fracture, divided by the fractured area. This method has previously been used in a large number of studies to establish the specific fracture energy of various wood species. In the present work, this experimental procedure is used to test 80 specimens, divided into four different series.

The second method employed is based on the same experimental tests, but in combination with numerical models. The difference between the numerically and experimentally obtained force-displacement responses is quantified in a so-called cost function, which further is minimized through a finite element model updating (FEMU) procedure using a Bayesian optimization framework, to establish the optimal set of material parameters. In addition, two commonly employed cost functions employed in FEMU are reformulated in the context of likelihood-based inference to estimate the uncertainty in the optimal set of parameters.

It is shown that neither of the two commonly employed cost functions can recover the variance in the experimental load-displacement curves. However, the mean behaviour is captured fairly well for both cost functions. (Less)
Please use this url to cite or link to this publication:
author
supervisor
organization
publishing date
type
Thesis
publication status
published
subject
keywords
fracture mechicans, finite element modeling, wood, cohesive zone modeling, finite element model updating, likelihood inference, optimization
pages
115 pages
publisher
Division of Structural Mechanics, Lund university
ISBN
978-91-8104-413-3
978-91-8104-414-0
language
English
LU publication?
yes
id
8e8686d8-5040-4871-bf49-f62c7227f0c9
date added to LUP
2025-02-25 13:20:02
date last changed
2025-04-04 14:23:54
@misc{8e8686d8-5040-4871-bf49-f62c7227f0c9,
  abstract     = {{This thesis covers the estimation of the fracture behaviour of the European wood<br/>species Norway spruce and birch. The fracture behavior of wood can be quantified<br/>through three material parameters: the stiffness, the strength, and the specific fracture<br/>energy. In the present work, these parameters are characterized through two methods.<br/>The first method is purely experimental and only estimates one parameter, by use<br/>of the so-called work-of-fracture method. In this procedure, a single-edge-notched<br/>beam is loaded in three-point-bending under stable crack propagation until complete<br/>failure. The specific fracture energy, i.e. the energy dissipated during the creation<br/>of a unit crack surface, can then be evaluated as the work carried out to complete<br/>fracture, divided by the fractured area. This method has previously been used in<br/>a large number of studies to establish the specific fracture energy of various wood<br/>species. In the present work, this experimental procedure is used to test 80 specimens,<br/>divided into four different series.<br/>The second method employed is based on the same experimental tests, but in combination with numerical models. The difference between the numerically and experimentally obtained force-displacement responses is quantified in a so-called cost function,<br/>which further is minimized through a finite element model updating (FEMU) procedure using a Bayesian optimization framework, to establish the optimal set of material<br/>parameters. In addition, two commonly employed cost functions employed in FEMU<br/>are reformulated in the context of likelihood-based inference to estimate the uncertainty in the optimal set of parameters.<br/>It is shown that neither of the two commonly employed cost functions can recover the<br/>variance in the experimental load-displacement curves. However, the mean behaviour<br/>is captured fairly well for both cost functions.}},
  author       = {{Jonasson, Johannes}},
  isbn         = {{978-91-8104-413-3}},
  keywords     = {{fracture mechicans; finite element modeling; wood; cohesive zone modeling; finite element model updating; likelihood inference; optimization}},
  language     = {{eng}},
  month        = {{02}},
  note         = {{Licentiate Thesis}},
  publisher    = {{Division of Structural Mechanics, Lund university}},
  title        = {{Experimental and Computational Analysis of Mode I Fracture in Wood : Likelihood-based inference and finite element model updating}},
  url          = {{https://lup.lub.lu.se/search/files/209181993/Lic_JJ.pdf}},
  year         = {{2025}},
}