Memorias de investigación
Communications at congresses:
Predicting particleboard modulus of rupture through artificial neural networks using production parameters
Year:2017

Research Areas
  • Forestry research,
  • Engineering,
  • Construction of engineering

Information
Abstract
One of the most important properties of particleboard for its structural use is modulus of rupture (MOR). However, the test for this property lacks the immediacy necessary for real-time application to production line control. To reduce the delay in obtaining results, which affects the entire industry, it would be beneficial to have modelling tools to enable determination of mechanical properties using parameters taken directly from the production line. In relation to this, the use of artificial neural networks (ANN) has grown in recent years in various fields of science and technology. The nature of ANNs as universal function approximators makes them powerful modelling tools, particularly when it is important to obtain a high degree of reliability rather than determining the relations between the variables involved. These mathematical models have been successfully applied to obtain the mechanical properties of wood-based products using more easily-measured properties or manufacturing parameters. In this study, an ANN was developed to model particleboard MOR using manufacturing parameters, obtaining sufficient reliability for application in the factory.
International
Si
Congress
CompWood 2017
960
Place
Viena
Reviewers
Si
ISBN/ISSN
978-3-903024-49-6
Start Date
07/06/2017
End Date
09/06/2017
From page
33
To page
33
Computational methods in Wood mechanics for material properties to timber structures
Participants

Research Group, Departaments and Institutes related
  • Creador: Grupo de Investigación: Tecnología de la Madera y el Corcho
  • Departamento: Ingeniería Civil: Construcción, Infraestructura y Transporte
  • Departamento: Sistemas y Recursos Naturales