Integrated method for virtual strength prediction of timber

Integrated method for virtual strength prediction of timber

Khaloian, Ani (Technische Universität München)
van de Kuilen, J.W.G. (TU Delft Bio-based Structures & Materials; Technische Universität München)

2021

Before timber boards can be used for engineering applications, they need to be strength graded. This step is currently done based on visual or machine grading methods. Each approach may face problems such as frequent measurement (human) errors or problems in dynamic measurements due to missing density values, respectively. To increase accuracy of the predictions, an advanced numerical method has been developed based on FE-analysis to predict tensile strength of the boards. By simulating the tensile test procedure virtually, stress developments around wood heterogeneities have been analysed, and identifying parameters (IPs) have been provided, representing the stress concentrations in 3D anisotropic space. Virtual dynamic-MoE has been derived after performing the stress-wave analysis, to be used as another IP for strength predictions. These parameters have been used in a non-linear multiple regression analysis with the tensile strength for the predictions. Similar approach has been performed, using the parameters of the visual and machine grading methods. The quality of strength prediction based on virtual method was in the same level/slightly higher than recently available methods, depending on the wood species. The model has been developed by considering a scatter for the quality range of 450 spruce, Douglas fir and beech boards. For the model verification, the approach has been used for strength prediction of a group of ash and maple boards, which provided satisfactory results.

Dynamic MoE
FEM
Prediction
Strength
Stress concentration factor

http://resolver.tudelft.nl/uuid:59bde83c-f3b2-40c8-971a-600c7552ba70

2022-02-27

World Conference on Timber Engineering 2021, WCTE 2021, 2021-08-09 → 2021-08-12, Santiago, Chile

Accepted Author Manuscript

Institutional Repository

conference paper

© 2021 Ani Khaloian, J.W.G. van de Kuilen