ECCOMAS 2024

An orthotropic 3D elasto-plastic damage model for wood

  • Seeber, Franziska (Technical University of Munich)
  • Khaloian-Sarnaghi, Ani (Technical University of Munich)
  • Benvenuti, Elena (University of Ferrara)
  • van de Kuilen, Jan-Willem (Technical University of Munich)

Please login to view abstract download link

Developing a generalized model that covers the general scatter of wood may help to better understand the mechanical behavior of this material. Besides the strong anisotropy of wood, heterogeneities e.g. in the form of locally varying fiber deviations and knots make the strength prediction demanding [1]. In addition, the loading-direction dependency and the response of the material under different loading configurations need to be taken into account for modeling. A 3D elasto-plastic continuum damage model was developed and validated to take into consideration the material’s ductile behavior under compression and the brittle material’s behavior under tension. Various orthotropic stress-based damage initiation criteria were implemented and compared in this study. Since mesh-dependency limits the approach, the model is enhanced by non-local damage [2]. Through gradient enhancement, a non-local variable is incorporated considering the effect of interaction from the micro-structure of wood. This is accomplished while maintaining anisotropy, stress-based damage initiation criteria, as well as an appropriate size of fracture process zone. In this way accumulating micro-cracks evolve into a macro-crack representing several failure modes within the CDM framework. The systematic validation is carried out by comparing numerical results to experiments on globally oriented clear wood samples, tensile tests with fiber deviations, as well as four-point bending tests with combined plastic and damaging material regions. After validating the model, a satisfying agreement was observed between the numerical and experimental results. In conclusion, the model targets not only to increase the accuracy of strength prediction but is developed for combination with other numerical models representing e.g. moistureor time-dependent behavior for understanding the complexity of wood.