Newsletter #2 topic - Modeling of TW

Transparent wood is an emerging optical material with sustainability advantages. It is generally prepared by decolorization (delignification or bleaching) followed by polymer infiltration. The refractive index of the polymer needs to be matched with decolored wood. Transparent wood can have a transparency up to 90% for thin materials, which is combined with high haze. Transparent wood and related biocomposites exhibit a complex microstructure originating from pristine wood. The resulting composite is heterogeneous across multiple scales, from the cell structure and growth rings at larger scales to cellulosic microfibrils at smaller scales, see Fig. 1. The microstructural fingerprint of transparent wood and related biocomposites determines the optical, mechanical, and thermal properties of the material, for instance, its orthotropic nature (characterized by a high strength and stiffness along the grain).

Fig 1: Hierarchy representation of transparent wood across different observation scales

In the AI-TranspWood Project, we adopt several modeling methods to understand and predict properties of Transparent Wood based on the microstructure. Both analytical and numerical tools are required to cover all microstructural aspects. Analytical tools allow to study large sets of composites with minimal computation time, while numerical tools allow for incorporating greater details, for instance a more exact geometrical representation of the cell architecture. With the help of AI-based surrogate models, modeling frameworks are created that help develop sustainable bio-based composite materials.Part of the effort was put on the modeling of wood structure which is highly related to transparent wood processing and properties. Wood has a complex 3D structures with different types of cells with various dimensions including fibers, vessels, and ray cells. A distortion-map-based model was constructed to simulate realistic 3D microstructures of wood and transparent wood, which allows easy refining. Figure 2 shows the steps to generate the 3D wood structure. Combing with geometrical optics, the scattering properties of 3D TW models of cellular microstructure could be investigated numerically. For example, effects from different material parameters on ray scattering could be analyzed.

Figure 2 Steps in the method to generate wood microstructures with fibers, vessels and ray cells with realistic dimensions and distribution. (Chen B, Montanari C, Popov S, et al. A distortion-map-based method for morphology generation in multi-phase materials-application to wood[J]. Composites Science and Technology, 2023, 244: 110262.)

Early results of the analytical micromechanical models highlight the importance of the bond between the infiltrated polymer and the delignified cellulose. Only if these interfaces are reasonably stiff and strong, the Transparent Wood is stronger than pristine wood. The numerical models so far enable to calculate realistic stress concentration fields around the microstructural defects, such as the large vessels or ray cells, where cracking is usually initiates.