Development and Evaluation of a Hexahedral Mesh-Morphing Strategy for the Knee Meniscus

The state-of-the-art for generating accurate and high-quality hexahedral meshes of irregular geometries for finite element (FE) simulations is a laborious and time-consuming endeavour. Mesh-morphing is a technique which modifies the vertices of an existing mesh to match the boundary prescribed by another. This technique has been chosen as it has the capacity to automatically create accurate and high-quality hexahedral meshes of irregular geometries from a single template. The meniscus was chosen as it is an important component of the knee and challenging structure to simulate. Hexahedral discretisation is often required due to: multi-body contact, large deformations and complex material properties (nearly-incompressible and anisotropic). In this research, a novel, automatic and general-purpose mesh-morphing strategy was developed. Additionally, several robust and thorough methodologies were developed to assess the sensitivity and validate the performance of the mesh-morphing strategy. This was achieved through performance comparisons against a state-of-the-art procedure - the multi-block method (IA-FEMesh). The performance metrics not only assessed the capacity of the mesh-morphing strategy (i.e. speed, accuracy and mesh-quality), but also the functionality for meaningful FE simulations. The mesh-morphing strategy generated 20 challenging meniscus meshes in under a minute compared to an average of 26 minutes, with comparable surface error and mesh-quality metrics to the state-of-the-art. Also, there was no significant difference in the FE simulation outcomes. The mesh-morphing strategy offers a faster, competitive and automated alternative to the semi-automatic state-of-the-art, and only required one template mesh. The strategy can already be applied to a diverse range of geometries, and with some trivial modifications can approach a larger proportion. The developed tool can be used to improve productivity and automate the development of FE models. This could enable the design of large-scale studies and assist the development of digital twins. A wide-range of industries that require automatic and accurate meshing from 3D scanning technologies could utilise this work.