An efficient approach for mesoscale virtual permeability predictions using realistic fiber geometry extracted via micro-ray computed tomography

Achieving realistic voxel geometry to accurately extract the regular structure and dimensions of mesh elements, along with properly resolved yarn curvature, from μXCT scans of composite fiber preforms remains a challenge. Creating such a model from the segmented μXCT images, present challenges towards maintaining the size, boundaries, and the periodicity of the unit cell potentially leading to inconsistencies in predicting the meso-scale permeabilities during infusion in liquid composite moulding. Such inconsistencies in the geometry creation can potentially be resolved by directly creating the flow simulation meshes from the extracted geometry of the μXCT scan, while incorporating the defined classes of yarn orientation and fiber 𝑉_𝑓. Present study proposes the creation of simulation mesh for the flow model by varying pixel densities for multilayer UD and NCF preform in presence of biax cross-plies. A comprehensive workflow, beginning with choice of thresholding and skip voxel has been used in advanced Darcy-Forchheimer model to predict the meso-scale permeabilities with assigned local fiber orientations and micro-scale permeabilities. A detailed study of skip voxel and choices of thresholding parameter proposes an efficient approach in prediction meso-scale virtual permeabilities.