Biomaterials are typical structures of poroelasticity in nature. Due to heterogeneous composition of microstructures, the asymptotic based homogenization was introduced to correlate the microscopic solid-fluid phase to the macroscopic balance laws. In this work, a microstructure informed computational method for modeling of porous biomaterials is developed. To construct microscopic models directly from medical images, the active contour model based on variational level set formulation for interface and boundary identification is first introduced, which overcomes the difficulties such as the jagged interface, background noise, and blurred objects commonly encountered in the numerical simulation. Inspired by the discretized image pixels, the direct strong form with collocation in conjunction with the reproducing kernel approximation is introduced to solve the level set equation. A gradient reproducing kernel collocation method (Chi et al. 2013) requiring only first order differentiation of approximation functions is then introduced for solving characteristic functions efficiently in the microstructures using pixels as the discretization points. Finally, an image based multiscale modeling of porous biological materials with application to bones is performed to demonstrate the proposed method (Yang et al. 2012).
|出版狀態||Published - 1 一月 2013|
|事件||13th East Asia-Pacific Conference on Structural Engineering and Construction, EASEC 2013 - Sapporo, Japan|
持續時間: 11 九月 2013 → 13 九月 2013
|Conference||13th East Asia-Pacific Conference on Structural Engineering and Construction, EASEC 2013|
|期間||11/09/13 → 13/09/13|