Scholarlink

<p>This work presents an integral non-local damage model, implemented in a finite element framework, for the objective prediction of the mechanical behavior up to failure of human femurs under single-leg stance loading. Patient-specific computational models were built by reconstructing the femoral geometries through semi-automatic segmentation of computed tomography images, from which personalized material properties have also been derived. Progressive loss of bone structural integrity was described using a continuum damage mechanics formulation, bridging the lack of objectivity inherent in local damage approaches with an integral non-local regularization. The available benchmark experiments, conducted on two different <i>ex vivo</i> femurs, were reproduced <i>in silico</i>, obtaining mesh-convergent numerical results that show very good agreement with the corresponding experimental findings. The proposed approach can potentially improve the predictive accuracy of femoral fracture risk and represents the first step towards the development of patient-specific computational tools aimed at supporting clinical decision-making, particularly in cases where current diagnostic standards lead to uncertain assessments due to their limited specificity.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

An integral non-local damage model for objective predictions of mechanical failure in human femurs

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月22日

查看全文
  • Pierfrancesco Gaziano

摘要

This work presents an integral non-local damage model, implemented in a finite element framework, for the objective prediction of the mechanical behavior up to failure of human femurs under single-leg stance loading. Patient-specific computational models were built by reconstructing the femoral geometries through semi-automatic segmentation of computed tomography images, from which personalized material properties have also been derived. Progressive loss of bone structural integrity was described using a continuum damage mechanics formulation, bridging the lack of objectivity inherent in local damage approaches with an integral non-local regularization. The available benchmark experiments, conducted on two different ex vivo femurs, were reproduced in silico, obtaining mesh-convergent numerical results that show very good agreement with the corresponding experimental findings. The proposed approach can potentially improve the predictive accuracy of femoral fracture risk and represents the first step towards the development of patient-specific computational tools aimed at supporting clinical decision-making, particularly in cases where current diagnostic standards lead to uncertain assessments due to their limited specificity.