基于有限体积法的光固化树脂黏弹性力学行为研究
INVESTIGATION OF VISCO-ELASTIC MECHANICAL BEHAVIORS OF UV-CURING RESIN BASED ON FINITE VOLUME METHOD
-
摘要: 光固化聚合物体系成分复杂, 随时间变化的黏弹性力学行为具有不确定性. 因此, 以代表性光固化树脂中的聚氨酯丙烯酸树脂(polyurethane acrylate, PUA)为例, 基于二阶位移展开有限体积法及分数阶黏弹性本构关系发展光固化树脂的力学模型. 首先, 根据边界条件确定胞元面的连续性条件和数值模型的总刚矩阵, 显式计算所有胞元的未知面位移. 为探究光固化树脂在不同应变率影响下的黏弹性力学行为, 引入分数阶有限变形Kelvin-Voigt黏弹性力学模型, 建立应力-应变之间的本构关系. 然后, 通过应变率分别为10−4, 10−3, 10−2以及10 s−1的单向拉伸试验完成本构模型中弹性项以及黏性项的参数识别. 最后, 开展不同应变率下的单向拉伸试验及数字图像散斑(digital image correlation, DIC)分析, 评估数值模型的计算误差. 结果表明, 文章所建立的数值模型能有效预测不同应变率下光固化树脂的力学行为, 并且平均预测误差仅为1.98%.Abstract: The complexity of the ultraviolet-curing (UV-curing) polymer system, comprised of various resins, contributes to the uncertainty in its viscoelastic mechanical behaviors as it evolves over time. By using polyurethane acrylate (PUA) as a representative material, the fractional order viscoelastic constitutive is incorporated into the second-order displacement expanded finite volume model to explore the mechanical behaviors of UV-curing resin in this paper. Firstly, the continuity conditions of cell surface and the global stiffness matrix of the numerical model are determined in accordance with loading conditions, enabling the explicit calculation of unknown surface displacements for all cells. To investigate the viscoelastic behavior of UV-curing resin with respect to varying strain rates, a fractional-order finite deformation Kelvin-Voigt viscoelastic model is introduced to establish the stress-strain constitutive relationship. Secondly, parameters of the elastic and viscous components in the constitutive model are determined through uniaxial tensile tests at strain rates of 10−4, 10−3, 10−2, and 10 s−1. Finally, digital image correlation (DIC) analysis is performed in conjunction with uniaxial tensile tests at various strain rates to compare the experimental results and evaluate the computational accuracy of the numerical model. It is indicated that the established numerical model can effectively predict the visco-elastic mechanical behaviors of light-cured resin under different strain rates, with an average prediction error of 1.98%.