Autoclave consolidation is used to manufacture continuous fiber composites in applications that have strict part porosity requirements. The applied positive pressure in this process is attractive for reduction in part porosity. However, some part geometries can cause porosity issues even under positive pressure. One is the concave corner seen in an L-bracket geometry. Higher porosity is seen in areas of high curvature, hindering part quality. Since the autoclave process takes several hours and prepreg material is expensive, trial-and-error methods of resolving issues are not practical. In this work, a unique physics-based viscoelastic model is proposed to describe the mechanical behavior of uncured continuous fiber thermoset prepreg undergoing consolidation under hydrostatic pressure. This model considers stress due to compaction of the fiber network, compression of voids in the resin, and viscous stress from resin flow relative to fibers. The constitutive expressions for these are coupled to important mechanisms that occur during autoclave consolidation. The viscoelastic model is incorporated into the finite element analysis software ABAQUS/Standard using a UMAT subroutine. Numerical results are validated by analytic solutions and experimental comparison for flat and L-bracket geometries. A parametric study identifies important process and material parameters that influence the quality of the manufactured part.

热压罐固结用于在具有严格零件孔隙率要求的应用中制造连续纤维复合材料。在此过程中施加的正压对于减少零件孔隙率具有吸引力。然而，某些零件的几何形状即使在正压下也会导致孔隙率问题。一种是 L 形支架几何形状中的凹角。在高曲率区域会出现较高的孔隙率，从而影响零件质量。由于热压罐过程需要几个小时并且预浸材料昂贵，因此解决问题的试错方法并不实用。在这项工作中，提出了一种独特的基于物理的粘弹性模型来描述未固化的连续纤维热固性预浸料在静水压力下固结的机械行为。该模型考虑了由于纤维网络压实、树脂中空隙压缩以及树脂相对于纤维流动产生的粘性应力而产生的应力。这些的本构表达式与高压釜固结过程中发生的重要机制耦合。使用 UMAT 子程序将粘弹性模型合并到有限元分析软件 ABAQUS/Standard 中。数值结果通过平面和 L 形支架几何形状的解析解和实验比较进行了验证。参数研究确定影响制造零件质量的重要工艺和材料参数。使用 UMAT 子程序将粘弹性模型合并到有限元分析软件 ABAQUS/Standard 中。数值结果通过平面和 L 形支架几何形状的解析解和实验比较进行了验证。参数研究确定影响制造零件质量的重要工艺和材料参数。使用 UMAT 子程序将粘弹性模型合并到有限元分析软件 ABAQUS/Standard 中。数值结果通过平面和 L 形支架几何形状的解析解和实验比较进行了验证。参数研究确定影响制造零件质量的重要工艺和材料参数。