COMPUTATIONAL VISCOELASTIC DAMAGE MODELING OF COMPOSITE MATERIALS IN EXTRUSION DEPOSITION ADDITIVE MANUFACTURING

Extrusion deposition additive manufacturing (EDAM) is a material extrusion method within the additive manufacturing technique, this method utilizes a screw and heaters to drive molten short fiber polymer composite through an orifice. The use of short fiber composite for EDAM has enabled large-scale 3D printing of tools and molds for traditional composite manufacturing processes, and structures. Additive manufactured (AM) short fiber composites (SFC) are both anisotropic and viscoelastic, with mechanical properties exhibiting strong non-linear behavior and temperature dependence. This means that the material processing history impacts the final residual stress and deformation states in a printed part. To simulate the in-manufacturing and in-service complex behavior of an AM SFC part, a homogenized physics-informed material model is required to be able to capture non-linear behavior stemming from sub-scale damage mechanisms. With this in mind, a viscoelastic damage model is developed within a thermodynamically consistent framework. This model is implemented into the implicit finite element code ABAQUS and integrated with the Additive3D framework previously developed at Purdue CMSC.

Damage refers to failure mechanisms associated with fracture, these mechanisms act to degrade the stiffness of the material. The proposed material model is developed under the continuum damage mechanics and thermodynamics frameworks; therefore, thermodynamic-consistency in enforced and representative orthotropic damage variables describe anisotropic damage nature. A damage surface is defined using the energy-norm, which is related to the thermodynamic forces conjugated to the damage variables, hardening function, and material-dependent coefficients. Temperature-dependent material parameters are used to capture experimentally observed attributes of the stress versus strain. Two approaches are presented, a model calibrated using simple uniaxial tensile experiments conducted at ambient and elevated temperatures. The second approach, considers the unilateral effect alongside extensive nonlinear behavior under shear by utilizing the sign of the normal strains to evolve normal damage variables that are dissociated into tensile and compressive modes, and using independent shear damage variables, respectively.

The material model is implemented as a user-defined material subroutine and exercised in the commercial finite element analysis software Abaqus/Standard. The performance of an additive manufacturing mold using the modified Additive3D framework is presented to demonstrate the viability of predicted progressive damage at elevated temperature for a compression molding tool.