Evaluation of the informed isotropic (IISO) viscosity model for compression molding of discontinuous fiber reinforced polymers

The informed isotropic (IISO) viscosity model has gained popularity as a surrogate for fourth-order tensor viscosity models in simulating injection and compression molding of discontinuous fiber reinforced composites in industrial applications, primarily to overcome numerical challenges that are especially pronounced in long fiber reinforced materials. In addition, the IISO model can be easily integrated into (commercial) isotropic frameworks. The central idea is to equate the energy dissipation rates resulting from a fully anisotropic and an isotropic viscous material model, which allows deriving a scalar surrogate viscosity that depends on the local fiber orientation. However, the model’s fundamental capability to predict anisotropic flow behavior in compression molding remains limited. This work comprehensively assesses the IISO viscosity model’s capabilities through analytical and numerical investigations of fundamental flow scenarios. We demonstrate that the IISO viscosity model cannot generate elliptical deformation in lubricated squeeze flow of initially cylindrical samples with spatially homogeneous and anisotropic initial fiber orientation states, confirming the model’s inherent limitation due to stress–strain-rate coaxiality. When we extend the analysis to non-lubricated squeeze flow, the results emphasize that the IISO viscosity model also fails to produce anisotropic flow regardless of the spatially homogeneous and (aligned) orthotropic initial fiber orientation state. Furthermore, we demonstrate that the compression-molding-style center-gated disk benchmark is inconclusive, as the flow trajectory depends on the magnitude of the imposed perturbation rather than the fiber orientation. The perturbation also introduces a physically implausible circumferential vortex. Finally, we discuss potential sources of apparent anisotropic behavior in numerical flow simulations in the literature and highlight the challenges of parameterizing the IISO viscosity model experimentally.