Experimental and Numerical Study of Rubber Flow in the Extrusion Die of a Weather Strip

Abstract

Extrusion is the main method used to produce rubber weather strips in automotive industries, and the quality of the nal product largely depends on the thermal properties of the process output. Therefore, precise thermal control of the process is the key to product quality control. This study establishes a three-dimensional model of the nonisothermal viscous ow of ethylene propylene diene monomer (EPDM) rubber melts through a power law rheological model and a mixed nite element method. The rheological properties of the lled rubber compound were characterized using a capillary rheometer (Rosand) at di erent temperatures to evaluate the required material parameters for numerical simulation. Curing characteristics were investigated using a rubber process analyzer (RPA-2000) to construct a curing curve at di erent temperatures. The pressure-stabilized Petrov–Galerkin (PSPG) method and streamline upwind/ Petrov–Galerkin numerical scheme were employed to solve the ow equations and increase numerical stability. The power law rheological model was combined with eld equations such as continuity, momentum, and energy equations to determine the complex ow behavior in an extrusion die of real geometry. Extrusion experiments were performed in an industrial extrusion line, and temperature and pressure were measured at di erent extruder speeds by using special sensors mounted on the extrusion die. The results con rmed that for EPDM rubber compound, the extruder speed exerted a remarkable e ect on the temperature rise and pressure drop in the extrusion die. The impact of viscous dissipation on the thermal behavior and pressure drop prediction of the rubber compound ow is also discussed. The obtained scorch time was compared with the estimated residence time in the ow domain to elucidate the in uence of extruder speed on the processing characteristic. The results suggested the lack of premature vulcanization or the start of scorching inside the ow domain within the studied extruder speed range. The validity of model prediction was veri ed by comparison between simulation and experimental results. The predicted results of the model showed good agreement with the experimental data.