Scraper factors investigation on Al2O3 paste flow based on marine predators algorithm-bidirectional gated recurrent unit pseudo-lattice Boltzmann method for stereolithography molding

Abstract

The coating process is an essential part of stereolithography technology to mold alumina paste, achieved using a scraper. The characteristics of the scraper (geometry and roughness) are investigated to discover the potential influence mechanism on the coating process. To complete the analysis, the bidirectional gated recurrent unit (BiGRU) optimized by the marine predator algorithm (MPA) is proposed to provide a predicted model for the rheological data instead of the traditional fitting by using the known rheological equations. Then, a pseudo-lattice Boltzmann (pseudo-LB) method is proposed to conduct the coating flow simulation without a specific rheological equation, where the predicted rheological model is applied to obtain the relaxation time for each iteration. Based on the above MPA-BiGRU pseudo-LB method, three geometries and different roughnesses are explored for the coating process. The results indicate that the rectangle geometry suits the large coating speed, and the trapezium geometry is recommended for conditions with the small coating speed. The roughness significantly affects the coating process with the rectangle geometry, and the smoother scraper generates a minor disturbance. However, it is almost ineffective in the cases with trapezium geometry.