Differential pressure Vacuum Casting is a new kind of rapid manufacturing technology. The filling speed is the key process parameter which decides the filling quality. In order to ensure the stabilization of the flowing speed, qualitative model has been set up. Both CAE software and BP Neural Network Models have been used to find the relation between the inlet pressure and the forward speed. Using the inverse BP Model and giving a constant velocity of every section in the cavity, the expected pressure values have been got. In terms of the Filling Length Measure Method (FLMM) proposed by the researchers, the time points of flowing stage have been evaluated. The relation between the expected values of inlet pressure and time points has been concluded from this study.