| In the precision injection-molded field, making use of CAE (Computer Aided Engineering) technology to control the final quality of the parts that have complex mold cavity by the filling processing parameters has become a very important method. While the complexity of the filling process and the instability that the polymer melt flow at the phase, how to design the optimal processing parameters such as injection pressure, filling time, injection temperature, injection velocity, is a problem that puzzles the researchers in the injection molding numerical simulation field. How to combine the modern optimizing theories with the technology of CAE and how to use it in the processing of polymers have become the hotspot of the researching field.The unsteadiness of the MFV (Melt-Front-Velocity) leads to the non-uniform shrinkage, inconsistent tropism and warpage in appearance of the parts. How to control MFV is the main problem in this researching field. In this dissertation a very good mind is presented that can correctly select the time points during the filling phase according to MFA-Filling time curve from flow simulation, and change the corresponding injection flow rate. We can make MFA advancing more steady by controlling the ram-speed. We get more uniform MFV (Melt-Front-Velocity) at last.Flowing simulation and Genetic Algorithms (GA) are two important foundation to solve the MFV optimized question, the former let us get some data easily, such as the pressure field distribution at any time in the mold, MFA advancing position and the corresponding MFV, these data provide the direct condition to outspread the optimized compute. GA has many merits that includes independence to the question, the global optimized, connotative parallelity, high efficiency and strong illegibility to solve the different non-linear question, to the multi-objective and non-linear MFV optimized question in the paper, the solution proves the standout optimized function ofGA.Aimed at the solution of MFV optimized question, the author makes a systematical study that includes searching information, engineering analysis, academic analysis, selecting solution way and realizing the result by program. The main work is as following:1. Based on thorough analysis in physical course of filling phase in complex cavity, the optimized question was divided into two parts to solve. At first, we optimize the control time point and zone number on the MFA-Filling time curve, then we optimize the injection flow rate. With this mind, we construct each mathematic model (objective function and restiction) according to the factual question.2. On the basis of the work of the filling simulation, we select GA to solve the question. The float code mode is used to code for chromosome according to the character of the MFV optimized question. A new kind of copying rules about filial generation is brought forward during the process of copying and the prematurity is avoided effectively. After the operation process of copying, intercross and aberrance, the optimist situation and the flow rate of the reference point are received.3. From the result that was optimized by sample parts, the object value can always decrease by 50%-70% after optimizing, while the time used by the optimization process is not over 5 seconds usually. It is nothing to speak of the time comparing to the time that simulate the flowing process, and the later is often from many minutes to hours.4. For the MFV optimized question, we have finished the software by which it was realized on computer. This program was written by Microsoft VC++, it works in multi-thread mode, moreover, it encapsulated the two function: Flow Simulation and GA, it made them work better for main thread together. At the same time, the interface is the standard one from MFC (Microsoft Foundation Class), and the document is serialized also. These make it more fashionable and standardizing. At last, the result was visual in this-v-program, it makes the users to check the middle results dynamica... |
PDF Full Text Request