Studies On Improving Weld Lines In Molded Parts By Exploiting Sequential Injection Molding

Weld lines are one of the critical defects in injection molding production. Weld lines usually have a negative effect on both the aesthetics and the performance. In order to improve injection molding quality, the location and length of weld lines should be controlled and positioned in the non-critical region. The strength of weld lines should also be increased in weld regions. To solve the above-mentioned problems, this paper proposes a sequential injection molding (SIM) or sequential valve gating (SVG) technique to control weld lines and provide balanced fill by adjusting programmed time sequence to the opening and closing of valve gates. Multi-gate sequential control is a complex issue due to multi-coupled facrors, and variable production conditions lead to the issue more difficult to be controlled. Therefore, how to find an effective multi-gate sequential control method to improve injection molded part quality is becoming a crucial issue that need be resolved. Accordingly, the Back Propagation Neural Network (BPNN) - Improved Artificial Fish Swarm Algorithm (IAFSA) optimum method is proposed and used to control time sequence to the opening of valve gates based on numerical simulation and experiment. The research in the paper is mainly consisted of four parts:1) Based on the principles of viscous fluid, theoretical models of the whole filling stages in runner/nozzle/cavity are developed based on some assumes and simplifications. The numerical models are developed by hybrid finite-element/finite-difference/control-volume method. The part quality indexes such as weld line, air trap, volumetric shrinkage, sink mark index and warpage are further predicted based on the simulation. Numerical simulation experiments of sequential injection molding are conducted with three typical parts to investigate relationships between quality indexes and time sequence of valve gates, and the optimum control modes are determined by evaluating simulaition results.2) The weighted sum of valve-gate number, injection time, maximum injection pressure, maximum calmping force and maximum temperature difference of the part are chosen as the optimization objective function, and the optimization model which can optimize both the number and locations of the valve gates is introduced. Batch numerical simulation of several valve-gate number/location configurations are conducted and evaluated. The minimum value of objective function can be found with the corresponding optimum valve-gate number/location configuration. The weighted sum of maximum injection pressure, maximum calmping force, total weld-line length, total air traps volume and maximum warpage are chosen as the optimization objective function, and the optimization model which can optimize time sequence of the valve gates is introduced. Orthogonal array experiments are conducted and evaluated. The minimum value of objective function with the corresponding optimum time sequence to the opening of valve gates can be found by using BPNN-IAFSA method. The weld lines in molding parts are eliminated by appling the BPNN-IAFSA method.3) Based on conventional methods,this paper presents an approach to effectively control weld-line locations by applying SIM. First, the method partitiones the moulding into sub-mouldings based on the specified weld-line locations. Second, the method determines the optimum valve gate location in each sub-moulding by applying equal-flow-path-length algorithm. Finally,the sum of difference absolute value between two indexes of conformity degree of adjacent sub-mouldings are chosen as the optimization objective function. The minimum value of objective function with the corresponding optimum time sequence of the valve gates can be found by using BPNN-IAFSA optimum method.4) Forming process of weld line and the reason of improving weld-line strength by using SIM process is further analyzed from temperature, pressure, and the distance between weld region and the locations of valve gates. A modified mathematical model is developed based on traditional prediction model of weld-line strength. In the model, the free energy difference can be calculated by applying Sanchez-Lacombe Lattice Theory. Furthermore, the degree of bonding affected by some factors such as temperature and pressure are all taken into consideration. Combined with the numerical simulation to obtain temperature and pressure in weld regions, the model can be utilized to predict the variation rules of weld-line strength of sequential injection molding parts, and prediction results accord well with the experimental results.