Study Of Penetration Mechanism And Robust Optimization Of Processing For Water-assisted Injection Molding

As a new technology to produce bigger and longer hollow parts, there are many unique advantages in water-assisted injection molding(WAIM) in contradistinction to gas-assisted injection molding, such as short cycle, thin wall thickness and better surface quality, etc. However, WAIM hasn’t developed for a long time, and it needs to intensive study for the molding mechanise, processing, mould and equipment.According to the real problems in WAIM, the main research contents are as follows:Based on rapid cooling and turbulent characteristics of the water, WAIM mathematical model is put forward. To make the discretization and solving ways of govering equations unify, the uniform differential fromat is built up. Several common numernical calculation solutions are compared, and PISO algorithm is selected to solve the equations.Base on the model, penetration behavior of the water is simulated by computational fluid dynamics(CFD) method. Aimed on free surface flow for the water and melt, and turbulence with high Reynolds for the water, the water/melt free interface is tracked by volume of fluid(VOF), and ? ?? and ? ?? turbulent model are compared to study the different water penetration. By simulation, the distribution of temperature, velocity and pressure are analysed, and the water penetraton mechanisms are discovered.Wall thickness distribution is an important indicator which aims at measuring the quality of WAIM parts. Based on typical structure of part, the distribution of residual wall thickness at long straight sections, expansion transitions, contraction transitions, transitions with fillent and curved sections are analysed. Furthermore, the effects of processing parameters on hollowed core ratio are researched. On this basis, measures are presented to improve the fundamental water channel structure, which provides the rationale for design of parts.WAIM experimental equipments are constructed, by which technology exploration is performed for cooling tube of an automobile brand. Based on radial basis function, an inverse neural network model are developed. Inputted the part quality of design expectations, including hollowed core ratio and wall thickness difference, the reasonable short shot size, melt temperature, water pressure, delay time and mold temperature can be predicted. By the inverse model, the aimlessness of processing decision can be avoided.An optimization strategy integrating design of experiment, surrogate model and optimization algorithm is proposed. Through the comparing of different design of experiment, optimal Latin hypercube is selected. To increase the efficiency and precision of computation, ridial basis function surrogate model is employed. and particle swarm optimization algorithm is adopted to find the optimal solution. It shows that the determined Pareto optimal solutions achieve the optimization goal of the hollowed core ratio as big as possible and the residual wall thick difference as small as possible.On the basis of the determined optimization solution, WAIM design for 6 Sigma through robust optimization is proposed. By Monte Carlo simulation, 6 Sigma analysis is processed on the point of determined optimization design. It states clearly that the point of determined optimization design is on the boundary of restraint, the Sigma level is low, and it can’t meet the robust design. As a result, 6 Sigma design by Monte Carlo simulation and PSO algorithm is performed. It indicates that the point of 6 Sigma design is far away from the boundary of restraint, the robustness and reliability is good.