Analysis of the Part Forming Process in Gas-Assisted Injection Molding and the Development of Design Guidelines for Plate-Like Parts

Gas-assisted injection molding (GIM) is one of the special injection molding processes and consists of an introduction of gas into a cavity that was previously only partially filled with melt. The gas displaces the melt towards the auxiliaries of the cavity while coring out the thicker sections. That way partially hollow molding can be manufactured. Several different aspects of the GIM process are being investigated and discussed in order to increase the process understanding. First an extensive investigation of the GIM part forming process and the penetration of the gas into thin areas, called fingering effect is being conducted. Furthermore an experimental GIM study using glass-fiber-reinforced thermoplastics is being carried out to examine the fiber orientation mechanism in a simple tubular geometry. Gas introduced directly into the mold cavity via a gas nozzle offers some advantages over the injection of gas through the machine nozzle. An analysis of the influence of different gas nozzle designs on the part forming process is being conducted. A detailed evaluation of a CAE software package for the simulation of the gas-assist process clearly states that state-of-the-art CAE simulation is a very valuable tool for the design of GIM parts allowing to predict fairly accurate the part forming process and the final gas bubble distribution. The main objective of this thesis is focued on the development of design guidelines for plate-like GIM parts by separately investigating different gas channel distributions as well as gas channel cross sections. A topic that received so far little attention is the quality assurance for GIM. The gas bubble distribution inside the part is not only very sensitive to any process changes but also difficult to locate if non-transparent materials are being molded. A useful contribution is made towards an implementation of quality control for the GIM process. It is shown that a statistical design of experiments and an additional evaluation is able to significantly reduce the standard deviation of the part quality.