Characterization of shrinkage effects in micro-injection moulding (µ-IM)

Abstract

This thesis characterizes the effects on shrinkage in microinjection moulding. The literature review considers four branches of investigation (material properties, processing parameters, mould design and specimen design). Two research gaps rise from the analysis of the literature review: the absence of a standardized methodology for measuring shrinkage of moulded parts at the micro-scale, and the absence of optimization stage that implements multiple quality criteria. Adequate research routes are set in order to address these gaps. The conventional standard for determining shrinkage at the macro scale is adapted to the micro-scale and this bridges the first gap. The micro-mould replicates the same design of the standard, and a preliminary stage solves some mouldability problems: the implemented mould extended the mouldability range of processing parameters for improving the reliability of results. After the micro-mould validation, the study of shrinkage at the micro-scale considers the influence of five processing parameters: the mould and melt temperature, the holding time and pressure, then the injection pressure. The design of experiment approach identifies the critical parameters that affect moulding, post-moulding and total shrinkage in parallel to and normal to the flow direction within an interval of confidence of 95% for POM and 90% for 316L feedstock. Statistical tools analyse the results, and the trends of critical factors found confirmation in the literature. This methodology at the micro-scale can fill the first gap because it is on purpose designed for the micro-scale. Moreover, the binder of feedstock is a mixture of POM based polymers, and the use of a common platform permits to compare directly the two materials and highlight the influence of powder loading. The optimization stage adopts desirability functions for achieving optimized values that simultaneously fulfil two requests: minimize shrinkage and maximize moulded part mass. The analysis of the literature review shows that few papers adopt multiple quality criteria approach as methodology for optimizing the results, and none consider jointly part mass and shrinkage. The optimized processing parameters allow moulding “optimized specimens”, and results demonstrate that their total shrinkage and part mass achieve the requests. Even if the use of desirability functions produce results thatrepresents a compromise between the requests, the results show that overall shrinkage decreases and part mass increases. This approach demonstrates its reliability and bridges the second gap. The last part of the thesis investigates the 316L feedstock behaviour for filling micro-features parallel to and normal to the flow oriented. The moulded features are investigated for studying the replication quality and the effect of the orientation of channels with dimension close to the feedstock lower mouldability value. These informations are available in the literature only for polymers, and the contribution of this part of thesis is to fill this gap by analysing a feedstock. The statistical approach permits to identify the critical factors that affect the feature replication quality. Optical investigations allow to identify the 316L feedstock lower mouldability value and to observe the influence of the orientation of features with dimensions near the lower limit.