Investigating the effect of process parameters on dimensional accuracy and ultimate tensile strength of micro injection moulded micro parts

• Main Author: Mani, Mohammad Reza
• Published: University of Nottingham 2016
• Subjects: 620.1; TP1080 Polymers and polymer manufacture
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.692713

This thesis presents two models for optimizing and guiding the micro injection moulding process. The models are generated by the use of a mathematical procedure, an understanding of the process, and empirical data obtained from several sets of experiments. Micro injection moulding is a well-known process that is heavily used in the mass production of micro polymer parts. It is a very reliable process and apart from the initial investment required for manufacturing a mould, the process is very low cost. Furthermore, polymer developments have led to the process being suitable for the production of micro parts in equipment used in several industries such as medical, automotive, aerospace and sensing. Due to these important industrial applications, several quality criteria have been the subject of research in recent years. One of the main challenges in micro moulding is the modelling of the process in terms of polymer flow and accuracy. This is because current available models use PVT data (pressure, volume, temperature) that is used for modelling of conventional injection moulding. Furthermore, these models ignore several factors in micro moulding such as the high shear rates and 3D flow of the polymer melt. Moreover, modelling of the mechanical properties of the micro parts based on mathematical systems used for macro parts leads to large errors. This study proposes a new method for modelling the effect of process parameters on the dimensional accuracy and UTS (Ultimate tensile strength) of micro walls. This results in reduction of risk and cost, and optimization of the process. The “accuracy model” relates the dimensional error to four process parameters (polymer melt and mould temperature, and injection velocity and pressure), polymer characteristics (density, specific heat capacity and thermal conductivity) and a characteristic of the machine (plunger diameter). The “mechanical model” relates the part’s UTS to the same parameters as in the accuracy model. In order to develop the “accuracy model” an understanding of the effect of process parameters on dimensional accuracy and the polymers needs to be obtained. Several sets of experiments were conducted to investigate and establish this effect. Two polymers, Polyoxymethylene (POM) and Polypropylene (PP), were used to conduct the study. The results showed that the polymer melt temperature had the highest effect, followed by injection pressure, injection velocity and mould temperature. Amongst these, injection velocity had an adverse effect on dimensional accuracy. Further analysis was done to investigate whether the effect was consistent for several sets of the parameters. Results of the experiments showed that while the effect was not linear, the trends obtained earlier were correct. The same procedure was applied to investigate the effect of process parameters on the UTS of the micro walls. Polymer melt temperature had the highest level of influence, followed by injection velocity, injection pressure and mould temperature. Increase in all parameters resulted in reduction of the UTS, except for the mould temperature. Next, the two models were developed through a method called dimensional analysis. Several dimensionless expressions were developed to form a general relationship between the parameters and the quality criteria. Then, the obtained results and data were used to find the constants and the specific form of the functions. The overall models were validated by a fresh set of selected experiments using an original brass insert. The achieved trends and models were validated experimentally, using a different mould insert with a micro channel with a different dimension. While the values for the dimensional error and UTS were different, the trends obtained before were correct for the new insert. The same trend was observed with the models. Again, predictions for PP parts had better agreement with experimental data compared to those of POM. In addition, the amount of error for the steel insert was higher, due to different thermal conductivity of the insert material and surface roughness of the micro channels.