Polymer Processing Technologies for Microinjection Moulding

MEMS and synonymous microsystems technology are ubiquitous drivers for manufacturing technologies. As the fundamental element of microfluidic chips, which are a new branch of MEMS applications, polymeric micro components and macro parts with micro-scale features are critical elements in these chips. They are mainly produced by the micro injection moulding process due to the mass production capability of this important manufacturing technology. However, micro injection moulding is not merely a scale down of conventional injection moulding; it requires a rethinking of each part of the process, especially for the replication of high aspect ratio micro features. This brings many more challenges for the simulation of filling, as well as for successful demoulding. This thesis has characterized the micro injection moulding process and simulation of the filling process with the help of on-line process monitoring. The specific objectives of this thesis are to understand the factors that affect the replication of micro features based on thin-wall substrates including the demoulding process and the factors that should be considered in simulating the filling process. An on-line monitoring system was developed to characterize key aspects of injection moulding of four different mould inserts using a reciprocating micro injection moulding machine, and the actual process data from it were calibrated before using it in subsequent simulations. The commercial software Moldex3D and Rhino were adopted for simulation of the micro injection moulding process. To acquire the actual heat transfer coefficient during the micro injection moulding process and investigate its influence on the simulation accuracy, an experimental method was presented to calculate it in the micro injection moulding process of a 600μm thick dogbone part by using a changeable mould insert and PT sensors. It was found that the experimentally measured heat transfer coefficient agreed closely with the automatically determined one in Moldex3D. The heat transfer coefficient could be affected by the injection velocity, packing pressure and mould temperature, while the value in packing stage was higher than that in the cooling stage. The demoulding characteristics in the micro injection moulding process were investigated. First, two designed machined aluminium mould inserts were used to calculate the demoulding force of an array of 4×5 free-standing micro ridges (200×200×5000μm) based on the DOE method. The moulded micro ridges with aspect ratio of 1 were demoulded successfully and mould temperature had a vital effect on the demoulding force of the micro ridges. The best set of process parameters was determined to minimise the demoulding force and the friction force was smaller than the adhesion force at the optimum parameter levels. Secondly, the demoulding characteristics of a cluster of micro ridges (more than 627) of different aspect ratios (1, 2.5 and 4; 100μm height) on a thin-wall substrate were investigated, including the variotherm and venting process conditions via the DOE method. It was found that variotherm had a negative influence on the demoulding force as a result of the increased filling height of micro features under variotherm, and venting could increase the demoulding force due to the air exhaust action. Both the filling simulations of the above parts with different micro ridges were conducted using the short shot method based on the actual online data from experiments. The simulation meshes were carefully constructed using hybrid and boundary layer mesh methods from Rhino and Moldex3D Designer. The influence of factors of heat transfer coefficient, the friction coefficient of wall slip and the initial air pressure of venting on simulation accuracy was studied. The simulation results agreed well with the experimental outcomes and finally, the reservoirs on a microfluidic single cell chip were selected to validate the effectiveness of such simulation method.