Evaluating and controlling process variability in micro-injection moulding

Microsystem technologies require relatively strict quality requirements. This is because their functionalities are usually dependent on stringent requirements of dimensions, masses or tolerances. When mass-producing micro-components, e.g. replication of disposable microfluidic diagnostics devices, the consistency of the produced components could be significantly affected by process variability. The variability could be associated with a specific process parameter or could be a result of process noise. This paper presents a methodology to assess and minimise process variability in micro-injection moulding, an example of well- established mass-production techniques for micro-components. A design-of- experiments approach was implemented, where five process parameters were investigated for possible effects on the process variability of two components. The variability was represented by the standard deviation of the replicated part mass. It was found that melt temperature was a significant source of variability in part mass for one of the components, whilst the other was affected by unsystematic variability. Optimisations tools such as response surfaces and desirability functions were implemented to minimise mass variability by more than 40%.