Monitoring Multistage Surface Spatial Variations Using Functional Morphing

In multistage manufacturing processes, the machined surface shape of a part changes as it goes through each stage. Process monitoring at multiple stages is necessary for root cause diagnosis and surface variation reduction. However, due to measurement time and capacity constraints, it is challenging to collect sufficient surface measurements at all intermediate stages for monitoring. This paper proposes a functional morphing based algorithm to monitor the surface variation propagation using end of line multi-resolution measurements supplemented with low resolution measurements at intermediate stages. The surface changes over multiple stages are captured by a functional morphing model which integrates geometric transformations with engineering insights. The model estimates a morphed surface prediction at an intermediate stage of interest using end-of-line surface measurements. This morphed surface is combined with the low-resolution measurements at that stage to improve the surface prediction accuracy. The model can be further improved by incorporating the effects of correlated process variables. Based on the model, abnormal surface variations can be detected and located by a single-linkage cluster monitoring algorithm as developed in our previous work. The case study of a two-stage machining process demonstrates that the method successfully monitors multistage surfaces using reduced measurement resolution at intermediate stages.