Modeling and predicting the performance of coordinate measuring machines

Researchers and practitioners may choose to improve the performance of coordinate measuring machines (CMMs) and systems with the eventual goal of reducing systematic errors by studying and compensating for the sources of such errors. However, this approach usually involves a substantial amount of time and cost. Another drawback of this approach is that considering only systematic errors does not provide accurate evaluation of machine performance and is not a good representative of the total error budget of the machine. Random errors, in many cases, have considerable contribution to machine performance. The uncertainty originating from a CMM is, in almost all cases, caused by random errors. Although uncertainty is unavoidable, it can be estimated using some statistical methods.; This research develops an alternative approach for the prediction of uncertainty of coordinate measuring machines using geometric error information in each axis or repeatedly measured coordinates. It also introduces a procedure to calculate the probability of accepting a measured position within a specified range.; The developed model has resulted in an improvement in the probability of accepting a measured position, compared to previously published models and results. The model may also be used for determining an optimal tolerance for the manufacturing and measurement processes and provides less costly approach for evaluating the performance of coordinate measuring machines.