Derivation of a cost model to aid management of CNC machine tool accuracy maintenance

Manufacturing industries strive to produce improved component accuracy while not reducing machine tool
availability or production throughput. The accuracy of CNC production machines is one of the critical factors in
determining the quality of these components. Maintaining the capability of the machine to produce in-tolerance
parts can be approached in one of two ways: run to failure or periodic calibration and monitoring. The problem is
analogous to general machine tool maintenance, but with the clear distinction that the failure mode of general
machine tool components results in a loss of production, whereas that of accuracy allows parts to be produced,
which are only later detected as non-conforming as part of the quality control processes. This distinction creates
problems of cost-justification, since at this point in the manufacturing chain, any responsibility of the machine is
not directly evident. Studies in the field of maintenance have resulted in cost calculations for the downtime
associated with machine failure. This paper addresses the analogous, unanswered problem of maintaining the
accuracy of CNC machine tools. A mathematical cost function is derived that can form the basis of a strategy for
either running until non-conforming parts are detected or scheduling predictive CNC machine tool calibrations.
This is sufficiently generic that it can consider that this decision will be based upon different scales
of production, different values of components etc. Therefore, the model is broken down to a level where these
variables for the different inputs can be tailored to the individual manufacturer.