D-optimal design of artifacts used in-machine software error compensation

A methodology for the design of artifacts used in software error compensation of rapid prototyping (RP) machines is presented. In software error compensation, an artifact is frequently utilized for measuring parametric errors of the machine axes. In the past, different artifacts have been used and designed in an ad hoc way. To date, no method is available to evaluate the design of an artifact based on its effect on the precision of the fitted models estimated using the artifact. This paper proposes a method to evaluate and improve the design of an artifact using design of experiments (DOE) techniques and numerical optimization. As demonstrated by the results, a D-optimality criterion optimized using simulated annealing techniques can be used to evaluate and improve the design of an artifact so that the components of the volumetric error are predicted more precisely, with the benefit of better error compensation over repeated use of the error models. Although in this paper the concept is only demonstrated for the SLA 250 RP machine, it can be applied to the design of any artifact used for machine error compensation.