OR

I'm looking for in the

Modeling and characterization of structural defects on 5-axis machine tools for in-process measurements

Doctoral thesis defended by Fabien VIPREY on November 18th 2016

Abstract

The principle behind performing metrology during machining is to directly obtain measurement data during the production process flow. This operating principle responds to the increasing need of industry to conduct line measurements as operations are unfolding, or between two machining operations, by employing the means of production to measure the machined part. An effective knowledge of measurement error sources, such as geometric errors, is a precondition to guaranteeing a dimensional metrology directly traceable on machine tools.

This works program pertains to the geometric modeling of machine tools, based on a normalized parameterization of geometric errors. This model has been simulated and simplified through the use of a virtual machine developed as a tool to improve understanding and visualization of the effects of geometric errors on the volumetric error.

A new heat-insensitive equipment standard has been developed: the Multi-Feature Bar. Correlated with the international definition of the meter by calibration plus a European inter-laboratory comparison, this standard allows envisaging traceable measurements on a machine tool within an aggressive environment. The identification of three parameters intrinsic to this standard, combined with an appropriate measurement procedure, ensures the comprehensive and traceable identification of linear axis movement errors. Subsequently, error identification between axes has been based on analyzing combinations of sufficient parameters for an optimal characterization of the volumetric error.

A model parameter identification procedure has been proposed by minimizing the structure's temporal drift and the previously determined movement error effects. A sensitivity analysis performed on measurement procedure settings as well as on noise effects serves to guarantee the quality of the proposed identification approach.