Abstract : In-process metrology consists in obtaining measurement data directly into the manufacturing process. This method results from an increasing need of manufacturers to carry out on-line measurements during one manufacturing task or between two manufacturing tasks by using the mean of
production to measure the machined part. Monitoring the sources of errors like geometric errors is one of the prerequisites to ensure the traceable dimensional metrology directly on the machine tool.
This thesis deals with the geometric modeling of 5-axis machine tool based on a standardized parameterization of geometric errors. This model is simulated and simplified by the use of a virtual machine developed in order to help understand and visualize the effects of geometric errors on the volumetric error. A new standard thermo-invariant material namely Multi-Feature Bar has been developed.
After its calibration and after a European intercomparison, it provides a direct
metrological traceability to the SI meter for dimensional measurement on machine tool in a hostile environment. The identification of three intrinsic parameters of this standard, coupled with a measurement procedure ensures
complete and traceable identification of motion errors of linear axes. The identification of position and orientation errors of axis is based on an analysis of combinations of necessary parameters to characterize volumetric error and at best. A model parameter identification procedure is proposed by minimizing the time drift of the structural loop and the effects of previously identified motion errors. A sensitivity analysis of the measurement procedure settings and of the noise effects ensures the quality of this proposed identification.

Mots clés : 5-axis Machine Tool, Modeling, Identification, Geometric Errors, Material Standard