Une approche unifiée pour la mesure et l'analyse des défauts de forme des pièces mécaniques

To reduce post-processing and improve acquired data quality, it is intended to work out a mesurement system allowing to deliver a description of the acquired geometry in a predefined base of defects. This system will lean on techniques of correlation of images and camera measurements. The surface texturing will be made by projecting a speckle patern by means of high definition projector, allowing the realization of a fine texture on mechanical part of big dimensions. This technique would have the advantage to supply directly exploitable measures for the qualification of the considered part and more specifically the identification of the shape defects. In this context, the presented doctoral work aim to propose and to validate an approach of 3D measurement allowing a post-processing time reduction and improving the interoperability of the data acquired with their environment of use (study of the conformity, the prediction of assembly, validation of sizing or performance, etc.). More particularly, these works must be developed for large scale parts like aeronautic structures or bodywork part for which the control of shape defect is a real stake for the assembly. It is then intended to turn to the automation of procedure of measure using digital image correlation (DIC) based on form defect description and speckle pattern projection.