A robust and automated FE-based method for fixtureless dimensional metrology of non-rigid parts using an improved numerical inspection fixture

Abstract

Dimensional inspection is an important element in the quality control of mechanical parts that have deviations from their nominal (CAD) model resulting from the manufacturing process. The focus of this research is on the profile inspection of non-rigid parts, which are broadly used in the aeronautic and automotive industries. In a free-state condition, due to residual stress and gravity loads, a non-rigid part can have a different shape compared with its assembled condition. To overcome this issue, specific inspection fixtures are usually allocated in industry to compensate for the displacement of such parts in order to simulate the use state and accomplish dimensional inspections. These dedicated fixtures, their installation and the inspection process consume a large amount of time and cost. Therefore, our principal objective has been to develop an inspection plan for eliminating the need for specialised fixtures by digitizing the displaced part’s surface using a contactless (optical) measuring device and comparing the acquired point cloud with the CAD model to identify deviations. In our previous work, we developed an approach to numerically inspect the profile of a non-rigid part using a non-rigid registration method and finite element analysis. To do so, a simulated displacement was performed using an improved definition of boundary conditions for simulating unfixed parts. In this paper, we will improve on the method and save time by increasing the accuracy of displacement boundary conditions and using automatic node insertion and finite element analysis. The repeatability and robustness of the approach will be also studied, and its metrological performance will be analysed. We will apply the improved method on two industrial non-rigid parts with free-form surfaces simulated with different types of displacement, defect and measurement noise (for evaluation of robustness).