Abstract
Form error evaluation of the machined honeycomb surfaces has the power to deliver tremendous benefits to improve the machining accuracy and guide the assembly process. The line laser measurement can meet the high-resolution requirement for thin-walled honeycomb cell walls, but faces a high proportion of noise data consisting of burrs and outliers. In this paper, dimensionality reduction combined with random sample consensus or regression analysis is adopted to distinguish the irregular noise from the measurement data. A directional extraction of honeycomb core surface form based on locally weighted regression is proposed. A highly realistic honeycomb surface simulation method is established to solve the lack of honeycomb benchmarks. The simulation experiment shows that the accuracy of the proposed method is ±7 μm. The comparison experiment with coordinate measuring machine demonstrates the surface form RMS error differences for the polygonal and curved honeycomb by the two method are 7 μm and 19 μm, respectively.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call