Quantitative evaluation of pyramid belt wear using light-reflection characteristic of agglomerate coating and image processing

Abstract

As a new generation of coated abrasive products with a unique pyramid-shaped structure, pyramid belt offers finer and more uniform grinding effects, and has been increasingly applied in precision grinding. However, belt wear occurs inevitably, challenging the control of shape accuracy and surface quality of ground components. Current research on pyramid belt wear is relatively limited, still lacking a universal, rapid method for observing and evaluating belt wear. In response, this paper proposes a quantitative evaluation method for pyramid belt wear using light-reflection characteristic of agglomerate coating and image processing. Firstly, a detailed analysis of the wear types and characteristics of pyramid belt is conducted. Then, focusing on the agglomerate flat wear, a method utilizing the light-reflection characteristic of agglomerate coating for observing belt wear morphology is proposed, enabling clear image capture of belt wear contours. Based on this, a wear coefficient based on the geometric characteristics of agglomerate is defined, and a quantitative evaluation method for belt wear using image processing is further proposed. The wear experiment of the belt is conducted on a robotic belt grinding platform, and experimental results show that the evaluation deviation of the proposed method from the evaluation results based on optical profilometer is within 5% at a 95% confidence interval, demonstrating the effectiveness of this method. Additionally, this method has advantages such as high computational efficiency, low cost, and good usability. Finally, the temporal evolution processes and spatial distribution patterns of belt wear and material removal rate are investigated and analyzed.