Theoretical modelling of surface roughness in robotic abrasive disc grinding

Robotic abrasive disc grinding is widely used for surface finishing of structural components in industries such as aerospace and automotive manufacturing, where controlling surface roughness is critical for ensuring product quality and performance. Surface roughness theoretical models play a foundational role in deepening the understanding of surface profile formation and optimizing machining processes. Despite studies in wheel and belt grinding, abrasive disc grinding-specific models are still comparatively rare. This paper addresses this gap by proposing a novel theoretical model that integrates disc machining theory with key process parameters—rotation speed, grinding force, and feed rate—to establish a quantitative relationship with surface roughness. By analysing the material removal process and surface formation mechanism, the model estimates surface roughness from both kinematic and geometric perspectives. The proposed model is validated through robotic abrasive disc grinding experiments conducted under various conditions, achieving an average absolute relative error of 7.3%, with most deviations below 10%. This theoretical model offers physical insights into the grinding process. In the future, the model could be combined with data-driven approaches to enhance generalization and accuracy, leveraging its interpretability to improve predictive performance.