Non-ideal surface modelling approach for enabling uncertainty representation

Non-ideal models have gained considerable attention in the design stage for evaluating product functionality and performance considering realistic conditions of surface. The evaluation generally involves deriving statistical distributions certain key characteristics, such as gaps distances, based on numerous assembly process simulations, thus necessitating a large number samples non-ideal surfaces. traditional practice generating is to model systematic random deviations each point part It difficult, however, represent randomness real surface formation mechanism using this point-wise approach, leading significant computational cost. As result, technique hinders wider adoption specified representations simulation analysis. deviation modes that feature surfaces are dominated by manufacturing error factors, which could be systematically represented corresponding mode functions. Furthermore, variations inevitably cause deviations. In paper, defined uncertainty reflected function coefficients. A conceptual characterises through coefficients phase space therefore proposed. multi-dimensional space, denotes group determines sample. Following distribution law coefficients, can generated sampling. This method reduces problem from modelling an infinite points finite coefficient dimension compact space. enables efficient reflect errors resulting facilitates analysis under effect randomness. To verify method, case study involving tapered manufactured taper-turning presented. discussed with respect typical factors process.