A Statistical Method of Distinguishing and Quantifying Tolerances in Assemblies

Hua Chen,Xin Li,Sun Jin
DOI: https://doi.org/10.1016/j.cie.2021.107259
IF: 7.18
2021-01-01
Computers & Industrial Engineering
Abstract:Distinguishing and quantifying the roles of tolerances in assemblies are very important to tolerance allocation and optimization. Some of geometric tolerances, such as orientation tolerances and form tolerances, are difficult to be considered in tolerance analysis and allocation processes. Three dimensional (3D) tolerance analysis methods explore a new way to deal with these tolerances. However, existing 3D models lack of effective methods to distinguish and quantify these tolerances in a statistical way. In this paper, based on a unified Jacobian-Torsor model, a statistical method of distinguishing and quantifying tolerances in assemblies is presented. The internal relation of tolerances inside a combinational tolerance is established by data fitting methods, and a calculating scheme used for percent contributions and their subdivisions is proposed. This method improves the utility of the unified Jacobian-Torsor model and has reference meaning to other 3D models. At last, this method has been illustrated on a simple two-block assembly. Sufficient comparison and discussion were made for the results, which demonstrated the effectiveness of the method.
What problem does this paper attempt to address?