Research on gear shaving error analysis based on knowledge graph

The influencing factors of shaving-induced mid-profile concavity error and tooth profile errors are numerous and complex, yet they lack effective interconnections, resulting in severe siloization issues. Existing literature research and practical manufacturing data have primarily accumulated the relationship between the influence of individual factors on gear shaving error and stored in the form of text dispersed. These discrete knowledges can't be systematically integrated and utilized and individual factor influence mechanism can't be effectively shared, resulting in the failure to quantitatively differentiate between the factors of the role of the mechanism, while the unclear underlying mechanisms obstruct the fundamental revelation of the mapping relationship between shaving-induced mid-profile concavity error and tooth profile errors. To address this, this study derives computational formulas for the improved gear shaving tooth profile cutting depth error and shaving allowance, solving the gear shaving tooth profile cutting depth error and shaving allowance by the shaving cutting parameter (radial feed), and analyzing the influence law of radial feed on the gear shaving tooth profile cutting depth error and shaving allowance. Experimental results demonstrate that when the degree of overlap ranges from 1.3 to 1.9, and the axis intersection angle error falls within 0.2°~0.5°, selecting a radial feed of 44 ~ 50 μm simultaneously reduces tooth profile middle-concave error and tooth profile errors in workpiece gears. To resolve the lack of factor correlation, the Neo4j graph database was employed to construct a knowledge graph of influencing factors for shaving-induced mid-profile concavity error and tooth profile errors. This framework enables unified management of shaving error factors, the expression of correlations between knowledge entities, and relevant search and reasoning. It effectively resolves the issue of missing intrinsic connections among shaving error factors, realizes efficient retrieval and reasoning of knowledge regarding the intrinsic mechanism through which shaving error factors induce gear shaving errors, and thereby better facilitates the solution to tooth profile errors in shaving processes.