Physics-driven precision reliability prediction modelling for the condition estimation of worm gear turntable

Abstract

The worm gear turntable is a crucial rotary indexing component for machining sculptured surfaces and plays a vital role in modern precision manufacturing. However, the limited availability of data during the design phase and the challenges posed by multi-source uncertainties in complex mechanical structures hinder the effectiveness of traditional physics-based precision reliability studies. To address these issues, this paper proposes a novel physics-driven precision reliability prediction model for worm gear turntables, accounting for multiple random and time-variant errors. Firstly, the accumulation process of initial random errors is explored using the meta-action method, and a precision model is constructed to estimate the initial condition of the worm gear turntable. Next, the mechanism by which time-variant factors reduce precision reliability is explained. Specifically, wear errors, derived from a wear model, and thermal errors, obtained through a numerical model, are incorporated into the precision model to establish a comprehensive precision reliability prediction framework. Finally, a precision reliability experiment is conducted to generate a test dataset for validation. A comparison of theoretical predictions and experimental results demonstrates the accuracy and effectiveness of the proposed approach.