Reliability-based design optimization of key components in the ball screw feed system

Abstract

As a critical component in high-precision mechanical transmission systems, the design of ball screws directly affects transmission accuracy and overall system stability. Aiming at the shortcomings of traditional deterministic design methods in processing machining errors and material uncertainties, this paper proposes a reliability‑based design optimization (RBDO) for a ball screw feed system (BSFS). The proposed multi‑objective framework maintains machining accuracy while minimizing system mass and maximizing dynamic performance, subject to nonlinear stiffness reliability constraints. Computational efficiency is improved by employing a Kriging surrogate model to approximate complex numerical calculations. A global optimum is identified by integrating an efficient meta‑heuristic algorithm. The proposed dynamic model is verified by static loading tests, modal hammering tests, and vibration response experiments. The numerical simulation demonstrates that, compared to traditional deterministic design, system stiffness reliability is increased from 74.25 % to 93.51 %, and system mass is reduced by about 8 %, while the natural frequency remains nearly unchanged. The potential value of this research is to guide the optimization of BSFS at the design stage.