Optimization of Flexible Rotor for Ultrasonic Motor Based on Response Surface and Genetic Algorithm

Bo Chen, Juntao Yang, Haoyu Tang, Y-s Wu, Haoran ZhangMicromachines 2024 · DOI

TL;DR

This paper introduces a novel optimization approach for the flexible rotor of traveling wave rotary ultrasonic motors (TRUMs). By combining Kriging response surface modeling with a multi-objective genetic algorithm (MOGA), they significantly improved rotor-stator contact uniformity and reduced stress, leading to enhanced motor performance and durability.

What Problem Does This Paper Address?

Traveling wave rotary ultrasonic motors (TRUMs) are known for their precision and quiet operation, making them valuable in various industries. A key component is the flexible rotor, which helps reduce radial friction. However, existing designs often suffer from uneven contact between the stator and rotor, leading to rotor deformation, increased stress, and ultimately, reduced performance and durability. Traditional methods for optimizing rotor dimensions are limited and can result in suboptimal designs. This paper aims to address these issues by developing a more effective method for optimizing the flexible rotor's design to ensure uniform contact and minimize stress.

Methods

The researchers propose a two-pronged approach:

  • Kriging Response Surface Model: To handle the complex interactions and non-linearities involved in rotor-stator contact, they first construct a Kriging response surface model. This statistical technique is adept at modeling complex relationships from limited data. They used Latin Hypercube Sampling (LHS) to generate a comprehensive and uniformly distributed set of design points for building this model. LHS is a more efficient sampling method than random sampling, ensuring better coverage of the design space with fewer points.
  • Multi-Objective Genetic Algorithm (MOGA): Once the response surface model is established, a Multi-Objective Genetic Algorithm (MOGA) is employed to optimize the rotor design. MOGA leverages principles of natural selection (crossover and mutation) to explore the design space and find optimal solutions. It's particularly useful for multi-objective problems, aiming to balance competing goals (in this case, uniform contact and minimal stress). The MOGA was implemented using ANSYS Workbench and Design Expert software.

Datasets

The study utilized experimental data generated from a finite element model of the TRUM60 ultrasonic motor. An orthogonal experimental design was performed, resulting in 147 sets of experimental data points. This data was used to train the Kriging response surface model. No new public benchmark datasets were introduced.

Key Insights and Ideas

  • The Rotor Matters: While much research focuses on stator design, this paper emphasizes the critical role of the flexible rotor in achieving optimal TRUM performance.
  • Kriging + MOGA Synergy: The combination of Kriging for modeling complex interactions and MOGA for robust multi-objective optimization proved highly effective in finding a superior rotor design.
  • Uniformity is Key: The primary optimization objective (M1) was focused on achieving uniform contact between the stator and rotor, directly addressing the root cause of many performance issues.
  • Stress Reduction is Significant: The optimization not only improved contact uniformity but also drastically reduced the maximum contact stress on the rotor, a crucial factor for durability.
  • Experimental Validation is Crucial: The paper goes beyond simulation by experimentally validating the optimized design, demonstrating its real-world effectiveness through wear pattern analysis.

Key Results

The optimization process yielded significant improvements:

  • Objective Function Value (M1): Decreased from an initial value of 0.631 to 0.036 (for Candidate 1), indicating a substantial improvement in contact uniformity.
  • Maximum Contact Stress (M2): Reduced from 32.77 MPa to 9.96 MPa (for Candidate 1), a nearly 70% reduction.
  • Experimental Wear Patterns: Visual comparison of wear patterns before and after optimization clearly showed a more uniform and shallower wear distribution on the optimized rotor, confirming improved contact uniformity and reduced localized stress.

Limitations and Future Work

The authors acknowledge the following limitations and suggest future research directions:

  • Model Specificity: The current structural design and optimization are limited to the TRUM60-type ultrasonic motor. Future work could explore the applicability of this approach to other motor models.
  • Microscopic Contact Behavior: The study did not investigate the microscopic contact behavior under ultrasonic frequencies. Future research will focus on microscopic analysis to address this gap.
  • Further Exploration: While the current structural design is improved, further research could explore other potential design variations and their impact on performance.

Summary updated

Abstract

The flexible rotor, as a crucial component of the traveling wave rotary ultrasonic motor, effectively reduces radial friction. However, issues such as uneven contact between the stator and rotor, as well as rotor-deformation-induced stress, still persist. This paper presents an optimization method that combines the Kriging response surface model with a multi-objective genetic algorithm (MOGA). Drawing on the existing rotor structure, a novel rotor design is proposed to match the improved TRUM60 stator. During the optimization process, the contact surface between the stator and rotor is taken as the optimization target, and an objective function is established. The Kriging response surface model is constructed using Latin hypercube sampling, and an MOGA is employed to optimize this model, allowing the selection of the optimal balanced solution from multiple candidate designs. Following stator optimization, the objective function value decreased from 0.631 to 0.036, and the maximum contact stress on the rotor inner ring was reduced from 32.77 MPa to 9.96 MPa. Experimental validation confirmed the reliability of this design, significantly improving the overall performance and durability of the motor.
BibTeX
@inproceedings{chen2024optimization,
  author = {Bo Chen and Juntao Yang and Haoyu Tang and Y-S Wu and Haoran Zhang},
  title = {{Optimization of Flexible Rotor for Ultrasonic Motor Based on Response Surface and Genetic Algorithm}},
  booktitle = {Micromachines},
  year = {2024},
  doi = {10.3390/mi16010054},
}
Figure