基于GA-BP神经网络的G20CrNi2MoA轴承套圈渗碳热处理工艺优化

杨智勇, 陶源彪, 陈奕帆, 王馨冉, 杜嘉政, 李志强

清华大学学报(自然科学版) ›› 2026, Vol. 66 ›› Issue (2) : 335-345.

PDF(12058 KB)
PDF(12058 KB)
清华大学学报(自然科学版) ›› 2026, Vol. 66 ›› Issue (2) : 335-345. DOI: 10.16511/j.cnki.qhdxxb.2026.27.013
机械工程

基于GA-BP神经网络的G20CrNi2MoA轴承套圈渗碳热处理工艺优化

作者信息 +

Optimization of carburizing heat treatment process for G20CrNi2MoA bearing ring based on genetic algorithm-backpropagation neural network

Author information +
文章历史 +

摘要

渗碳钢轴承的滚动接触疲劳寿命与渗碳层硬度及其分布密切相关,而渗碳热处理工艺决定了渗碳层的硬度分布,进而影响渗碳层深度。为延长高速列车轴箱轴承的滚动接触疲劳寿命,该文以国产G20CrNi2MoA轴承外圈为研究对象,结合试验与热处理仿真技术,系统研究了渗碳层硬度分布对外圈滚动接触疲劳性能的影响规律。首先通过滚动接触疲劳试验明确了适中的渗碳层深度是延长外圈滚动接触疲劳寿命的关键;进而结合遗传算法(GA)与BP神经网络,优化并获得了使套圈具有高滚动接触疲劳性能的渗碳层硬度曲线,其关键参数为:表层硬度693 HV,渗碳层深度1.71 mm;基于此,通过对套圈渗碳热处理工艺的GA-BP神经网络优化,提出了一种旨在实现轴承外圈长寿命的渗碳热处理工艺优化方案,并通过G20CrNi2MoA轴承外圈的渗碳热处理试验验证了其可行性,其关键工艺参数为:强渗时间16.5 h、扩散温度930 ℃、扩散时间6.54 h、扩散碳势1.05%、等温时间3.6 h。渗碳热处理工艺优化后,轴承外圈的滚动接触疲劳寿命比现有轴承外圈的寿命延长约4.7%。

Abstract

Objective: In recent years, China's high-speed train industry has developed rapidly. As a key core component, the process of localizing axle box bearings is limited by insufficient fatigue performance. At present, high-speed train bearings rely entirely on imports, and fatigue of domestically manufactured bearing rings has become an urgent problem that must be addressed. The fatigue performance of axle box bearing rings is mainly determined by the metallurgical quality of the bearing steel and the heat treatment process. G20CrNi2MoA steel is a high-quality carburized bearing steel for manufacturing the inner and outer rings of this type of bearing. Although domestic smelting levels have gradually caught up with international levels, there is still a significant gap in carburizing heat treatment technology, and the hardness distribution of the carburized layer is the key factor determining the rolling contact fatigue life of carburized steel bearings. To address this dilemma, the outer ring of the domestic G20CrNi2MoA bearing is the core research object of this study, aiming to explore the influence of the hardness distribution of the carburized layer on the bearing's rolling contact fatigue performance and to optimize the process. Methods: A systematic research system of experimental verification, simulation, and intelligent optimization was constructed. First, the key factors affecting performance were determined through a rolling contact fatigue test, after which the formation mechanism of the carburized layer was analyzed via simulated heat treatment. Subsequently, a model combining a genetic algorithm (GA) and a backpropagation (BP) neural network was introduced to optimize the parameters. Results: Using this system, innovative results were achieved. The results confirm that a carburized layer with a moderate depth is key to extending the fatigue life of the outer ring, and the model provides a more accurate solution for the hardness distribution curve of the optimal carburized layer (surface hardness: 693 HV; depth of carburized layer: 1.71 mm). Based on this reverse optimization, a complete carburized heat treatment scheme is obtained. The key parameters include a long infiltration time of 16.5 h, a diffusion temperature of 930 ℃, a diffusion time of 6.54 h, a carbon diffusion potential of 1.05%, and an isothermal time of 3.6 h. Targeted tests verify that the rolling contact fatigue life of the outer ring of the bearing treated using this process is approximately 4.7% higher than that of the existing domestic bearing outer ring. Conclusions: First, there is a significant nonlinear correlation between the rolling contact fatigue performance of the domestic G20CrNi2MoA bearing outer ring and the hardness distribution of the carburized layer, the depth of the carburized layer, and other parameters. Accurately regulating these parameters is key to improving performance. Second, the model combining a GA and a BP neural network provides an efficient and accurate technical approach for optimizing the carburizing heat treatment of bearings, thereby greatly reducing the cost and cycle time of traditional trial-and-error methods. Third, the carburizing heat treatment process proposed in this study effectively improves the fatigue performance of domestic bearing rings. It provides important support, with both theoretical value and practical guiding significance, for addressing the technical gap in the heat treatment of high-speed bearings in China, breaking the monopoly of foreign technology, and promoting the localization of high-speed train bearings. It also lays a foundation for subsequent research and the development of higher-speed grade bearings.

关键词

渗碳钢轴承套圈 / 滚动接触疲劳寿命 / 渗碳层硬度曲线 / 渗碳层深度 / 热处理工艺

Key words

carburized steel bearing rings / rolling contact fatigue life / hardness curve of the carburized layer / carburized layer depth / heat treatment process

引用本文

导出引用
杨智勇, 陶源彪, 陈奕帆, . 基于GA-BP神经网络的G20CrNi2MoA轴承套圈渗碳热处理工艺优化[J]. 清华大学学报(自然科学版). 2026, 66(2): 335-345 https://doi.org/10.16511/j.cnki.qhdxxb.2026.27.013
Zhiyong YANG, Yuanbiao TAO, Yifan CHEN, et al. Optimization of carburizing heat treatment process for G20CrNi2MoA bearing ring based on genetic algorithm-backpropagation neural network[J]. Journal of Tsinghua University(Science and Technology). 2026, 66(2): 335-345 https://doi.org/10.16511/j.cnki.qhdxxb.2026.27.013
中图分类号: TP183   

参考文献

1
李秋泽, 谌亮, 徐建新, 等. 高速动车组用滚动轴承失效模式及对策展望[J]. 轴承, 2024 (3): 1- 8.
LI Q Z , CHEN L , XU J X , et al. Failure modes of rolling bearings for high-speed EMUs and prospects for countermeasures[J]. Bearing, 2024 (3): 1- 8.
2
赵恒喆, 杨晓英, 汤洁, 等. 高速铁路动车组轴箱轴承故障分析[J]. 轴承, 2022 (7): 1- 8.
ZHAO H Z , YANG X Y , TANG J , et al. Fault analysis of axle box bearings for high speed railway EMUs[J]. Bearing, 2022 (7): 1- 8.
3
SADEGHI F , JALALAHMADI B , SLACK T S , et al. A review of rolling contact fatigue[J]. Journal of Tribology, 2009, 131 (4): 041403.
4
ROY S. Investigating the effect of retained austenite on wear and fatigue behavior of AISI 8620 steel[D]. Ames: Iowa State University, 2018.
5
ROY S , SUNDARARAJAN S . Effect of retained austenite on spalling behavior of carburized AISI 8620 steel under boundary lubrication[J]. International Journal of Fatigue, 2019, 119, 238- 246.
6
ZHANG S G , WANG W Z , ZHANG H B , et al. The effect of hardness distribution by carburizing on the elastic-plastic contact performance[J]. Tribology International, 2016, 100, 24- 34.
7
LORENZ S J , SADEGHI F , WANG C . Effect of spatial hardness distribution in rolling contact fatigue performance of bearing contacts[J]. Tribology International, 2022, 171, 107550.
8
景国荣, 曹广惠, 薛经仙, 等. G20CrNi2MoA渗碳后热处理工艺的研究[J]. 轴承, 1986 (2): 16-22, 63.
JING G R , CAO G H , XUE J X , et al. Research on heat treatment process of G20CrNi2MoA after carburizing[J]. Bearing, 1986 (2): 16-22, 63.
9
李明, 刘晓丽. 渗碳热处理的缺陷研究与对策[J]. 世界有色金属, 2017 (21): 274- 275.
LI M , LIU X L . Study on the defects of carburizing heat treatment and Countermeasures[J]. World Nonferrous Metals, 2017 (21): 274- 275.
10
邹胜超. G20CrNi2MoA轴承套圈热处理仿真优化与试验研究[D]. 北京: 北京交通大学, 2021.
ZOU S C. Simulation optimization and experimental study on heat treatment of G20CrNi2MoA bearing rings[D]. Beijing: Beijing Jiaotong University, 2021. (in Chinese)
11
HERNANDEZ-MORALES B , BARBA-MENDEZ O , INGALLS-CRUZ A , et al. Mathematical modelling of temperature and stress evolution during cooling of a stainless steel Navy C-ring probe[J]. International Journal of Materials & Product Technology, 2005, 24 (1-4): 306- 318.
12
LI Z C , GRANDHI R V , SRINIVASAN R . Distortion minimization during gas quenching process[J]. Journal of Materials Processing Technology, 2006, 172 (2): 249- 257.
13
KIM D W , CHO Y G , CHO H H , et al. A numerical model for vacuum carburization of an automotive gear ring[J]. Metals and Materials International, 2011, 17 (6): 885- 890.
14
ZHANG X , TANG J Y , ZHANG X R . An optimized hardness model for carburizing-quenching of low carbon alloy steel[J]. Journal of Central South University, 2017, 24 (1): 9- 16.
15
KHAN D , GAUTHAM B . Integrated modeling of carburizing-quenching-tempering of steel gears for an ICME framework[J]. Integrating Materials and Manufacturing Innovation, 2018, 7 (1): 28- 41.
16
ZHANG Y T , WANG G , SHI W K , et al. Optimization of process parameters for the axial distortion and distortion range during die quenching of a spiral bevel gear[J]. Journal of Materials Engineering and Performance, 2018, 27 (8): 3803- 3811.
17
GUO Y F , WANG C R , MA Z , et al. A new mesh smoothing method based on a neural network[J]. Computational Mechanics, 2022, 69 (2): 425- 438.
18
JIA Y F , FU R , LING C , et al. Fatigue life prediction based on a deep learning method for Ti-6Al-4V fabricated by laser powder bed fusion up to very-high-cycle fatigue regime[J]. International Journal of Fatigue, 2023, 172, 107645.
19
SHI T , SUN J Y , LI J H , et al. Machine learning based very-high-cycle fatigue life prediction of AlSi10Mg alloy fabricated by selective laser melting[J]. International Journal of Fatigue, 2023, 171, 107585.
20
KATOCH S , CHAUHAN S S , KUMAR V . A review on genetic algorithm: Past, present, and future[J]. Multimedia Tools and Applications, 2021, 80 (5): 8091- 8126.
21
RAZAVI S A , ASHRAFIZADEH F , FOOLADI S . Prediction of age hardening parameters for 17-4PH stainless steel by artificial neural network and genetic algorithm[J]. Materials Science and Engineering: A, 2016, 675, 147- 152.
22
KAMAL BABU K , PANNEERSELVAM K , SATHIYA P , et al. Parameter optimization of friction stir welding of cryorolled AA2219 alloy using artificial neural network modeling with genetic algorithm[J]. The International Journal of Advanced Manufacturing Technology, 2018, 94 (9-12): 3117- 3129.
23
ABERE A E , TSEGAW A A , NALLAMOTHU R B . Process parameters optimization of bobbin tool friction stir welding on aluminum alloy 6061-T6 using combined artificial neural network and genetic algorithm[J]. Journal of the Brazilian Society of Mechanical Sciences and Engineering, 2022, 44 (11): 566.
24
MENGISTIE A K , BOGALE T M . Development of automatic orbital pipe MIG welding system and process parameters' optimization of AISI 1020 mild steel pipe using hybrid artificial neural network and genetic algorithm[J]. The International Journal of Advanced Manufacturing Technology, 2023, 128 (5-6): 2013- 2028.
25
中华人民共和国国家质量监督检验检疫总局, 中国国家标准化管理委员会. 渗碳轴承钢: GB/T 3203-2016[S]. 北京: 中国标准出版社, 2016.
General Administration of Quality Supervision, Inspection and Quarantine of the People's Republic of China, National Standardization Administration. Carburizing steels for bearing: GB/T 3203-2016[S]. Beijing: Standards Press of China, 2016. (in Chinese)
26
中华人民共和国国家质量监督检验检疫总局, 中国国家标准化管理委员会. 钢件渗碳淬火硬化层深度的测定和校核: GB/T 9450-2005[S]. 北京: 中国标准出版社, 2006.
General Administration of Quality Supervision, Inspection and Quarantine of the People's Republic of China, National Standardization Administration. Steels—determination and verification of the depth of carburized and hardened cases: GB/T 9450-2005[S]. Beijing: Standards Press of China, 2006. (in Chinese)
27
中华人民共和国国家发展和改革委员会. 滚动轴承材料接触疲劳试验方法: JB/T 10510-2005[S]. 北京: 机械工业出版社, 2005.
National Development and Reform Commission. Test method for contact fatigue of rolling bearing material: JB/T 10510-2005[S]. Beijing: China Machine Press, 2005. (in Chinese)
28
YANG Z Y , WEI J L , LIU T , et al. Influence of carbon layer hardness characteristics on RCF performance of carburized bearing steel and optimization of heat treatment process[J]. Engineering Failure Analysis, 2024, 160, 108164.
29
PAVLINA E J , VAN TYNE C J . Correlation of yield strength and tensile strength with hardness for steels[J]. Journal of Materials Engineering and Performance, 2008, 17 (6): 888- 893.
30
ZAVADSKAS E K , TURSKIS Z . A new logarithmic normalization method in games theory[J]. Informatica, 2008, 19 (2): 303- 314.
31
邓攀科. 城轨车辆铝合金铸造牵引梁材料与成型工艺研究[D]. 北京: 北京交通大学, 2020.
DENG P K. Research on materials and forming technology of aluminum alloy cast traction beam for urban rail vehicle[D]. Beijing: Beijing Jiaotong University, 2020. (in Chinese)

基金

国家自然科学基金面上项目(52372345)
中央高校基本科研业务费专项基金(2024JBZY006)

版权

版权所有,未经授权,不得转载。
PDF(12058 KB)

Accesses

Citation

Detail

段落导航
相关文章

/