| 陈升,刘烨,景国玺,等.基于反向传播神经网络的增压器涡轮转子过盈连接结构优化设计[J].内燃机工程,2024,45(1):10-17. |
| 基于反向传播神经网络的增压器涡轮转子过盈连接结构优化设计 |
| Optimization Design of the Interference Connection Structure for Turbocharger Turbine Rotors Based on Back Propagation Neural Network |
| DOI：10.13949/j.cnki.nrjgc.2024.01.002 |
| 关键词:钛铝合金 涡轮 过盈连接 结构优化 |
| Key Words:titanium–aluminum alloy turbine interference connection structural optimization |
| 基金项目:国防科技重点实验室基金项目（61422120301） |
|
| 摘要点击次数: 1125 |
| 全文下载次数: 465 |
| 摘要:针对钛铝合金涡轮与K418合金套的过盈连接结构存在的应力问题，以涡轮转子过盈连接结构为研究对象，设计型线结构并建立过盈连接结构应力预测模型，基于过盈连接结构应力预测模型分析不同型线对应力分布状态的影响。研究结果表明：过盈连接结构应力预测模型具有较高的精度，对平均接触应力的预测误差为0.67%，对最大Mises应力的预测误差为8.25%。采用过盈连接结构应力预测模型得到的型线优化设计方案使涡轮轴最大Mises应力减少了42.01%，而平均接触应力只减少了10.22%，增压器涡轮转子过盈连接结构可靠性显著提高。 |
| Abstract:To solve the stress problems existing in the interference connection structure of the titanium-aluminum alloy turbine and the K418 alloy sleeve， the turbine rotor interference connection structure was taken as the research object. A profile and a stress prediction model of the interference connection structure were designed and established. And the influences of different profiles on the stress distribution state were analyzed based on the stress prediction model of the interference connection structure. The research results show that the prediction error of the interference connection structure stress prediction model for average contact stress is 0.67%， and the prediction error for the maximum Mises stress is 8.25%， which has high accuracy. The optimized design scheme of the profile obtained by using the interference connection structure stress prediction model reduces the maximum Mises stress of the turbine shaft by 42.01%， while the average contact stress only decreases by 10.22%. The reliability of the interference connection structure of the turbocharger turbine rotor is significantly improved. |
|
查看全文 HTML
查看/发表评论 下载PDF阅读器 |