Scholarlink

<p>In the present work, a hybrid model based on combination of machine learning and theoretical micromechanical model has been developed to predict the fatigue life of the sample machined by ultrasonic assisted turning (UAT) process. Here, firstly the UAT factors viz. vibration amplitude in different directions, tool flank wear, cutting speed and feed rate were correlated to surface integrity aspects (SIAs) viz. hardness, residual stress, grain size and roughness using artificial neural network (ANN). Then, the ANN modeled SIAs were used as input to predict the fatigue life using crack propagation-based fatigue model namely Navaro Rios (NR). The developed model was confirmed taking into account low cycle fatigue life of machined Inconel 718 using the fully reversal rotating bending and axial fatigue tests. Comparing the modeled and predicted values, it was found that the developed hybrid model is more matched with the experimental fatigue tests measured by the rotating-bending than axial tests as the former is based on the gradient loading where the surface integrity have more determining effect on it. Also, it was found from the results that the vibration amplitude in Z (lateral) direction has the most significant impact on fatigue life followed by feed rate, vibration amplitude in Y direction (longitudinal), also, the impact of cutting velocity and tool flank wear were not found very deterministic. The application of vibration in Z direction corresponds to more compressive residual stress distribution, harder surface, finer grain size, and less roughness which their combination yields further fatigue life. In addition, it was found from the sensitivity analysis that the residual stress, followed by roughness (Rsm and Rt), grain size and hardness have greatest impact on determining the fatigue life. It is worth to add that ironing and peening effects of ultrasonic assisted turning process in lateral and longitudinal directions were found as underlying mechanisms of SIAs and subsequently fatigue life enhancement were former corresponds to better surface finish, while the latter yields more compressive residual stress, finer grain size and generation of hardened layer.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Effect of direction of vibration on fatigue life of Inconel 718 in single and multi-directional ultrasonic assisted machining: Machine learning augmented theoretical model

  • 期刊论文

  • 发表日期：

    2025年11月15日

查看全文
  • Naif Alharbi

摘要

In the present work, a hybrid model based on combination of machine learning and theoretical micromechanical model has been developed to predict the fatigue life of the sample machined by ultrasonic assisted turning (UAT) process. Here, firstly the UAT factors viz. vibration amplitude in different directions, tool flank wear, cutting speed and feed rate were correlated to surface integrity aspects (SIAs) viz. hardness, residual stress, grain size and roughness using artificial neural network (ANN). Then, the ANN modeled SIAs were used as input to predict the fatigue life using crack propagation-based fatigue model namely Navaro Rios (NR). The developed model was confirmed taking into account low cycle fatigue life of machined Inconel 718 using the fully reversal rotating bending and axial fatigue tests. Comparing the modeled and predicted values, it was found that the developed hybrid model is more matched with the experimental fatigue tests measured by the rotating-bending than axial tests as the former is based on the gradient loading where the surface integrity have more determining effect on it. Also, it was found from the results that the vibration amplitude in Z (lateral) direction has the most significant impact on fatigue life followed by feed rate, vibration amplitude in Y direction (longitudinal), also, the impact of cutting velocity and tool flank wear were not found very deterministic. The application of vibration in Z direction corresponds to more compressive residual stress distribution, harder surface, finer grain size, and less roughness which their combination yields further fatigue life. In addition, it was found from the sensitivity analysis that the residual stress, followed by roughness (Rsm and Rt), grain size and hardness have greatest impact on determining the fatigue life. It is worth to add that ironing and peening effects of ultrasonic assisted turning process in lateral and longitudinal directions were found as underlying mechanisms of SIAs and subsequently fatigue life enhancement were former corresponds to better surface finish, while the latter yields more compressive residual stress, finer grain size and generation of hardened layer.