Scholarlink

<p>This study aimed to enhance ultimate tensile strength (UTS) and impact strength (IS) for dissimilar AISI 304 stainless steel and AISI 1018 mild steel joint using vibration assisted robotic gas metal arc welding (GMAW)&#xa0;process. Three variant levels were employed to five input parameters namely, current, voltage, welding speed, gas flow rate, and amplitude. 32 experiments were designed and carried out based on central composite design in response surface methodology. According to the experimental investigation, maximum UTS of 517.67 MPa was achieved at current 85 A, voltage 30 V, gas flow rate 8 L/min, welding speed 5 mm/sec, and amplitude 0.6 mm; and maximum IS of 51.92 J was obtained at current 100 A, voltage 26 V, gas flow rate 6 L/min, welding speed 7 mm/sec, and amplitude 0.4 mm. A hybrid ANN-GA model was constructed to forecast and optimize the response variables and the corresponding control parameters. The ANN modeling had best predicted responses with an MSE of 0.006 and coefficient of regression R of 0.98805 for normalized data and an MSE of 214.44 on original data for a 5-13-2 configuration back-propagation network using Bayesian Regularization algorithm. The optimized GA has resulted UTS and IS of 579.2 MPa and 68.69 J respectively. From the confirmation test, an average responses of UTS 556.7 MPa and IS 71 J were obtained with error percentages of 4.05% and 3.24% respectively.</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

Enhancing tensile and impact strengths in vibration-assisted robotic GMAW of dissimilar AISI 304 stainless steel and AISI 1018 mild steel using hybrid ANN-GA optimization

  • 期刊论文

  • 发表日期：

    2026年4月7日

查看全文
  • Samuel Belete Asfaw,
  • Teshome Mulatie Bogale,
  • Assefa Asmare Tsegaw,
  • Addisu Negash Ali

摘要

This study aimed to enhance ultimate tensile strength (UTS) and impact strength (IS) for dissimilar AISI 304 stainless steel and AISI 1018 mild steel joint using vibration assisted robotic gas metal arc welding (GMAW) process. Three variant levels were employed to five input parameters namely, current, voltage, welding speed, gas flow rate, and amplitude. 32 experiments were designed and carried out based on central composite design in response surface methodology. According to the experimental investigation, maximum UTS of 517.67 MPa was achieved at current 85 A, voltage 30 V, gas flow rate 8 L/min, welding speed 5 mm/sec, and amplitude 0.6 mm; and maximum IS of 51.92 J was obtained at current 100 A, voltage 26 V, gas flow rate 6 L/min, welding speed 7 mm/sec, and amplitude 0.4 mm. A hybrid ANN-GA model was constructed to forecast and optimize the response variables and the corresponding control parameters. The ANN modeling had best predicted responses with an MSE of 0.006 and coefficient of regression R of 0.98805 for normalized data and an MSE of 214.44 on original data for a 5-13-2 configuration back-propagation network using Bayesian Regularization algorithm. The optimized GA has resulted UTS and IS of 579.2 MPa and 68.69 J respectively. From the confirmation test, an average responses of UTS 556.7 MPa and IS 71 J were obtained with error percentages of 4.05% and 3.24% respectively.