Open Access Open Access  Restricted Access Subscription Access

Estimation of Geometry and Properties of Weld Bead Using Artificial Neural Networks


Affiliations
1 Department of Mechanical Engineering, Kalyani Government Engineering College, Kalyani-741235, India
 

Investigations on prediction or estimation of output responses in welding using artificial neural networks (ANN) have become popular among researchers. In this work, metal active gas (MAG) welding was implemented to join EN-8D medium carbon steel plates together by varying welding current, weld voltage and torch traverse speed as input parameters. Depth of penetration, reinforcement, hardness and bend angle at failure were the responses. Then the input parameters and output parameters are used to train the neural networks in this work. Feed forward network with Levenberg-Marquardt training function is implemented. 3-10-4 model of ANN is used for the prediction of depth of penetration, reinforcement, hardness and bend angle. From the regression chart, it is found that the designed model predicted the results of both replications with quite less error and hence, the effectiveness of the technique.

Keywords

GMAW; ANN; Neural network; MAG welding; Modeling; MATI_AB
User
Notifications
Font Size


  • Khanna, O. R, A Text Book of Welding Technology, Dhanpat Rai Publications, New Delhi, 2001.
  • Nadkarni, S.V., Modern Arc Welding Technology, Oxford & IBM Publishing Co. Pvt. Ltd., New Delhi.
  • Bera, T, The History of Development of Gas Metal Arc Welding process, Indian Science Cruiser, Vol 34, No 4, pp. 64-66, 2020.
  • Chan, B., Pacey, J. and Bibby, M., Modelling Gas Metal Arc Weld Geometry Using Artificial Neural Network Technology, Journal of Canadian Metallurgical Quarterly, Vol. 38, pp. 43-51,1999.
  • Lee, J. and Um, K., A Comparison in a Back-Bead Prediction of Gas Metai Arc Weiding Using Muitipie Regression Anaiysis and Artificiai Neurai Network, Journai of Optics and Lasers in Engineering, Voi. 34, pp. 149-158,2000.
  • Sreeraj, P, Kannan, T. and Maji, S., Simuiation and Parameter Optimization of GMAW Process Using Neurai Networks and Particle Swarm Optimization Algorithm, Internationai Journai of Mechanicai Engineering and Robotic Research, Voi 2, pp. 131-146,2013.
  • Shah, J., Patei, G. and Makwana, J., Optimization and Prediction of MiG Weiding Process Parameters Using ANN, Internationai Journai of Engineering Deveiopment and Research, Voi. 5, pp. 1487-1491,2017.
  • Sreeharan, B.N., Kannan, T. and Aravind, P, Process Optimization of GMAW over AA6351 Aiuminium Aiioy Using ANN, Voi. 8, pp. 208-218,2017.
  • Ates, H., Prediction of Gas Metai Arc Welding Parameters Based on Artificiai Neural Networks, Materials and Design, Vol. 28, pp. 2015-2023,2007.
  • Pal, S., Pal, S.K. and Samantaray, A.K., Artificial Neural Network Modeling of Weid Joint Strength Prediction of a Puised Metai Inert Gas Welding Process Using Arc Signais, Journai of Materiais Processing Technoiogy, Voi. 202, pp. 464-474,2008.
  • Singh, V., Chandrasekaran, M. and Thirugananasambandam, M., Artificial Neural Network Modelling of Weid Bead Characteristics during GMAW of Nitrogen Strengthened Austenitic Stainiess Steei, AlP Conference Proceedings, 2128, 020024,2019.
  • Nagesh, D.S. and Datta, G.L., Modeiing of Fiiiet Weided Joint of GMAW Process: integrated Approach Using DOE, ANN and GA, International Journal on Interactive Design Manufacturing, Vol. 2, pp. 127-136, 2008.
  • Casalino, G., Hu, S.J. and Hou, W., Deformation Prediction and Quality Evaluation of the Gas Metal Arc Welding Butt Weld, Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture, Vol. 217, No. 11, pp. 1615-1622,2003.
  • Nagesh, D.S. and Datta, G.L, Prediction of Weld Bead Geometry and Penetration in Shielded Metal-Arc Welding Using Artificial Neural Networks, Journal of Materials Processing Technology, Vol. 123, No. 2, pp. 303-312,2002.
  • Bera, T. and Das, S., Application of Artificial Neural Networks in Predicting Output Parameters of Gas Metal Arc Welding of Dissimilar Steels, Indian Science Cruiser, Vol. 35, No. 3, pp. 26-30,2021.
  • Kanti, K.M. and Rao, P.S., Prediction of Effect of Welding Process Parameter of MIG Process on Weld Bead Geometry, Journal of Materials Processing Technology, Vol. 200, No. 1-3, pp. 300- 305,2008.
  • Das, A. and Das, S., Prediction of Bead Geometry of Gas Metal Arc Welded Workpiece Using Artificial Neural Networks, Proceedings of the National Seminar on 'Welding Science and Technology- Present Status & Future Direction' (NSWEST 2021), July 23-24 2021, pp.79-80,2021.
  • Saha, M.K., Dhara, L.N. and Das, S., Variation of Bead Geometry of 316 Austenitic Stainless Steel Weld with Varying Heat Input Using Metal Active Gas Welding, Recent Advances in Mechanical Engineering, 2021.
  • Saha, M.K., Hazra, R., Mondal, A. and Das, S., Effect of heat input on geometry of austenitic stainless steel weld bead on low alloy steel. Journal of the Institution of Engineers (India), Series C, Vol. 100, No.4, pp. 607-615,2019.
  • Sarkar, A. and Das, S., Selection of Appropriate Process Parameters for Gas Metal Arc Welding of a Steel under 100% Carbon Dioxide Gas Shield, Indian Welding Journal, Vol.49, No.4, pp.61-70,2016.
  • Ramos-Jaime, D. and Lopez-Juarez, I., ANN and Linear Regression Model Comparison for the Prediction of Bead Geometrical Properties in Automated Welding, 1st International Congress on Instrumentation and Applied Science, pp. 1-10,2010.
  • Sabiruddin, K., Das, S. and Bhattacharya, A., Application of the analytic hierarchy process for optimisation of process parameters in GMAW, Indian Welding Journal, Vol.42, No.1, pp.38-46,2009.
  • Das, A. and Das, S., Tungsten Arc Welded Workpiece Using Artificial Neural Networks, Proceedings of the National Welding Meet (NWM 2021), Tiruchirapalli, October 07-08 2021.
  • Das, S., Roy R. and Chattopadhyay, A.B., Evaluation of Wear of Turning Carbide Inserts Using Neural Networks, International Journal of Machine Tools and Manufacture, Vol.36, No.7, pp.789-797, 1996.
  • Das, S., Bandyopadhyay, P.P. and Chattopadhyay, A.B., Neural-Networks- Based Tool Wear Monitoring in Turning Medium Carbon Steel Using a Coated Carbide Tool, Journal of Materials Processing Technology, Vol.63, No.1-3, pp. 187-192,1997.
  • Das, S., On Wear Monitoring of TIN Coated Tools- Part II: With Neural Networks, Proceedings of the 14th International Conference on Robotics and Factories of the Future (CAR& FOF1998), Coimbatore, India, pp.645-653,1998.
  • Mukherjee, A. and Das, S., A Simple Online Tool Condition Monitoring System Using Artificial Neural Networks, lOP Conf. Series; Materials Science and Engineering, Vol.1080,No.012021,2021.
  • Kartik, C.S., Suryaganesh, G., Joshi, N.R., Ghanta, K.C. and Das, S., Application of Neural Networks to an Esterification Process, Proceedings of the IIChE Golden Jubilee Congress (CHEMCON-1997), New Delhi, pp.1019-1029,1997.
  • Ghanta, K.C. and Das, S., Neural Networks Based Modeling of Viscosity for Facilitating Transportation of Magnetite Ore-Water Slurry, Journal of the Association of Engineers, India, Vol.83, No.2, pp.43-54, 2013.
  • Misra, D., Das, S., Mondal, N. and Saha, P.P., Estimation of Drilling Burr Formation with Artificial Neural Network Analysis, Indian Science Cruiser, Vol.34, No.3, pp.23-31,2020.
  • Lera, G. and Pinzolas, M., Neighborhood Based Levenberg-Marquardt Algorithm for Neural Network Training, IEEE Transactions on Neural Networks, Vol. 13, No. 5, pp. 1200-1203,2002.

Abstract Views: 93

PDF Views: 23




  • Estimation of Geometry and Properties of Weld Bead Using Artificial Neural Networks

Abstract Views: 93  |  PDF Views: 23

Authors

Tapas Bera
Department of Mechanical Engineering, Kalyani Government Engineering College, Kalyani-741235, India
Santanu Das
, India

Abstract


Investigations on prediction or estimation of output responses in welding using artificial neural networks (ANN) have become popular among researchers. In this work, metal active gas (MAG) welding was implemented to join EN-8D medium carbon steel plates together by varying welding current, weld voltage and torch traverse speed as input parameters. Depth of penetration, reinforcement, hardness and bend angle at failure were the responses. Then the input parameters and output parameters are used to train the neural networks in this work. Feed forward network with Levenberg-Marquardt training function is implemented. 3-10-4 model of ANN is used for the prediction of depth of penetration, reinforcement, hardness and bend angle. From the regression chart, it is found that the designed model predicted the results of both replications with quite less error and hence, the effectiveness of the technique.

Keywords


GMAW; ANN; Neural network; MAG welding; Modeling; MATI_AB

References





DOI: https://doi.org/10.21843/reas%2F2021%2F46-56%2F212373