DEVELOPMENT OF INCOMPLETE PENETRATION PREDICTIVE MODELS USING RESPONSE SURFACE METHODOLOGY AND ARTIFICIAL NEURAL NETWORK

Authors

  • HENRY ODION OJIKA
  • JOSEPH ACHEBO
  • ANDREW OZIGAGUN

Keywords:

incomplete penetration, prediction, tungsten inert gas welding, artificial neural network, mild steel

Abstract

This paper seeks for forecasting partial penetration inside tungsten inert gas welding utilizing reaction surface technique plus artificial neural network. It handles the statistical approach of main composite design, Analysis of Variance (ANOVA), surface plots as well as cooks distance and also the quasi newton neural network to examine along with forecast the reaction aim. The outcome extracted from the all model suggests that the design predicts the partial penetration adequately. Even so the artificial neural network is a greater predictive model with least mean square error. The statistical method for prediction used have discovered boosting applications in an assortment of areas of sciences and engineering, the techniques outlined in this paper for incomplete penetration prediction is able to eradicate the requirement for performing experiments on the groundwork of the conventional technique which happens to be time consuming and economically not justifiable.

References

Azar, A.S., Ås, S.K., Akselsen, O.M. (2012): Determination of welding heat source parameters from actual bead shape. – Computational materials science 54: 176-182.

Chan, B., Pacey, J., Bibby, M. (1999): Modelling gas metal arc weld geometry using artificial neural network technology. – Canadian Metallurgical Quarterly 38(1): 43-51.

Dey, V., Pratihar, D.K., Datta, G.L., Jha, M.N., Saha, T.K., Bapat, A.V. (2009): Optimization of bead geometry in electron beam welding using a genetic algorithm. – Journal of Materials Processing Technology 209(3): 1151-1157.

Gunaraj, V., Murugan, N. (1999): Application of response surface methodology for predicting weld bead quality in submerged arc welding of pipes. – Journal of materials processing technology 88(1-3): 266-275.

Kim, I.S., Son, J.S., Kim, I.G., Kim, J.Y., Kim, O.S. (2003a): A study on relationship between process variables and bead penetration for robotic CO2 arc welding. – Journal of Materials Processing Technology 136(1-3): 139-145.

Kim, I.S., Son, K.J., Yang, Y.S., Yaragada, P.K.D.V. (2003b): Sensitivity analysis for process parameters in GMA welding processes using a factorial design method. – International Journal of Machine Tools and Manufacture 43(8): 763-769.

Mostafa, N.B., Khajavi, M.N. (2006): Optimization of welding parameters for weld penetration in FCAW. – Journal of achievements in materials and manufacturing engineering 16(1-2): 132-138.

Murugan, N., Parmar, R.S. (1994): Effects of MIG process parameters on the geometry of the bead in the automatic surfacing of stainless steel. – Journal of Materials Processing Technology 41(4): 381-398.

Raveendra, J., Parmar, R.S. (1987): Mathematical models to predict weld bead geometry for flux cord arc welding. – Metal Construction 19(1): 31R-35R.

Shumovsky, A. (1952): Controlling welding shrinkage and distortion. – The Canadian Welder 2p.

Downloads

Published

2021-02-16

How to Cite

OJIKA, H. O., ACHEBO, J., & OZIGAGUN, A. (2021). DEVELOPMENT OF INCOMPLETE PENETRATION PREDICTIVE MODELS USING RESPONSE SURFACE METHODOLOGY AND ARTIFICIAL NEURAL NETWORK. Quantum Journal of Engineering, Science and Technology, 2(1), 1–9. Retrieved from https://qjoest.com/index.php/qjoest/article/view/17

Issue

Section

Articles