Prediction of HAZ width and toughness of HY85 steel using artificial neural network

ABSTRACT

Physical weld simulation of single pass welding of HY 85 steel was performed using Gleeble® 3800 thermo-mechanical simulator at a peak temperature of 1300ºCwith heating rate of 200ºCs⁻¹, and cooled with cooling rates ranging from 0.5 to 120ºCs⁻¹ and corresponding heat inputs ranging from 1137 to 4.7 kJ/cm. The impact toughness of coarse grain heat-affected (CGHAZ) zone of simulated HAZ specimens was determined at a temperature of −50°C and related with their microstructures. Back propagation neural network (BPN) model was used to model heat-affected zone width and impact toughness of simulated heat-affected zone specimens, using heat input and cooling rate as input parameters. Predicted data of HAZ width and impact toughness is compared with the experimental data and it is observed that the back propagation neural network model with 4 hidden neurons adequately predicts the impact toughness and HAZ width. Calculated relative percentage error for HAZ width is ±3.15, whereas for impact toughness it is ±7.93%.

KEYWORDS:

• Weld heat affected zone
• impact toughness
• feed forward back propagation
• artificial neural network

Acknowledgments

The authors show gratitude to Indian Institute of Technology, Roorkee and Bundelkhand Institute of Engineering and Technology, Jhansi for its support received to complete the whole work.

Disclosure statement

No potential conflict of interest was reported by the authors.