Prediction of hardness in HAZ of multi-pass welds in low-alloy steel using Neural Network

Abstract

In temper bead welding, hardness is one of the key criteria to evaluate the tempering effect. A neural network-based method for hardness prediction in the heat affected zone (HAZ) of low-alloy steel has been investigated in the present study to evaluate the tempering effect in temper bead welding. The new hardness prediction system was constructed by using a neural network based on the experimentally obtained hardness database. On the basis of the thermal cycles numerically obtained by FEM, hardness distribution in HAZ of low alloy steel welded with temper bead welding method was calculated. The predicted hardness was in good accordance with the experimental results. It follows that our new prediction system is effective for estimating the tempering effect in HAZ during temper bead welding and hence enables us to assess the effectiveness of temper bead welding.