Artificial neural network prediction of weld distortion rectification using a travelling induction coil
Artificial neural network prediction of weld distortion rectification using a travelling induction coil
复制标题
DOI：
10.1007/s00170-012-4713-z
复制
发表时间：
2013-01
期刊：
影响因子：
--
通讯作者：
Colin Barclay;S. Campbell;A. Galloway;N. McPherson
中科院分区：
文献类型：
--
作者：
Colin Barclay;S. Campbell;A. Galloway;N. McPherson
通过试验研究，确定了采用移动感应线圈的感应加热工艺矫正角变形的适用性。从实验中获得的结果已被用来创建人工神经网络（ANN）模型的能力，以预测焊接引起的变形和变形矫正实现使用移动感应线圈。实验结果表明，该方法能有效地减小8 mm和10 mm厚DH36钢板的角变形，并能有效地消除6 mm厚钢板的角变形。6 mm板的结果也表明存在一个临界感应线圈的旅行速度，在该速度下发生最大的校正弯曲。人工神经网络已被证明有能力预测焊接和感应加热后的板的最终变形。该模型还被用来作为一种工具，以确定最佳的速度，以尽量减少钢板后，受到焊接和感应加热过程中产生的变形。
An experimental investigation has been carried out to determine the applicability of an induction heating process with a travelling induction coil for the rectification of angular welding distortion. The results obtained from experimentation have been used to create artificial neural network (ANN) models with the ability to predict the weld-induced distortion and the distortion rectification achieved using a travelling induction coil. The experimental results have shown the ability to reduce the angular distortion for 8- and 10-mm thick DH36 steel plates and effectively eliminate the distortion on 6-mm thick plates. Results for 6-mm plates also show the existence of a critical induction coil travel speed, at which maximum corrective bending occurs. ANNs have demonstrated the ability to predict the final distortion of the plate both after welding and induction heating. The models have also been used as a tool to determine the optimum speed to minimise the resulting distortion of a steel plate after being subjected to both welding and induction heating processes.