Application of Neural Network Technology in Welding Procedure Qualification for Joint

DOI

10.2991/icaita-18.2018.14How to use a DOI?

Keywords

neural network; yield stress; welded joint

Abstract

Welding Procedure for joint has a very important position in the manufacture of all kinds of important products. Application of neural network technology could improve the existing in welding procedure qualification for joint process. It is shown that the yield strength of welded joint could be predicted effectively by using the welding parameters of welded joint, the temperature of the welding parameters and the heat treatment after welding as the input parameters by the neural network training. But the P, B and S element could be ignored in the process of predicting yield strength due to little influence on yield stress. The yield strength of steel could be increased by the increase of C content, which could be decreased by the increase of the heat treatment temperature. Neural network model predicts the yield strength of materials with high accuracy, which could be utilized to the simulation of welding procedure qualification for joint.

Copyright

© 2018, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Ji’ai Xue
AU  - Weiping Ouyang
AU  - Yannan Du
AU  - Huiqing Ouyang
PY  - 2018/03
DA  - 2018/03
TI  - Application of Neural Network Technology in Welding Procedure Qualification for Joint
BT  - Proceedings of the 2018 2nd International Conference on Artificial Intelligence: Technologies and Applications (ICAITA 2018)
PB  - Atlantis Press
SP  - 54
EP  - 57
SN  - 1951-6851
UR  - https://doi.org/10.2991/icaita-18.2018.14
DO  - 10.2991/icaita-18.2018.14
ID  - Xue2018/03
ER  -