STM Journals Peer-reviewed research from Consortium e-Learning Network

Journal of Mechatronics and Automation

Artificial Neural Network Based Defect Prediction in Casting

  1. Anoop Sankar
  2. P Karunakaran

Abstract

The main problems which are facing by most of the casting industries is loss of productivity which is due to casting defects occurred during the production time. The main casting defects are cracks, misruns, blowholes scabs and airlocks. Most of the investigations made in this area is only discussing the defects occurred after a cast is made and no method has yet been developed to prevent the defects before casting. In this work the data which was collected from an industry is used as input to train the neural network. After training, if a new set of data is entered the network will automatically predict the defects, and then by altering the molding parameters the defects can be prevented.

Keywords

Support