Statistical modelling of deposition rate in twin-wire submerged arc welding

The present article offers statistical modeling of deposition rate in twin-wire submerged arc welding. 'Best subset selection method' and 'Mallows criterion' have been used for model development. The deposition rate has been experimentally measured under varying process parameters including welding current, welding voltage, welding speed, contact-tube-to-workpiece-distance, and wire-diameter. Models for two polarities - direct current electrode positive and direct current electrode negative - have been presented. The models have been validated and have been further confirmed with the data points other than that used for model development. The models have been checked for their predictability and also compared with the models developed with conventionally used technique. The developed models of deposition rate in twin-wire submerged arc welding can be used for practical purposes with admissible error in prediction and better predictability compared with conventional models. The effects of process parameters on the deposition rate have also been presented.