Prediction of Plasma Arc Cutting Performance for SS-304 Material using Artificial Neural Network
Journal: International Journal of Mechanical and Production Engineering Research and Development (IJMPERD ) (Vol.9, No. 5)Publication Date: 2019-10-31
Authors : A. H. Patel; Dr A. B. Pandey;
Page : 551-558
Keywords : Plasma Arc Cutting Process; Artificial Neural Networks & Modeling;
Abstract
A number of cutting parameters are responsible for the quality of cut in plasma arc cutting (PAC) process, so the prediction of process performance is important to set the control parameters for achieving the adequate cut quality. This paper attempts to develop predictive models for Current input parameters, stand-off distance, pressure and cutting speed and their effects on output responses like material remove rate (MRR), top kerf width, bottom kerf width, straightness, and bevel angle during PAC. All the experiments were carried out on 6 mm thick SS-304 material, different output responses were measured and various artificial neural network (ANN) architecture models were developed in Easy NN software for prediction and were determined by calculating various errors and variances between actual experiments. The limiting value for all the errors over the entire data is selected as 5% and the maximum number of training cycles are limited to 1000000 for each learning set. In the present work, it is found that 4-6-5 ANN architecture is the best model structure for selected input parameters.
Other Latest Articles
- Promethee Based Distributed Coverage and Connectivity Preserving Scheduling Algorithm for MWSN
- Experimental Investigation of Spring Back and Wrinkling Phenomena in Square Pipes during Bending
- Mechanical Characterization of Aluminum 6061 with B4c and High Entropy Alloys
- Microstructure and Mechanical Properties of Hot Extruded Mg-3zn-1cu-0.7mn Alloy Produced by Powder Metallurgy
- Experimental Investigation of Twisted Tape Insert in a Double Pipe Heat Exchanger using Aluminium Oxide Nanofluid
Last modified: 2019-11-13 17:47:38