Evaluation of Surface Roughness of Carbon Nanotube TMT Nanosteel Material Using Taguchi Analysis and Neural Networks

Author(s): 
S. Prabhu, B. K. Vinayagam

Source:

S. Prabhu & B. K. Vinayagam, “Evaluation of Surface Roughness of Carbon Nanotube TMT  Nanosteel Material Using Taguchi Analysis and Neural Networks”, Journal of Mechanical Engineering Research and Developments, vol. 39, no. 3, pp. 718-729, 2016. 
 

ABSTRACT: Neural network analysis is used to predict the surface roughness in CNC lathe machining process of Thermo Mechanically Treated (TMT) steel. L9 orthogonal array was used to optimize the machining parameters using Taguchi design of experiment technique. The data used for the training and checking of the networks performance derived from experiments conducted. Analysis of Variance (ANOVA) and F-test were used to determine the significant parameter influencing the output parameters of surface roughness. The statistical analysis implies that the cutting speed was an utmost parameter on surface roughness. TMT nanosteel are tested using XRD analysis to obtain voids structure of nanosteel. Using the feed forward Artificial Neural Networks (ANNs) the experimental values are trained with the Levenberg-Marquardt algorithm is used the most influential factors were determined. ANN based predicted the surface roughness with a mean squared error is 2.90% for with CNT nanosteel.

Keywords :  Nanosteel; Surface roughness; Taguchi analysis; ANOVA; Neural network.