Experimental Investigation on Reverse Engineering Techniques Employing Response Surface Methodology for Freeform Surfaces

Abstract

Reverse engineering techniques are being applied in different streams of engineering especially in manufacturing, repairing activities and for reconstructing freeform surfaces. The method originates from object digitization, progresses to CAD model creation by using Non Uniform Rational B-Splines and later the generation of freeform surface; by employing the curve-D algorithm the deviations between CAD model and reverse engineered freeform surface are estimated. The present work investigates the impact of various process parameters in reverse engineering viz memory, process time, deviation, defects and curvature deviation affected during the freeform surface construction. The factors governing the responses are noise levels, smoothing level and number of triangles. The factors are chosen at 5 levels by employing response surface methodology technique with central composite design matrix. The optimum process conditions were identified by constructing the respective response surface contours. It was observed that the deviation has increased with the smoothing level for various sample sizes undertaken in the investigation, for a constant ‘noise reduction’ parameter value. It is also observed that the curvature deviation has reduced with the reduction of smoothing levels and increase of triangle count.