MODEL MATEMATIK UNTUK MEMPREDIKSI KEKASARAN PERMUKAAN HASIL PROSES CNC BUBUT TANPA PENDINGINAN

Abstract

Turning is one of the machining operations that widely used in variety of manufacturingindustries. The quality of surface plays very important role in the performance of turningthat significantly improves fatigue strength, and corrosion resistance. It is also affectsfunctional attributes of parts such as light reflection, coating and heat treatment. Thisresearch aim to develop a mathematical model to predict surface roughness based onmachining parameter selected (spindle speed, feed, and depth of cut) with cutting coditiondry. The experiment was tested by using Trun Master TMC 320 Sinumerik 802 Smachining center. After 27 specimens were cut for experimental purposes, they weremeasured off-line with a surface roughness tester to obtain the roughness average value.The surface roughness (Ra) could be predicted effectively by applying spindle speed (n),feed rate (Vf) and depth of cut (a) in the multiple regression model as: Y = -0.1228 –0.0001 n + 20.940 Vf + 0.7444 a. The  average percentage deviation of the testing set was6.429 %  for training data set This showed that the statistical model could predict thesurface roughness with about 93.571 % accuracy of the testing data set
Keywors: CNC Turning, Surface Roughness, Multiple Linier Regression