Volume 48, No 2, 2026, Pages 339-351
Download full text in PDFSemi-Empirical Modeling of Surface Roughness in Turning
Authors:
Miroslav Radovanovic 
DOI: 10.24874/ti.2070.11.25.03
Received: 30 March 2026
Revised: 6 May 2026
Accepted: 19 May 2026
Published: 15 June 2026
Abstract:
The study presents a novel semi-empirical dimensionless mathematical model for predicting surface roughness (Ra) in the finish turning of AISI 316LVM stainless steel. The proposed model integrates dimensional analysis (DA), design of experiments (DoE), and regression analysis (RA) to obtain a realistic and industrially applicable model. Dimensional analysis was applied to derive a dimensionally homogeneous mathematical model. Buckingham's p theorem was used to express the relationship between surface roughness and the influencing factors in dimensionless form. As a result, the normalized surface roughness Ra/(f²/rε) is expressed as a function of the normalized cutting parameters: ap/ap,rec, f/frec, and vc/vc,rec,rec, which are normalized with respect to the tool manufacturer’s recommendations. Taguchi’s design of experiments and analysis of variance (ANOVA) were used to investigate the effects of four process factors (corner radius, depth of cut, feed, and cutting speed) on surface roughness. Regression analysis was then used to determine the unknown coefficient and exponents in the mathematical model. The proposed approach is physically consistent and industrially relevant, providing a practical framework for developing predictive surface roughness models in turning operations.
Keywords:
Turning Surface roughness Semi-empirical model Dimensional analysis Buckingham’s p theorem Design of experiments
