EXPERIMENTAL AND STATISTICAL ASSESSMENT OF POCKET MILLING PARAMETERS FOR SS304 USING TAGUCHI EXPERIMENTAL DESIGN

Abstract

Purpose: Pocket milling of AISI 304 stainless steel requires appropriate selection of cutting conditions to obtain acceptable surface quality without unnecessarily reducing material removal. This study examines the effects of spindle speed, feed rate, and depth of cut on surface roughness and identifies the best observed parameter combination within the tested range.

Design/methodology/approach: Nine pocket-milling trials were conducted on a VMC 850 using a 12 mm TiAlN-coated carbide end mill. A Taguchi L9 orthogonal array was employed with spindle speeds of 1000, 1250, and 1500 rpm, feed rates of 300, 600, and 900 mm/min, and depths of cut of 0.2, 0.4, and 0.6 mm. Surface roughness was measured using a Mitutoyo SJ-301 in terms of arithmetic average roughness, root mean square roughness, and maximum height. Material removal rate was calculated from the machining conditions. Minitab was used for signal-to-noise analysis, main-effects analysis, analysis of variance, first-order regression, and response-surface visualization.

Findings: Feed rate showed the clearest practical effect on surface roughness, although none of the three factors was statistically significant at the 5% level. The lowest measured roughness occurred at 1250 rpm, 300 mm/min, and 0.4 mm depth of cut, giving 1.992 µm arithmetic average roughness, 2.456 µm root mean square roughness, and 11.101 µm maximum height. The highest calculated material removal rate was 6480 mm³/min at 900 mm/min feed and 0.6 mm depth of cut. The first-order regression model gave an R² of 14.3%, indicating limited predictive capability. No independent confirmation experiment was conducted.

Practical implications: The identified setting provides a practical starting point for achieving lower surface roughness under the investigated machine, tool, material, and pocket geometry. The results also show the trade-off between surface quality and material removal rate and provide a basis for replicated and broader multi-response machining studies.