Optimization of Machining Parameters for Improved Surface Integrity in Chromium–Nickel Alloy Steel Turning Using TOPSIS and GRA

Abstract

Interest in surface integrity has grown in the manufacturing industry; indeed, it has become an integral part of the industry. It can be studied by examining surface roughness parameters, hardness variations, and microstructure. However, evaluating all these parameters together can be a challenging task. To address this multi-criteria decision-making model (MCDM), techniques such as Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) and Grey Relational Analysis (GRA) provide a suitable solution for optimizing the machining parameters that lead to improved product quality. This work investigated surface roughness parameters, including arithmetic average surface roughness (2D) (Ra), mean surface roughness depth (2D) (Rz), area arithmetic mean height (3D) (Sa), and maximum surface height (3D) (Sz), in conjunction with Vickers macrohardness (HV) and optical micrographs, to analyze machined surfaces during the turning of X5CrNi18-10 steel. The results suggest that machining with a spindle speed (N) of 2000 rpm or v_c of 282.7 m/min, a feed rate (f) of 0.1 mm/rev, and a depth of cut of 0.5 mm yields the best surface, achieving an “A” class surface finish. These parameters can be applied in manufacturing industries that utilize chromium–nickel alloys. Additionally, the method used can be applied to rank the quality of the product.

1. Introduction

Surface integrity, a widely used term in the manufacturing industry to refer to the functional performance and reliability of machined products in industries, was coined by Michael Field and F. Kahles in 1964 [1]. It is quantified based on mechanical, metallurgical, chemical, and topological states [2]. These qualities are studied through surface roughness, hardness variation, microstructure, and corrosion resistance assessments.

Studying surface integrity is essential for determining the quality, performance, and longevity of machined product, particularly in high-reliability industries such as the automotive, marine, and aerospace industries. Enhanced surfaces and functionalities are essential, as they directly impact product performance [2,3,4,5]. Additionally, surface integrity is essential for ensuring safe and efficient operations under complex service conditions [6].

Maintaining surface integrity can be challenging when machining hardened steels, such as X5CrNi18-10, due to their properties. X5CrNi18-10, an austenitic chrome nickel steel, contains up to eighteen percent chromium and ten percent nickel, giving the material properties like toughness and corrosion resistance, which are valuable traits for the engineering industry. The material is widely used in the automotive, food, and marine industries but it is challenging to machine due to its complex properties [7,8]. This steel presents several challenges, such as work hardening and low thermal conductivity [9,10]. The combination of work hardening and low thermal conductivity results in high cutting forces, leading to increased energy consumption during machining [11]. The workpiece material exhibits high toughness and shows resistance during machining, making it essential to study the effects of the cutting conditions and to evaluate surface integrity. Chi et al. [12] mentioned that surface defects may initiate crack propagation during operations, resulting in premature component failure, significantly reducing the material’s performance and limiting this steel’s large-scale engineering application. To avoid such risks, achieving a high-quality finish becomes important. The class “A” surface finish is a high-quality surface finish with a lack of imperfections like scratches or tearing [13].

Labuda et al. [14] studied the surface roughness parameters arithmetic average (AA) surface roughness (R_a), total peak-to-valley height (R_t), core roughness depth (R_k), max peak–valley (3D) (R_pv), and reduced peak height (R_pk) to analyze the surface of X5CrNi18-10 steel. The most favorable results, with an R_a of 0.77, were obtained at a cutting speed of 200 mm/min. Das et al. [15] investigated the arithmetic average (AA) surface roughness (R_a), root mean square roughness (R_q), and mean surface roughness depth (R_z) of AISI 4140 steel during hard turning. It was found that the best surface finish, with an R_a of 0.714, R_q of 0.844, and R_z of 3.741 μm, was obtained at a feed (f) of 0.05 mm/rev, cutting speed (v_c) of 170 m/min, and 0.2 mm cut depth. Another study [16] presented the results of finish turning on X5CrNi18-10 steel, where R_a, R_t, R_k, R_pv, and R_pk were measured after the machining operation and burnishing was used to improve the surface integrity. These studies briefly investigated surface roughness parameters such as R_a, R_t, R_k, R_pv, and R_pk; however, another factor that affects surface integrity is hardness variation. The factors that, in turn, influence hardness variation are cutting speed, feed rate, depth of cut, and tool material. Higher cutting speeds generate heat, which leads to thermal softening of the material, potentially reducing its hardness; however, increased feed rates cause greater heat generation and strain, which counteract this effect by inducing hardening [17,18]. The depth of cut has a significant impact on the cutting forces and resulting surface hardness. A larger depth of cut increases the cutting forces and the heat generated, which influence the hardness of the machined surface [19].

Individual studies focusing solely on surface roughness parameters do not provide comprehensive insights into surface integrity, as the evaluation of surface integrity in difficult-to-machine materials also involves the assessment of surface defects such as surface tearing, clean feed marks, and topographical and microstructural changes. It can be challenging and time-consuming to evaluate all these responses individually and select the optimal combination of machining parameters.

To address this, multi-criteria decision-making (MCDM) techniques, such as Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) and Grey Relational Analysis (GRA), are used to optimize machining parameters. Tzeng et al. [20] used Grey Relational Analysis (GRA) to analyze roughness average, roughness maximum height, and roundness. Kumar et al. [21] used TOPSIS to analyze the material removal rate. Abhang et al. [22] employed Grey Relational Analysis (GRA) to investigate the AA surface roughness (R_a) and chip thickness. Similarly, researchers [23,24] have applied these techniques to analyze various machining variables and their effects to draw conclusions.

This work aimed to evaluate surface integrity through the analysis of surface roughness parameters (R_a, R_z, S_a, and S_z) and Vickers hardness variation, as well as through optical micrographs of the machined surface. To identify the influence of the machining parameters (spindle speed and feed rate) and machining environment (dry and wet) on surface integrity, TOPSIS and GRA methods were used to optimize the machining parameters that produce better surface quality. This work can be used in manufacturing industries to achieve an “A” class surface roughness in X5CrNi18-10 steel.

2. Materials and Methods

The test material for this study was X5CrNi18-10 (an ISO M-type material grade) steel. The material influenced the cutting-edge, heat generation, notch wear, and built-up edge on the tool insert. The selection of inserts was a crucial aspect for this material’s properties. Based on the aforementioned workpiece material properties, the selected tool insert was the DNMG150604-MF1 CP500 (Seco Tools, Björnbacksvägen, Sweden), which is suitable for machining ISO M-type materials, as per the manufacturer’s guide. The insert was made of carbide with a PVD coating, which improved wear resistance and cutting performance. The insert was used to ensure close tolerance, good finish, and increased tool life due to the coating. These inserts were used for finishing and semi-finishing operations for longitudinal turning profiling. The nose radius of the insert was 0.40 mm. The CP500 grade was a hard micro-grain grade primarily intended for finishing superalloys. The MF1 chip breaker was designed for medium-finishing operations, and machining using this insert is well-suited for achieving a good surface finish. The tool holder used in the experiment was the DDJNL2525M15 from Seco tools (Seco Tools, Björnbacksvägen, Sweden) [25]. The tool shank was 25 mm in height (H) and breadth (B), providing stability while turning. It is essential to analyze the machining parameters that must be chosen for machining. The selection of cutting parameters requires a study of the workpiece characteristics and the properties of the cutting tool. Based on the workpiece material, the DNMG150604 tool was selected.

The manufacturer’s guide for the tool insert indicated suitable cutting parameters, with a spindle speed up to (N_s) 2000 rpm, a feed (f) of 0.1 mm/rev, and a depth of cut up to 2 mm. In previous work, these parameters were examined and it was concluded that the varying depth of cut plays a minor role in surface generation. Additionally, feed rates of 0.08, 0.16, and 0.24 have already been investigated and found to influence the surface integrity. In contrast, cutting speed had a minor effect [26].

To understand the effect of machining parameters on surface integrity, these parameters were used as a reference and varied. The spindle speed (N_s) was varied at three levels—1000 rpm (v_c = 141.4 m/min), 1500 rpm (v_c = 212.1 m/min), and 2000 rpm (v_c = 282.7 m/min). The selection of spindle speed was favored because it was directly controllable on the machine and easier to adjust. Feed (f) varied at two levels, 0.1 and 0.3 mm/rev, and depth of cut (a_p) varied at one level. Two environments were chosen: dry and wet machining conditions.

Table 1. Level of machining parameters.

Cutting Parameters	Level 1	Level 2	Level 3
Spindle Speed—N (rpm)	1000	1500	2000
Feed Rate—f (mm/rev)	0.1	0.3	-
Environment	Dry	Wet	-

explains the levels of machining parameters.

Based on three levels of spindle speed, two levels of feed, and two levels of machining environment, Taguchi’s L12 orthogonal array was created.

Table 2. L₁₂ Taguchi’s design of experiment.

Surface Number	Spindle Speed N (rpm)	Cutting Speed vc (m/min)	Feed Rate f (mm/rev)	Machining Environment
S1	1000	141.4	0.1	Dry
S2	1500	212.1	0.3	Dry
S3	2000	282.7	0.1	Dry
S4	1000	141.4	0.3	Dry
S5	1500	212.1	0.1	Dry
S6	2000	282.7	0.3	Dry
S7	1000	141.4	0.1	Wet
S8	1500	212.1	0.3	Wet
S9	2000	282.7	0.1	Wet
S10	1000	141.4	0.3	Wet
S11	1500	212.1	0.1	Wet
S12	2000	282.7	0.3	Wet

presents the design of experiments for this study.

Taguchi’s orthogonal array approach for machining parameter optimization provides effective results [27,28,29]. The design of the experiment table provides sufficient variation in the parameters. Table 2 represents S1 as surface no. 1, spindle speed in rpm, feed rate in mm/rev, depth of cut in mm, and the machining environment. To implement the experiment’s design, proper workpiece preparation was necessary.

According to the experimental design, a workpiece was prepared to collect data and understand the variation in surface integrity. Figure 1 highlights the dimensions of the workpiece. It was divided into five equal surfaces, each 30 mm in length, with a 4 mm gap between the surfaces and a 5 mm depth, creating a groove that parted the surfaces. These surfaces were divided to experiment with the different machining parameters, where feed, depth of cut, and spindle speed were varied according to the design of the experiments. Initially, the specimen was a 45 mm cylinder with a surface length of 30 mm. The total number of specimens used in this experiment was four, giving twelve surfaces.

A finished outer diameter (OD) turning operation was carried out to obtain experimental data. The HAAS ST-20 Y series machine tool (Haas Automation, Inc, Leoben, Austria) was used to conduct all the experiments, as shown in Figure 2. This machine is a Y-axis CNC turning Center with a maximum spindle speed of 4000 rpm [30]. For wet conditions, 5% CIKS HKF 420 oil emulsion (CIKS Kft. Miskolc, Hungary) was used.

In this work, the surface profile parameters R_a, R_z, S_a, and S_z of the machined workpiece were measured using an Alti Surf 520. The surface profile parameters were measured for all 12 machined surfaces, i.e., for 12 experiments, with a sampling length of 10 mm at a scale of 100 μm using a CL2 confocal chromatic sensor. The examined surface roughness parameters were R_a, i.e., the average surface roughness, and R_z, i.e., the average roughness depth. The surface 2D profile was determined according to ISO 21920-2 [31], and a filter (λc) was applied as specified in ISO 21920-3 [32]. Surface roughness was measured using the Arithmetic roughness average (R_a) or Arithmetic Average roughness (R_a). This metric is often referred to as the arithmetic mean roughness value, arithmetic average (AA), or centerline average (CLA). The arithmetical average roughness is defined as the area between the roughness profile and its centerline or the integral of the absolute value of the roughness profile height over the evaluation length [33]. R_z is the average absolute value of the five highest peaks, and the five lowest valleys were defined on the sampling length. This parameter was used to frequently check whether the profile had protruding peaks that might have affected the static or sliding contact function. Surface profile measurements included Area Average Roughness (S_a) and Maximum height (S_z). Area Average Roughness was the arithmetic mean of the absolute height deviations from the mean plane over a defined area. It provided an overall measure of surface texture and served as a 3D analog to the commonly used R_a parameter in 2D measurements. Maximum height was the vertical distance between the highest peak and the lowest valley within the measured area, representing the extreme height variation on the surface [34]. The setup parameters for the measurements were selected in accordance with ISO 21920-2 [31]. A 4 mm × 4 mm area was selected for measurement in the X and Y axes. These parameters are key 3D surface roughness parameters used to quantitatively describe the texture of turned surfaces. Studying S_a and S_z provides a comprehensive understanding of surface quality, helps optimize machining processes, and supports the prediction and control of functional properties in manufactured parts. Their measurements support the development of empirical and predictive models for surface roughness, helping in making practical measurements of surface finishes based on parameter variation [35]. A light optical microscopy examination was performed to validate surface profile measurements.

Microscopic analysis using optical micrographs is essential for the detailed examination of turned surfaces as it provides insights into surface topography and ensures precision in manufacturing processes. Chen et al. [36] noted that microscopic surface profiles are crucial for ensuring strict geometric compliance in precision manufacturing. A light optical inverted metallurgical microscope in brightfield mode with Axio Observer 5, (Carl Zeiss Microscopy GmbH, Germany), with magnifications of 100×, was used to understand the topography of the turned surface. The optical microscopy setup was connected to a computer, and the scaling was adjusted in the settings to enhance image precision. The other parameter used to analyze surface integrity was hardness measurement.

The Vickers hardness test was performed on the specimen using a Tukon 2100B (Berg Engineering & Sales Company, Inc., Rolling Meadows, IL, USA) to analyze the surface hardness, in accordance with ISO 6507 [37], on the shaft surface with a 10-s dwell time and an HV 10 load. The test was performed at three different points on each surface. Vickers hardness analysis was essential for examining and understanding the mechanical properties by analyzing the variation on the turned surfaces due to different cutting parameters. Vickers microhardness is correlated with contact hardness and relative contact pressure, which depend on surface roughness parameters [38]. One study [39] used coolant during the turning operation and found that the cooling effect had several advantages on the turned surface. It was noted that the minimum quantity lubricant (MQCL) reduced the depth of the hardened layer compared to dry cutting, indicating less structural change. Another study [40] mentioned that wet cutting decreased the hardened depth of the surface integrity.

The value was then compared to the surface integrity results, surface roughness parameters, and macro-Vickers hardness. MCDM models TOPSIS and GRA were used to rank the surfaces. Also, comparisons were performed.

The methodology of this study is presented through a flow chart, as shown in Figure 3.

3. Results

The primary results present the 2D and 3D profile analyses, followed by Vickers hardness. Additionally, microscopic analyses of surfaces were used to gain a deeper understanding of surface quality.

3.1. 2D & 3D Profile Analysis

Arithmetic average (AA), surface roughness (R_a), and mean surface roughness depth (R_z) were calculated for each surface.

Table 3. The measured values of R_a, R_z, S_a, and S_z.

Surface	Sa	Sz	Ra			Rz			Ra	Rz
Number	Area 4 mm × 4 mm	Area 4 mm × 4 mm	Point 1	Point 2	Point 3	Point 1	Point 2	Point 3	Average	Average
	µm	µm	µm	µm	µm	µm	µm	µm	µm	µm
S1	0.868	6.039	0.876	0.868	0.877	4.701	4.462	4.72	0.87	4.63
S2	3.82	17.814	3.902	3.902	3.891	16.589	16.341	16.632	3.9	16.52
S3	0.8	5.704	0.825	0.851	0.82	4.839	4.897	4.849	0.83	4.86
S4	4.187	20.496	4.297	4.283	4.266	16.601	17.189	16.84	4.28	16.88
S5	0.854	6.181	0.849	0.853	0.868	4.89	4.83	4.918	0.86	4.88
S6	4.361	17.55	4.413	4.4	4.367	15.944	16.014	16.066	4.39	16.01
S7	0.698	4.654	0.706	0.699	0.693	3.717	3.951	3.819	0.7	3.83
S8	4.038	18.071	3.929	3.964	3.999	16.553	16.558	16.694	3.96	16.6
S9	0.585	4.296	0.589	0.598	0.61	3.311	3.425	3.507	0.60	3.41
S10	4.094	18.607	4.154	4.168	4.213	17	16.916	16.959	4.18	16.96
S11	0.642	4.788	0.666	0.662	0.653	3.831	3.934	3.798	0.66	3.85
S12	4.311	20.264	4.203	4.254	4.216	17.987	17.817	18.027	4.22	17.94

provides surface roughness measurements for the 12 surfaces used in the experiment. R_a and R_z were estimated at three different points on the surface to give reliability and accuracy in the measurements. The surfaces could be classified into two groups: those with low surface roughness and those with high surface roughness. The average R_a was less than 1 µm; for high surface roughness, the average R_a was greater than 3 µm.

For 2D profile measurements, Arithmetic average surface roughness R_a and Mean surface roughness depth R_z were used; for 3D profile measurements, Area Average surface (S_a) and Roughness and Maximum Height (S_z) were used. S1–S6 were machined under dry conditions, and S7–S12 were machined under wet conditions. It can be observed that the feed rate had a significant impact on the surface roughness. Surface machined with a higher feed rate, i.e., 0.3 mm/rev, showed higher values of R_a, R_z, S_a, and S_z. It was also noticeable that the difference between R_a and S_a was minimal; this was particularly evident in S2 and S8. The lower feed rate of 0.1 mm/rev provided better R_a, R_z, S_a, and S_z. S9, S7, and S11 had R_a ≤ 0.8, reaching a precision class accuracy. These surfaces were machined under wet machining conditions.

It was also noticeable that changing the spindle speed from 1000 to 2000 rpm improved the surface finish. However, this effect was relatively minor, unlike the feed rate, which was identified as the most dominating factor impacting surface roughness. A higher feed rate produced rougher surfaces. Spindle speed variation had a minor influence on surface roughness. Wet machining conditions helped reduce roughness, especially at lower feed rates.

A 3D topography analysis was performed to better understand the profile results; 3D topography helps assess a surface in detail, highlighting the peaks and valleys in the surface profile.

3.2. 3D Topography Analysis

Figure 4 illustrates the 3D topography, describing the surface height and waviness under various cutting parameters, spindle speeds, and feed rates. It also highlights the machining conditions, i.e., dry and wet. A 4 mm × 4 mm area was measured in the X and Y axes, as shown in the plot in Figure 4. The Z axis is in micrometers (µm), representing depths at different points. The color gradients in Figure 4 represent the depths and the peaks: green and blue areas indicate lower depths or smoother regions, while red areas indicate peaks or rough regions. These variations show the topographical features of the turned surface. The depth of cut was 0.5 mm for all experiments.

S1, with cutting parameters N = 1000 rpm and f = 0.1 mm/rev exhibited moderate surface waviness, characterized by some isolated peaks, indicating a relatively smoother finish with non-uniform peaks. S2, with N = 1500 rpm and f = 0.3 mm/rev represented a rough surface, with high peaks and depth. This was understandable because of the high feed rate combined with dry machining. S3, with N = 2000 rpm and f = 0.1 mm/rev, i.e., a combination of a high spindle speed and low feed, showed moderate rough surfaces with medium peaks. In this cases, the peaks were more organized than with S2. This suggested a better surface feed despite dry machining in stainless steel. S4, with N = 1000 rpm and f = 0.3 mm/rev, showed random surface spikes. This indicated a poor surface finish due to the combination of a low spindle speed and high feed rate. S5, with N = 1500 rpm and f = 0.1 mm/rev, exhibited a uniform surface finish with a consistent pattern. S6, with N = 2000 rpm and f = 0.3 mm/rev, showed the roughest surface with dry machining, presenting high peaks and uneven patterns. High speed and a high feed rate in dry machining cause instability. S7–S12 comprise machining surfaces with wet machining conditions using coolant. S7, with cutting parameters N = 1000 rpm and f = 0.1 mm/rev, presented smooth surfaces with low peaks. The good surface finish indicated that the coolant effectively reduced friction and tool wear. S8, with N = 1500 rpm and f = 0.3 mm/rev, showed consistent peaks. In this case, feed rate effects were visible; however, the coolant minimized the impact of roughness. S9, with N = 2000 rpm and f = 0.1 mm/rev, i.e., a combination of high and low feed rates, presented the smoothest surface overall, with minimum peak variation. Overall, a high speed with low feed and wet machining conditions provided excellent surface finishes. S10, with N = 1000 rpm and f = 0.3 mm/rev, showed a larger spacing between the peaks. This was understandable due to the slower spindle speed and higher feed rate. S11, with N = 1500 rpm and f = 0.1 mm/rev, showed a smooth surface with slight regular waviness. It could be observed that the combination of suitable parameters resulted in a uniform surface texture. S12, with N = 2000 rpm and f = 0.3 mm/rev, showed more defined peaks and depths than S9 and S11. This indicated that the high feed rate limited the smoothness, even with a coolant.

3.3. Optical Microscopy Analysis

An optical micrograph of the surfaces is presented in Figure 5.

Figure 5 illustrates machined surfaces under dry and wet cutting conditions at different spindle speeds and feed rates. Surfaces are titled S1–S12. Red arrows in the images highlight surface tearing, surface deformation, and deformation due to heat generation during machining. The green double-sided arrow in Figure 5 represents the feed marks. The yellow arrows highlight the clean feed marks, which showed no deformation in the wet machining environment. The vertical axis represents spindle speed, and the horizontal axis represents the feed rates.

S1, with cutting parameters spindle speed (N) = 1000 rpm and feed rate (f) = 0.1 mm/rev showed heavy surface tearing with scraps and stains. The R_a value measured was 0.87 µm, and the R_z was 4.63 µm. Despite the modest surface roughness, the microscopy image revealed that the surface was damaged. The variance in R_a values and the optical micrograph showed that surface tearing was not always reflected in the average R_a values. S2, with N = 1500 rpm and f = 0.3 mm/rev, showed tool marks and scratches in the box. It could be observed that high roughness values corresponded to poor surface finishes. The high feed rate and dry conditions of hardened steel machining can therefore result in a poor surface finish. S3, with N = 2000 rpm and f = 0.1 mm/rev, showed many micro defects and pitting, marked with red circles. The defects were visible, but R_a remained relatively low and showed uniform surfaces. S4, with N = 1000 rpm and f = 0.3 mm/rev, showed material tearing and deformation, as highlighted in red boxes. It also showed that a higher feed rate provided a poor surface finish. S5, with N = 1500 rpm and f = 0.1 mm/rev, presents small pits and surface defects; however, the surface looked better, with minimal wear. S6, with N = 2000 rpm and f = 0.3 mm/rev, exhibited a rough texture (multiple red boxes indicate deformation and pits). Surface defects and machining marks support high R_a and R_z values. A high feed rate with a high spindle speed in dry conditions caused deformation due to high heat generation during machining.

S7–S12 represent the surfaces created under wet machining conditions. Optical microscopy images showed that S7, with N = 1000 rpm and f = 0.1 mm/rev, was smooth, with consistent grain marks, minimal scratches, and a clear feed mark without deformation, highlighted with a green arrow. The lubricant used helped minimize the wear, resulting in a good finish. S8, with N = 1500 rpm and f = 0.3 mm/rev, showed clean, linear cuts with no significant damage. Despite the clean micrograph, R_a was high due to deep but consistent grooves from a high feed rate. S9, with N = 2000 rpm and f = 0.1 mm/rev, showed minimal damage, clean peaks, and minor pitting. The visualization results and measured results showed similarities. A low feed rate and high spindle speed were the best combination with wet conditions. S10, with N = 1000 rpm and f = 0.3 mm/rev, presented considerable tool marks. A high feed rate with a low feed rate impacted the surface finish, even in wet conditions. S11, with N = 1500 rpm and f = 0.1 mm/rev, showed clean lines, minimal pitting, and surface tearing. S12, with N = 2000 rpm and f = 0.3 mm/rev, i.e., a high spindle speed and feed rate, presented consistent feed marks and some micro defects; however, although it appeared to be of better quality, it had high R_a and R_z values, primarily due to the higher feed rate.

3.4. Vickers Hardness Analysis

The Vickers hardness was measured for each surface under varying machining conditions, including both dry and wet conditions. The hardness was measured at three points on each surface, and the values of each point are presented in

Table 4. Measured Vickers hardness values.

Surface Number	Point 1 HV 10	Point 2 HV10	Point 3 HV 10	Average
S1	256	249	249	251.3
S2	308	317	311	312.0
S3	259	258	260	259.0
S4	314	280	289	294.3
S5	245	249	255	249.6
S6	297	310	320	309.0
S7	246	249	250	248.3
S8	322	310	325	319.0
S9	276	263	266	268.3
S10	333	326	323	327.3
S11	252	261	263	258.7
S12	305	327	303	311.7

. The average of all three points was taken to analyze the results. The peak hardness, i.e., HV 327.3, was measured with S10. The lowest hardness, i.e., HV 248.3, was measured at S7.

In dry machining conditions, the peak hardness was measured on S2, with HV 312.0, and the lowest hardness was measured on S5, with HV 249.6. The peak hardness in each condition was measured at a 0.3 mm/rev feed rate, and the lowest hardness was measured at a 0.1 mm/rev feed rate. This observation showed the effect of feed rate on hardness variation.

To easily visualize the effect of machining parameters and environment, Figure 6 highlights the results of the hardness variation tests. With dry machining (S1–S6), the average hardness measured was 279 HV; in wet machining conditions, the average hardness measured was 288.8 HV. The results show that coolant usage increased surface hardening due to improved thermal control and a reduced thermal softening effect during machining. The results show that the hardness increased at a feed rate of 0.3 mm/rev in both machining conditions. It is noticeable from Figure 6 that at higher feed rates and lower speeds, the achieved hardness was more significant than at medium and high speeds, as seen in S10. It can be seen that as the spindle speed increased with a higher feed, the hardness was reduced. At higher spindle speeds and feed rates, the average hardness was lower than the others, as seen in S12. With an increased feed rate and spindle speed, more heat was generated, resulting in a softer surface.

A trend was observed for a low feed rate of 0.1 mm/rev. The hardness value increased with an increase in the spindle speed. S9, with a low feed rate and high spindle speed, showed improved hardness. Similar results were observed with S3 (wet condition); as the tool feds slowly, the surface became more hardened with increased spindle speed due to rapid strain hardening, and there was less time for thermal cooling.

The factors that affected surface integrity, hardness variation, microstructure, and surface roughness parameters were mainly influenced by cutting parameters.

3.5. Main Effect Plots

The effect of cutting parameters, i.e., feed rate and spindle speed on the surface profile, arithmetic average surface roughness (R_a), mean surface roughness depth (R_z), area average surface roughness (S_a), and maximum height (S_z) was analyzed using main effect plots. A main effect plot is a graphical tool used to visualize how individual cutting parameters affect surface integrity measures, such as surface roughness and hardness. Main effect plots help identify which parameters have the most significant impact on surface integrity, guiding optimization for improved machining outcomes [41]. Figure 7 represents the main effect plot between cutting parameters, i.e., the influence of N and f on surface profile parameters R_a, R_z, S_a, S_z. The main effect plot suggests that N = 1000 rpm and N = 2000 rpm had less influence on the AA surface roughness (R_a); however, N = 1500 rpm decreased the value of R_a. Similar results could be seen between N and S_a. Increasing spindle speed from 1000 to 1500 and 2000 showed minimal variation and a lesser influence of R_z. However, for S_z, the variation changed. By increasing N from 1000 to 1500 rpm, the S_z value decreased and then increased slightly at 2000 rpm. In all parameters, the feed rate showed similar results, primarily because it varied at only two points; however, as the feed rate increased, the R_a, R_z, S_a, and S_z values also increased, indicating a rougher surface. This was also evident in the results mentioned above. The medium spindle rate showed better results for R_a, R_z, S_a, and S_z. Also, lower feed rates provided better results for R_a, R_z, S_a, and S_z.

The main effect plot in Figure 8 presents the influence of spindle speed and feed rate on Vickers hardness. Hardness exhibited an increasing trend with variations in spindle speed. The increase in N from 1000 to 1500 showed a slight increment in the hardness value, and further increasing the N to 2000 increased the value of hardness. However, the feed rate showed a strong influence on surface hardness. Increasing the feed rate from 0.1 mm/rev to 0.3 mm/rev increased the hardness. The higher feed rate resulted in greater plastic deformation of the material. Feed rate plays an essential role in determining the surface hardness after machining. A higher feed rate increases the hardness, improving surface wear resistance. Similarly, higher spindle speed enhances the surface hardness. Therefore, careful control of these parameters is necessary to improve surface integrity.

3.6. Implementation of Multi-Objective Optimization Methods

The objective of this analysis was to optimize surface integrity by minimizing R_a and maximizing HV. An equal priority was assigned to all the response variables during the analysis.

3.6.1. TOPSIS Analysis

The responses for AA surface roughness and Vickers macrohardness were used to perform TOPSIS and GRA analyses. The other surface parameters were not considered, as R_a is the most widely used parameter. Additionally, if the R_a was lower, other values followed accordingly.

In Figure 9, $\mathrm{i}$ is the number of surfaces $(\mathrm{i}=\mathrm{1,2},3\dots ,12)$, $j$ is the number of responses $(j=\mathrm{1,2},3)$, $X_{ij}=$ value normalized for the $i\mathrm{t}\mathrm{h}$ experiment and $j\mathrm{t}\mathrm{h}$ response; $\overline{X_{ij}}$ is a normalized decision matrix and $W_j$ is the weight assigned to the $j\mathrm{t}\mathrm{h}$ response. In the present work, the weights were selected through direct assignment. Each data point was assigned a weight of 0.50 so that the summation equaled 1. $V_{ij}$ is the weighted normalized matrix and $V_j^{+}$ is the ideal best value. In the current study, the ideal best value for roughness is the minimum value, while for hardness, it is the maximum value. $V_j^{-}$ is the ideal worst value. The maximum value is the ideal worst value for roughness, and for hardness, it is the minimum value. Geometrically, in a multi-dimensional space where each dimension represents a criterion, the positive ideal solution $A^{+}$ and the negative ideal solution, $A^{-}$ are points in that space. Each alternative is also a point. The Euclidean distances $S_i$ ⁺ and $S_i$ ⁻ represent the straight-line distances from each alternative to the positive ideal and negative ideal points, respectively. The final decision is made by comparing these distances, with the preferred alternative being the one closest to the positive ideal solution and farthest from the negative ideal solution. The Euclidean distance for the ideal best and worst values was calculated using the formulas depicted in Figure 9.

Table 5. TOPSIS analysis.

Machining Parameters				Response variables
Surface Number	Spindle Speed N (rpm)	Feed Rate f (mm/rev)	Environment	Average Ra µm	Average HV
S1	1000	0.1	Dry	0.87	251.3
S2	1500	0.3	Dry	3.90	312.0
S3	2000	0.1	Dry	0.83	259.0
S4	1000	0.3	Dry	4.28	294.3
S5	1500	0.1	Dry	0.86	249.6
S6	2000	0.3	Dry	4.39	309.0
S7	1000	0.1	Wet	0.70	248.3
S8	1500	0.3	Wet	3.96	319.0
S9	2000	0.1	Wet	0.60	268.3
S10	1000	0.3	Wet	4.18	327.3
S11	1500	0.1	Wet	0.66	258.7
S12	2000	0.3	Wet	4.22	311.7
				Euclidean Distance
Normalized Matrix		Weightage Normalized Matrix		Si+	Si-
0.0843	0.2540	0.0211	0.0635	0.0203	0.0849
0.3763	0.3154	0.0941	0.0788	0.0797	0.0200
0.0803	0.2618	0.0201	0.0655	0.0182	0.0860
0.4133	0.2975	0.1033	0.0744	0.0893	0.0119
0.0827	0.2523	0.0207	0.0631	0.0206	0.0853
0.4241	0.3124	0.1060	0.0781	0.0917	0.0153
0.0675	0.2510	0.0169	0.0628	0.0201	0.0891
0.3826	0.3225	0.0957	0.0806	0.0812	0.0206
0.0578	0.2713	0.0145	0.0678	0.0149	0.0917
0.4033	0.3309	0.1008	0.0827	0.0864	0.0206
0.0637	0.2615	0.0159	0.0654	0.0174	0.0901
0.4078	0.3151	0.1019	0.0788	0.0876	0.0165
Si+ + Si-	Pi	Surface No.	Ranking	Ranking Dry	Ranking Wet
0.1053	0.8069	S1	5	2	-
0.0998	0.2009	S2	8	4	-
0.1041	0.8256	S3	3	1	-
0.1012	0.1178	S4	12	6	-
0.1060	0.8055	S5	6	3	-
0.1070	0.1433	S6	11	5	-
0.1093	0.8159	S7	4	-	3
0.1019	0.2027	S8	7	-	4
0.1066	0.8601	S9	1	-	1
0.1070	0.1928	S10	9	-	5
0.1075	0.8380	S11	2	-	2
0.1041	0.1587	S12	10	-	6

provides a detailed analysis and calculation of the ranking.

According to the TOPSIS analysis results, S9 (wet machining, 2000 rpm, 0.1 mm/rev) ranked first overall and had the highest performance index, i.e., Pᵢ = 0.8601, indicating that it was the best among all the surfaces analyzed. S3 and S7, created under high-speed, low-feed conditions, presented low surface roughness. S2 and S4, which had high feed rates, exhibited lower performance indices, i.e., Pᵢ = 0.2009 and 0.1178, due to increased roughness. S3 ranked highest among the dry machining condition samples. S8 (wet, 1500 rpm, 0.3 mm/rev) was the lowest ranking surface, i.e., P_i = 0.2027, demonstrating that high feed rates degrade multi-performance results, even in wet conditions. This analysis demonstrates that the best outcomes for surface roughness parameters and hardness were obtained with low feed rates and higher spindle speeds in wet conditions. Table 5 presents the overall ranking of dry and wet conditions to facilitate understanding of the decision-making results.

3.6.2. Grey Relational Analysis (GRA)

Grey Relational Analysis (GRA) is an excellent approach for modifying a multi-response variable problem into a single array. It enables a set of input parameters to be used to rewrite outputs. The analysis can be performed in four primary steps, as shown in Figure 10 [43]. The response values are then standardized or normalized to within a range of 0 to 1, with 1 representing the optimal result. The normalized value is denoted by $x_i^{\mathrm{*}}\left(k\right)$, where $x_i^0\left(k\right)$ represents the original value for the $kth$ response. Normalization for R_a values was performed using the formula presented in Figure 10. For R_a, the smaller the better, while for the hardness, the larger the better. The maximum and minimum of $x_i^0\left(k\right)$ were used to perform normalization. The grey relational coefficient, denoted as ${\xi }_i\left(k\right)$, was calculated based on ${\mathsf{\Delta }}_{0}i\left(k\right)$, i.e., the absolute difference between the normalized value and the ideal (reference) normalized value. Here, ${\mathsf{\Delta }}_{\mathrm{m}\mathrm{i}\mathrm{n}}$ and ${\mathsf{\Delta }}_{\mathrm{m}\mathrm{a}\mathrm{x}}$ represent the minimum and maximum values of ${\mathsf{\Delta }}_{0}i\left(k\right)$, respectively. The distinguishing coefficient, denoted by $\mathsf{\Psi }$, was within the range $\mathsf{\Psi }\in \left[\mathrm{0,1}\right]$. The distinguishing coefficient (Ψ) was set to 0.5 to maintain balance. The grey relational grade (GRG) is represented by ${\alpha }_i$, and $n$ denotes the number of responses considered.

Table 6. GRA analysis.

Surface Number	Spindle Speed N (rpm)	Feed Rate f (mm/rev)	Environment (Dry/Wet)	Normalization		Deviation Sequence
S1	1000	0.1	Dry	0.9276	0.0376	0.0724	0.9624
S2	1500	0.3	Dry	0.1305	0.8059	0.8695	0.1941
S3	2000	0.1	Dry	0.9386	0.1350	0.0614	0.8650
S4	1000	0.3	Dry	0.0293	0.5819	0.9707	0.4181
S5	1500	0.1	Dry	0.9321	0.0160	0.0679	0.9840
S6	2000	0.3	Dry	0.0000	0.7679	1.0000	0.2321
S7	1000	0.1	Wet	0.9736	0.0000	0.0264	1.0000
S8	1500	0.3	Wet	0.1132	0.8945	0.8868	0.1055
S9	2000	0.1	Wet	1.0000	0.2532	0.0000	0.7468
S10	1000	0.3	Wet	0.0567	1.0000	0.9433	0.0000
S11	1500	0.1	Wet	0.9838	0.1308	0.0162	0.8692
S12	2000	0.3	Wet	0.0445	0.8017	0.9555	0.1983
GRC Calculation		GRG Calculation	Surface Number	Ranking Overall		Ranking Dry	Ranking Wet
0.8735	0.3419	0.3039	S1	7		3	-
0.3651	0.7204	0.2714	S2	9		4	-
0.8906	0.3663	0.3142	S3	5		1	-
0.3400	0.5446	0.2211	S4	12		6	-
0.8804	0.3369	0.3043	S5	6		2	-
0.3333	0.6830	0.2541	S6	11		5	-
0.9498	0.3333	0.3208	S7	4		-	4
0.3605	0.8258	0.2966	S8	8		-	5
1.0000	0.4010	0.3503	S9	1		-	1
0.3464	1.0000	0.3366	S10	2		-	2
0.9687	0.3652	0.3335	S11	3		-	3
0.3435	0.7160	0.2649	S12	10			6

provides a detailed calculation of GRG and ranking.

Surface S9 (wet machining, 2000 rpm (v_c = 282.7 m/min), 0.1 mm/rev) ranked first in the Grey Relational Analysis (GRA) and showed the best overall performance, i.e., αᵢ = 0.3503, indicating its good combination of high hardness ad low surface roughness. S7, with αᵢ = 0.3208, and S3, with αᵢ = 0.3142, likewise had good machining conditions with low f and high N, ranking a close second and third, respectively. S3 had the highest ranking among the dry-machined surfaces, suggesting that good performance can be attained in dry conditions with the appropriate cutting parameters. On the other hand, due to their high roughness values, S4 and S6, both created at high feed rates, exhibited the lowest grey relational grade αᵢ, with values of 0.2211 and 0.2541 respectively, placing them at the bottom of the ranking. Overall, GRA demonstrated that the optimum multi-response performance regarding surface integrity, mechanical, and dynamic parameters could be achieved with a feed rate of 0.1 mm/rev and a spindle speed of 2000 rpm (v_c = 282.7 m/min), particularly in wet environments.

4. Discussion

Among all parameters, S9, with machining parameters of N = 2000 rpm (v_c = 282.7 m/min) and f = 0.1 mm/rev, with wet a machining environment, achieved the best surface quality, with R_a 0.60 µm, minimal tearing, and a hardness of 268.3 HV at a feed of f = 0.1 mm/rev. This was most likely due to reduced friction during of tool–workpiece contact and reduced thermal stress due to the use of a coolant. Similar parameters can be found in a previous study [13]. Surfaces machined at higher feed rates, i.e., f = 0.3 mm/rev, such as S2, S4, S6, S8, S10, and S12, consistently showed higher R_a and R_z values, regardless of coolant use. An optical micrograph showed more tearing and deformation in dry conditions, suggesting high friction and heat generation during machining. S12, despite being machined under wet conditions, still showed high roughness, highlighting that the feed rate is more dominant than the use (or not) of coolant. Hardness generally increased with feed rate due to plastic deformation and strain hardening.

The material removal rate defines the observed correlation between feed rate and surface roughness during turning. At lower feed rates, the tool engages less material per revolution, leading to smaller chip thickness and reduced cutting force. This results in lower heat generation and plastic deformation near the surface, thus minimizing tearing and maintaining grain boundary integrity. The wet machining condition reduces friction and keeps the tool and workpiece material cool, which overall tends to increase the quality of the surface. In comparison, high feed rates increase the material removal rate, which often increases surface waviness and leads to the appearance of peaks and valleys in the surfaces. In these cases, a lower feed rate may provide a better surface finish. High spindle speeds yield improved surfaces, because at higher spindle speeds, the tool spends less time in contact with the workpiece, which reduces the chances of BUE formation on the tool, which can affect the surface. In addition, at higher speeds, heat dissipation also increases, preventing the workpiece and tool from building up heat. The use of coolant also improves surface finish, as it reduces friction, dissipating heat. It also prevents the workpiece from overheating, leading to a smoother surface during the machining.

In comparison with previous studies, it was observed that higher cutting speeds typically led to thermal softening of the material, thereby reducing surface hardness. In contrast, increased feed rates tended to generate greater heat, contributing to an increase in hardness. The present study highlights that feed rate has a more pronounced influence on surface hardness than cutting speed, as evidenced by the consistent increase in hardness with higher feed values [17,18]. Notably, the authors of [14] reported an R_a value of 0.77 µm at a cutting speed of 200 m/min, while in the current investigation, a lower R_a of 0.60 µm was achieved at a higher cutting speed of 282.7 m/min.

The consistency in the results of both TOPSIS and GRA validated the finding that the selection of a high spindle speed, i.e., N = 2000 rpm (v_c = 282.7 m/min), and a low feed, i.e., f = 0.1 mm/rev, in a wet machining environment achieves the best surfaces and is optimal for X5CrNi18-10 steel.

5. Conclusions

The study aimed to understand and evaluate the surface integrity of X5CrNi18-10 steel under varying machining parameters and environments and analyze variations thereof using MCDM techniques, i.e., TOPSIS and GRA. Based on our experimental analysis, the following conclusions were drawn.

• Machining parameters, i.e., spindle speed and feed rate, exhibited a strong influence on surface integrity. A high spindle speed, i.e., N = 2000 rpm (v_c = 282.7 m/min), and a low feed rate, i.e., f = 0.1 mm/rev, with a wet machining environment resulted in good surface finishes and minimum surface tearing. This parameter combination enhanced the quality of the machined surface while reducing the tearing deformation and thermal damage.
• Feed rate was the dominant factor influencing both surface roughness and hardness. An increase in feed rate increased the value of surface roughness parameters, which decreased the quality of the surface; however, the hardness increased due to higher plastic deformation.
• Wet machining condition improved the surface characteristics in X5CrNi18-10 turning. The use of coolant during machining helped with heat dissipation, reducing built-up edge formation and tool wear. This led to smoother chips and less surface tearing.
• Both TOPSIS and GRA ranked the optimal machining conditions as follows: N = 2000 rpm (v_c = 282.7 m/min) and f = 0.1 mm/rev, with a wet machining environment. As such, we recommend a high spindle speed, low feed, and wet machining conditions as an effective combination of parameters for good surface finishes in chromium nickel alloy turning.

These results may guide the selection of machining parameters for chromium–nickel alloy steel, where surface finish and mechanical reliability are critical, e.g., in the marine and aerospace industries.