Modeling and optimization of machining parameters in milling of INCONEL-800 super alloy considering energy,productivity,and quality using nanoparticle suspended lubrication

Abstract

Inconel-800 super alloy is a newly difficult-to-cut material. To improve the cutting conditions for this metal, sustainable methods in which minimum quantity lubrication enhanced with suspended nanoparticle were employed. This work also aims to model the relationship between process parameters (cutting speed, feed per tooth, depth of cut, and corner radius of cutting tool) and machining responses (surface roughness, specific cutting energy, cutting power, and material removal rate) using orthogonal array design of experiment and response surface methodology. Non-dominated sorting genetic algorithm was used to solve the multi-objective optimization problems in terms of energy, productivity, and quality of the machining process. The results indicate that the application of the response surface methodology model in combination with non-dominated sorting genetic algorithm is appropriate for this study due to the goodness of fit of response surface methodology and the global optimum solution of genetic algorithm. Because multi-objective optimization gives multiple solutions, Pareto plot and data mining are employed to support the selection of process parameters that can save time and cost and increase energy efficiency, meanwhile, simultaneously improve productivity and surface quality. The research results show that the specific cutting energy and energy consumption can be reduced up to 20.2% and 6.4%, respectively.

Keywords

Minimum quantity lubrication nanofluid specific cutting energy energy consumption machining optimization

Introduction

Super alloys play an important role in the manufacturing industry. A very important super alloy employed in the aerospace industry is a nickel–iron–chromium alloy, namely Inconel 800. Besides aerospace, this super alloy is also used in the defense, heat exchanger, automotive, and marine industries. The Inconel-800 super alloy has excellent properties (Table 1) than other materials consisting of eminent toughness, high tensile strength, good corrosion resistance, and high-temperature strength;^1–3 however, this super alloy is a difficult-to-cut material.⁴ Therefore, for machining the Inconel-800 super alloy, consideration of researchers’ finding is important because of the alloy’s poor thermal conductivity, hot hardness, and chemical reaction with cutting tools.⁴ A feasibility study for the machining of super alloys involves the use of auxiliary equipment to enhance machinability to achieve the best performance.⁵ Some studies include the application of laser-assisted machining (LAM).^6,7 This method makes use of a laser beam to increase the machinability of super alloys. However, the cost of investing in equipment for LAM is quite high. In addition, the laser beam can be dangerous for machine operators. Because of that, this method cannot be widely implemented. Other works have investigated some sustainable cooling and lubrication techniques, for instance, the minimum quantity lubrication (MQL) technique,⁸ cryogenic technique,⁹ and nanofluid MQL technique.⁴

Table 1.

Mechanical properties of Inconel 800 alloy.

Temperature (°C)	Tensile strength (MPa)	Yield strength (MPa)	Elongation (%)	R.A. (%)
27	448–549	172–226	30–48	76

The approach of nanofluids-assisted MQL is a high-grade technique that is being accepted by some model manufacturing industries due to its good characteristics.^10–13 The approach is capable of overcoming the disadvantages of dry machining and machining with a flood or MQL technique. Dry machining is economical as well as ecological; however, in dry machining, it is difficult to cut super alloy materials like Inconel-800 due to several limitations. It is the cause of poor surface quality and a higher cutting temperature leading to reduced tool life.^2,11,14 Cutting fluids in flood form can reduce the temperature in the cutting zone, but the use of a large amount of cutting fluid causes an increase in costs and many problems for human health.¹⁵ MQL is considered to be a cooling technique that replaces conventional lubrication methods with a very small quantity of fluid at a flow rate of 50–500 mL/h under the required high air pressure varying from 2 to 6 bar.^15–18 The MQL technique contributes to the reduction of the amount of lubricant, decreasing product costs, as well as increasing machining performances. The fluid used in the MQL technique should be decomposable and eco-friendly. Therefore, the MQL technique is an economical and environmentally friendly lubrication technique.^2,18–21 However, since the heat transfer coefficient of cutting fluid, such as oil, water, and an ethylene glycol mixture, is low, the temperature dissipation in the cutting zone of the lubrication is limited. The so-called nanofluids can enhance thermal conductivity up to twofold.²² Several other studies also illustrate this.^23–26 Furthermore, the addition of nanoparticles into the cutting fluid also increases the surface quality, lessens the cutting force, diminishes friction between two contact surfaces, and extends the tool life.^27–32 Therefore, this work focuses on using the MQL technique enhanced by nanoparticle suspended lubrication when end milling a new nickel-based alloy.

Energy reduction in the modern manufacturing industry is an essential requirement not only because the energy source in the world is gradually being exhausted but also because reducing energy consumption minimizes disadvantages to the living environment.^19,33 Considering energy consumption in the manufacturing industry, especially for machining tools, the milling process can be separated into two levels: machine enhancements and process parameter optimization.³⁴ The first approach is sometimes not feasible; the second is much more possible because it not only saves in cost but also demands less effort. Therefore, saving energy consumption by optimizing technological parameters in processing is an attractive concern for many researchers.

Specific cutting energy (U_c) is a quantity to evaluate the impact of the cutting process on the environment.^35,36U_c is the energy required to remove a quantity of material in a unit of time. Warsi et al. develop and analyze U_c-based process data and map facilitates high-speed machining of Al 6061 T6 alloy.³⁷ Jang et al.¹⁹ also research U_c to accomplish environmentally conscious manufacturing (ECM) in the MQL milling process. Campatelli et al. also focus on the worldwide diminution of environmental impacts by minimizing the power consumption including energy, specific energy, and total energy using response surface methodology (RSM) in the milling of carbon steel.³⁸

The literature review shows that researches on modeling and optimization of process parameters for saving energy in the field of metal processing draw particular attention to the manufacturing industries. The models usually used to render the relation between input parameters and technological responses are RSM, artificial neuron network (ANN), the Kriging model, and radial basis function (RBF). Among them, RSM is most commonly employed by researchers in engineering because of its flexibility and reliability.^34,39–41 After modeling the relationship between process parameters and machining responses, an optimization algorithm will be applied to solve the optimization problems. Some studies use the Taguchi method for optimization of energy consumption. For example, Camposeco-Negrete et al. used the Taguchi method to determine the level of three factors that affects energy consumption with a constant material removal rate.⁴² However, the limitations of the Taguchi method do not warrant a “real optimal” solution because it only addresses the discrete control factor.⁴³ Similarly, Gray relational analysis (GRA) does not offer a “real optimal” solution.⁴⁴ It is also optimal to determine the best level of the factor. Recently, the evolutionary optimization algorithm was applied in engineering as the particle swarm optimization (PSO), non-dominated sorting genetic algorithm (NSGA-II), and archive-based micro genetic algorithm (AMGA). Jang et al.¹⁹ used ANN couple PSO to optimize the single objective that is specific cutting energy. However, optimizing for a single objective function is not realistic in metal cutting. Nguyen employed the Kriging model and AMGA to optimize dry milling in terms of machining energy, surface quality, and production rate.⁴⁵ Park et al.³⁴ also optimized energy efficiency using the hybrid RSM and NSGA-II to determine the best solution for specific cutting energy and energy efficiency. It can be seen that the modeling and optimization for energy efficiency in the nanofluid MQL machining process are still of interest.

Optimization techniques applied to engineering are diverse and can be classified into two main groups: traditional and advanced optimization. Conventional optimization techniques can be involved Taguchi method, GRA, and RA coupled with RSM. The advanced optimization techniques consist of PSO, NSGA-II, AMGA, teaching-learning-based optimization (TLBO). The advanced optimization techniques are more robust than conventional ones because they base on global search algorithm and can solve both constrained and unconstrained multi-objective optimization. In addition, advanced optimization techniques can effectively solve complex optimization problems in which the global optimum result is guaranteed. The summaries of the literature review of conventional and advanced optimization techniques that were used in the engineering field are shown in Table 2. The application of a suitable optimization technique fosters the optimization process and gives a reliable optimal solution.

Table 2.

Summary of conventional and advanced optimization techniques.

Authors	Optimization techniques	Application areas	Remarks
Conventional optimizationtechniques
Pramanik et al.,⁴⁶Singh et al.,⁴⁷Vu et al.⁴⁸	Taguchi	WEDM, fused filamentfabrication, compliant	Taguchi method and ANOVA can be proved the significant impact of processing conditions of wire EDM. The results of Taguchi method are good agreement with Weibull analysis. Taguchi method is capable of evaluating the optimal value with acceptable deviation error by 7.15%.
Yan and Li⁴⁴	GRA	Milling	GRA is a strong tool for multi-objective optimization in machining to optimize cutting parameters.
Gajalakshmi et al.⁴⁹	GRA coupled with RSM	Dry sliding wear	GRA couple with RSM in optimizing dry sliding wear parameters can reduce wear, coefficient of friction, and frictional force.
Advanced optimization techniques
Ravichandran and Uthirapathy⁵⁰	TLBO	Position errors inCNC machine	TLBO optimization it can be significantly improved of R² for X, Y, Z axes are 1%–2.6%, 1%–14.3%, and 10.3%–71.1%, respectively
Gülhan and Erbatur⁵¹	GA	Robotics	The application of GA optimization technique has improved higher results than previous studies
Kavitha and Venkumar⁵²	SSO-GA	Job shop scheduling	Social spider optimization in conjunction with GA (SSO-GA) is effectively controlled to diminished makespan time compare with other hybrid of optimization algorithm
Prakash et al.^53,54	NSGA-II	PMEDM machining	NSGA-II was proposed to enhance the finding of the optimal technological responses simultaneously
Iqball et al.,⁵⁵, Prakash et al.,⁵⁶ Prakash et al.⁵⁷	PSO	Wind turbine system, EDM machining	PSO was used for stabilizing the system more quickly and effectively PSO technique gives the optimal input parameters and captures the influence of machining parameters. PSO improved the MRR and R_a respect to the worst case without optimization by 14.89% and 15.94%, respectively
Liu and Zhang⁵⁸	LS-SVM combine with PSO	Shield machine	Parallel cooperative particle swarm optimization-partial least squares support vector machine significantly enhanced in modeling capability and precision
Prakash et al.⁵⁹	Hybrid PSO-SA	Gas tungsten arcwelding	Hybrid PSO-simulated annealing algorithm is capable of enhancing significantly of the confidence interval by 95%

WEDM: wire electrical discharge machining; ANOVA: analysis of variance; GRA: gray relational analysis; RSM: response surface methodology; TLBO: teaching-learning-based optimization; CNC: computer numerical control; GA: genetic algorithm; SSO-GA: social spider optimization–genetic algorithm; NSGA-II: non-dominated sorting genetic algorithm; PMEDM: powder mixed electric discharge machining; PSO: particle swarm optimization; MRR: material removal rate; LS-SVM: least-squares support vector machine; PSO-SA: swarm optimization–simulated annealing.

According to the aforementioned literature reviews, it can be seen that overcoming the difficulties in machining in the new nickel-based alloy that has high applicability remains a big challenge. Furthermore, the application of MQL nanolubrication to enhance the machinability of the Inconel-800 super alloy, as well as multi-objective optimization to the proper selection of the cutting condition in terms of cutting energy reduction, high productivity, and good quality, is essential for researchers. Therefore, with this aim, this paper will perform multi-objective optimization of milling for Inconel-800 using carbon nanotube nanoparticle-based CT232 cutting oil in MQL in order to obtain the efficiency of energy, a good surface finish, and high productivity.

To obtain the optimal machining conditions for the milling of Inconel-800 alloy, the cutting parameters (cutting velocity, feed rate, depth of cut, and corner radius of cutting tool) which affect the technological responses (specific cutting energy, energy consumption, and surface roughness) should be figured out scientifically. To solve the research problem, an RSM model is employed to render the relationship between the cutting parameters and technological responses. Furthermore, the NSGA-II technique, an evolutional algorithm, will be utilized to solve the multi-objective optimization in this work. The remainder of this paper will present in detail the systematic research procedure, experimental setup, data processing, modeling, and process parameter optimization. Subsequently, the results and discussion will be shown, and finally, the conclusions and major contributions will be made.

Materials and methods

The workpiece used in the experiment is made of an Inconel-800 super alloy, with the chemical composition as tabulated in Table 3. A milling machine, namely, the Victor Vertical Machining Center 4, was utilized in this work. The type of milling operation is end milling (slot milling) and the type of depth of cut is axial depth of cut. The radial depth of cut was kept constant that equal to the diameter of the cutting tool. Workpiece dimensions were 225 mm (L) × 100 mm (W) × 60 mm (H) with the 100 mm dimension being the cutting length. The tool holder, namely a CoroMill R390 square shoulder milling cutter with two flutes, was used to perform all experiments. The inserts were entitled R390-11 T3 with three kinds of radii of noses consisting of 0.4, 0.8, and 1.2 mm. The diameter of the milling tool was Φ16 mm (Figure 1).

Table 3.

Chemical composition (wt%) of Inconel-800 super alloy.

Element	Ni	Cr	Fe	C	Al	Ti	Ti + Al	Mn
% weight	30.0–35.0	19.0–23.0	46.99	0.1 max	0.5	0.15–0.60	1.01	1.5 max

Figure 1.

Cutting tool for experiments.

Experiments were processed under minimum lubrication conditions with a flow rate of MQL oil, angle of nozzle, air pressure, and distance of nozzle parameters fixed as 100 mL/h, 60°, 3 kg/cm², and 20 mm, respectively. Multi-walled carbon nanotube (MWCNT) with particle sizes smaller than 30 nm was dissolved into the cutting oil CT 232. This cutting oil can increase accuracy, the efficiency of cost and energy, and tool life. It also has higher stability and difficult to deterioration. Specifications of CT232 cutting oil are described in Table 4. The percent of MWCNT in the cutting oil was 0.5 vol%. The selection of nanoparticle concentrations is based on the initial experiments of this work, and the reference of related studies.^60,61 The solution was stirred continuously using a magnetic stirrer for about 48 h. MWCNT was chosen because it has higher thermal conductivity (300 W/m K) than other nanoparticles.⁶² The machining environment of the experiments is shown in Table 5.

Table 4.

Specifications of CT232 cutting oil.

Color	Viscosity cSt. at 40°C	Flash point (°C)	Fatty (%)	CL (%)	Stability and resistance	Copper corrosion, 100°C, 1 h
Yellow	26	200	11	9.6	Good	1

Table 5.

Machining environment of experiments.

1	Machine	Victor Vertical Machining Center 4
2	Workpieces	Material: Inconel 800
		Dimensions: 225 mm (L) × 100 mm(W) × 60 mm (H)
3	Cutting tools	Diameter: Φ16 mmCoroMill R 390 square shouldermilling cutter (R390-016A16-11L)Insert: R390-11 T3 04,08,12
4	Cooling condition	MQL with MWCNT nanofluidDistance of nozzle: 20 mmAir pressure: 3 kg/cm²Angle of nozzle: 60°Flow rate: 100 mL/hCutting oil: Super E.P. CT232MWCNT with size < 30 nm.Concentration in lubrication: 0.5%

MQL: minimum quantity lubrication; MWCNT: multi-walled carbon nanotube; CL: Chlorine.

Regarding the design of experiment (DOE), this work designates four main controllable variables, including cutting velocity (v_c), feed per tooth (f_z), depth of cut (a_p), and corner radius (mm). The values of the input parameters are divided into three levels in which the levels are chosen based on the cutting tools recommendation and the experience of operators. The factors and their levels are presented in Table 6. Figure 2 is systematic procedure of modeling and optimization of machining parameters for this work. To reduce the number of experiments and to save experimental costs, the orthogonal array design is assumed to generate a set of 27 experiments, as shown in Table 7.

Table 6.

The factors and their levels.

Factor	Units	Levels
		1	2	3
Cutting velocity (v_c)	m/min	90	120	150
Feed per tooth (f_z)	mm/tooth	0.03	0.06	0.09
Depth of cut (a_p)	mm	0.4	0.7	1.0
Corner radius (r)	mm	0.4	0.8	1.2

Figure 2.

Systematic procedure of modeling and optimization of machining parameters.

Table 7.

DOE and the results of experiment.

No.	Machining parameters				Machining responses
	v_c (m/min)	(a_p) (mm)	f_z (mm/tooth)	r (mm)	R_a (µm)	P_c (W)	U_c (J/mm³)	MRR (mm³/min)
1	90	0.4	0.03	0.4	0.206	636.8	0.926	687.9
2	150	0.4	0.09	0.4	0.507	817.6	0.238	3439.5
3	120	0.4	0.06	0.4	0.351	727.2	0.396	1834.4
4	120	0.7	0.03	0.4	0.342	724.0	0.451	1605.1
5	150	0.7	0.06	0.4	0.368	867.2	0.216	4012.7
6	90	0.7	0.09	0.4	0.434	815.2	0.226	3611.5
7	150	1.0	0.03	0.4	0.441	812.8	0.284	2866.2
8	90	1.0	0.06	0.4	0.450	804.8	0.234	3439.5
9	120	1.0	0.09	0.4	0.599	1011.2	0.147	6879.0
10	120	0.4	0.03	0.8	0.201	659.2	0.719	917.2
11	150	0.4	0.06	0.8	0.297	766.4	0.334	2293.0
12	90	0.4	0.09	0.8	0.363	712.0	0.345	2063.7
13	150	0.7	0.03	0.8	0.313	776.8	0.387	2006.4
14	90	0.7	0.06	0.8	0.342	734.4	0.305	2407.6
15	120	0.7	0.09	0.8	0.504	850.4	0.177	4815.3
16	90	1.0	0.03	0.8	0.453	698.4	0.406	1719.7
17	120	1.0	0.06	0.8	0.531	856.8	0.187	4586.0
18	150	1.0	0.09	0.8	0.620	1053.6	0.123	8598.7
19	150	0.4	0.03	1.2	0.164	728.0	0.635	1146.5
20	90	0.4	0.06	1.2	0.224	662.4	0.481	1375.8
21	120	0.4	0.09	1.2	0.383	732.8	0.266	2751.6
22	90	0.7	0.03	1.2	0.303	660.0	0.548	1203.8
23	120	0.7	0.06	1.2	0.400	790.4	0.246	3210.2
24	150	0.7	0.09	1.2	0.500	927.2	0.154	6019.1
25	120	1.0	0.03	1.2	0.504	778.4	0.339	2293.0
26	150	1.0	0.06	1.2	0.504	957.6	0.167	5732.5
27	90	1.0	0.09	1.2	0.541	954.4	0.162	5159.2

DOE: design of experiment; MRR: material removal rate.

The machining responses which depend on the cutting parameters are surface roughness (R_a), specific cutting energy (U_c), cutting power (P_c), and material removal rate (MRR). The Mitutoyo SJ-400 surf-test instrument was employed to measure the surface roughness. The voltage and current were measured by a Lutron 3 Phase Power Analyzer DW-6092. Three-phase and four-wire electrical circuits of power analyzers were connected with suitable direct to get signals of voltage and current. These signals were processed through Lutron data acquisition software installed on the computer. From there, the values of power consumption were determined. The cutting force was estimated by a piezoelectric and a Vishay 6100 scanner. The signals were collected on the computer and processed by installed software. The MRR (mm³/min) in cutting process be calculated from equation (1)

MRR = \frac{a_{p} * a_{e} * v_{c} * f_{z} * z * 1000}{3.14 * d}

(1)

where a_e is the width of cut (mm), z is the flute of cutter, and d is diameter of cutting tool (mm). The details of the experimental setup are illustrated in Figure 3.

Figure 3.

The details of the experimental setup.

Generating energy in a certain way will adversely affect the environment, so (U_c) is often adopted to assess the influence of machining on the environment.^19,63 The magnitude of the specific cutting energy (U_c) is calculated as¹⁹

U_{c} = \frac{F_{c} * v_{c}}{MRR}

(2)

where F_c denotes the total cutting force.

To demonstrate the relationship between the machining parameters and technological responses, this paper uses RSM model that is widely employed for modeling and analyzing in mechanical engineering. In this model, the second order is utilized to approximate the response as follows

\begin{matrix} ψ = c_{0} + \sum_{i = 1}^{k} c_{i} X_{i} + \sum_{i = 1}^{k} c_{ii} X_{i}^{2} + \sum_{i = 1}^{k - 1} \sum_{j = i + 1}^{k} c_{ij} X_{i} X_{j} + ε \end{matrix}

(3)

where (c₀) is the constant, and (c_i), (c_ii), and (c_ij) are coefficients. (X_i) and (X_j) are the variables. (ε) is the random error of the experiment.

Afterward, an evolutional multi-objective optimization algorithm frequently used to optimize in mechanical engineering is the NSGA-II applied to solve the optimization problems given by the author. The framework of NSGA-II methodology to optimal cutting parameters in this paper is shown in Figure 4.

Figure 4.

Framework of NSGA-II methodology to optimize cutting parameters.

The optimal objective of this work is to minimize energy consumption and maximize surface quality or maximize machining productivity (material removal rate). Therefore, two scenarios are considered:

The first scenario, we minimize U_c and R_a.

The second scenario P_c is minimized and MRR is maximized with constraint of surface roughness.

Results and discussion

The data for the physical experimental results in this paper were collected and analyzed to render a relationship among process parameters and surface roughness, specific cutting energy, and cutting power. The results of 27 experiments based on the Taguchi orthogonal array are presented in Table 6. As mentioned in the “Materials and methods” section, RSM was employed to capture the influence of input parameters on technological responses based on these data.

The authenticity of the RSM model to the experimental data is displayed by the coefficient of determination for regression analysis (R²). The coefficient values of R² for R_a, P_c, and U_c were 0.959, 0.953, and 0.970 respectively. It can be seen that the coefficient values of R² for the RSM model are greater than 0.9 and are shown in Figure 5. Therefore, it can be concluded that the Kriging model is an appropriate model.

Figure 5.

Confirmation of the suitability of RSM model estimated by the coefficient R².

Figures 6, 8, and 9 depict the three-dimensional (3D) graph showing the influence of the input parameters (v_c, a_p, f_z, r) on the machining response (R_a, P_c, U_c), respectively. Figure 6 shows the influence of the process parameters on the surface roughness. From this figure, it can be inferred that R_a increases with a corresponding rise in a_p and f_z. On the other hand, increasing v_c and r leads to a decrease in R_a. It can be explained that machining at lower cutting velocity causes a built-up edge (BUE) phenomena on the cutting edge of the cutting tool. The BUE is formed which harms the cutting edge, lowering the surface roughness quality. When the corner radius ranges from 0.4 to 1.2 mm, the quality of the surface roughness is markedly improved. This phenomenon helps us understand that increasing the corner radius leads to a rising plowing effect that expands the material side flow on the surface finish.⁶⁴ Cutting with a large a_p increases the cutting force and cutting temperature. An increase in f_z causes the expansion of the furrows which are generated by the rotation of the cutting tool on the machining surface. These have led to reduced surface roughness quality. The results agree well with previous studies.^2,65

Figure 6.

Surface roughness versus machining parameter.

Some experiments on the wear of inserts with different corner radii as the levels of the corner radius are performed. The results indicated that the higher the corner radius, the greater the wear of inserts. Figure 7 shows that abrasion phenomena such as crater wear or flank wear in milling inserts are increased as corner radius increases. This is in line with the metal cutting theory.

Figure 7.

Optical micrographs of worn inserts: (a, b) R = 0.4 (mm), (c, d) R = 0.8 (mm), and (e, f) R = 1.2 (mm).

It is found from the 3D graph shown in Figure 8 that increasing the feed rate, depth of cut as well as the corner radius creates an increase in cutting power. This can be interpreted as follows. An increase in feed rate and depth of cut results in an enlargement in the MRR per unit of time, which means that the cutting force or cutting power is raised. Our results are in line with previous works.^40,66

Figure 8.

Cutting power versus machining parameter.

Figure 9 indicates the effect of the machining process on specific cutting energy (U_c). It can be seen that U_c increases as a_p and f_z decrease. This can be explained by the fact that the energetic effect of the machined volume increment is greater than that of the cutting energy.¹⁹ However, machining with larger a_p and f_z values leads to an increased cutting force. This causes a reduced tool life, so these parameters need to be limited. Figure 9 also shows that v_c and r have little effect on U_c.

Figure 9.

Specific cutting energy versus machining parameter.

The specific cutting energy U_c of Inconel-800 super alloy is compared with other works.^1,4,8 The comparison shows that U_c of milling and turning is little different (U_c range from 1.16 to 1.86 J/mm³ for turning compared to 0.12–0.93 J/mm³ for milling). The reason for these slight differences is due to the difference between the cutting movement and geometry of the milling and turning tools. In addition, the specific cutting energy U_c in this work is smaller than that of Gupta et al.¹ because of the contribution of MQL lubrication with MWCNT. This technique has a good lubrication effect so it diminishes the friction leading to a reduction in cutting force. It can be concluded that our research results are reliable. Our study is intended as a contribution to demonstrate the superiority of MWCNT for MQL lubrication.

Figure 10 presents the global effects (or can be called the main effect) of process parameters on the technological response. The values on the horizontal axis of Figure 10 are calculated by percentage (%). As seen in Figure 10, the greatest effects on R_a, P_c, and U_c are depth of cut, cutting velocity, and feed per tooth, respectively. Following that, in terms of influence to technological response, R_a, P_c, U_c are feed per tooth and depth of cut. The rest of the parameters also have significant or little effect on the machining response.

Figure 10.

The global effect of process parameters on (a) Ra, (b) P_c, and (c) U_c (%).

Regression equations that were determined based on RSM model for surface roughness, cutting power, and specific cutting energy are presented in Table 8.

Table 8.

Regression equation for surface roughness, cutting power, and specific cutting energy.

Response	Regression equations
R_a	–0.35 + 0.009 v_c + 0.125 a_p – 0.875 f_z – 0.07 r + 0.20 a_p² + 27.4 f_z² – 0.009 r² – 0.0003 v_c a_p + 0.024 v_c f_z – 0.0003v_c r – 2.85 a_p f_z + 0.20 a_p r – 0.63 f_z r	(4)
P_c	633.11 – 0.06v_c – 226.64a_p – 93.63f_z – 131.39r + 0.008v_c² + 48.89a_p² + 888.89 f_z² + 66.67r² + 0.7v_ca_p – 3.52v_c f_z – 0.2v_cr + 4616.3a_pf_z + 89.93a_pr – 287.41f_zr	(5)
U_c	2.98 – 0.01v_c – 2.09a_p – 24.59f_z – 0.34r + 0.60a_p² + 86.48f_z² + 0.05r² + 0.003 v_c a_p + 0.03v_c f_z + 0.0015v_c r + 7.22a_p f_z + 0.068a_p r + 0.228f_z r	(6)

After modeling using the RSM model, three available statistical regression equations show the relationship between input parameters and output parameters. Based on these equations, multi-objective optimization was employed to find the optimal solutions for scenarios presented under Materials and methods. The NSGA-II algorithm is implemented to find the optimal solution for this problem. The parameters of the NSGA-II were recorded as follows:

Population size = 20;

Number of generations = 50;

Crossover probability = 0.9;

Crossover distribution index = 10;

Mutation distribution index = 20.

What are the optimal solutions of machining parameters that minimize U_c and R_a is a study question for the first scenario. Figure 11 shows the Pareto plot of U_c and R_a. The blue points represent the Pareto point, and the black points show a feasible solution. The results show that when R_a increases, U_c decreases and vice versa. It can be seen that these two optimal objectives are conflict. The selection of the optimal value on the Pareto plot depends on the machine operator making decisions based on the trade-off between the two objectives. Figure 11 also shows the advantages of using NSGA-II optimally over arbitrary options. For example, for the same R_a = 0.35 (µm), the variation of U_c (ΔU_c) can be up to 0.268–0.214 = 0.054. This optimum result is improved over the other result by around (0.054/0.268) × 100 = 20.2%.

Figure 11.

Pareto front generated by NSGA-II base on RSM model for the first scenario.

How can one determine the corresponding input parameters after selecting the optimal solution? The answer is shown in Figure 12 which displays data mining to find the optimal process parameters for the first scenario. The iSight version 5.8 software was used for data mining. For example, when the optimal solutions selected in Figure 11 are U_c = 0.453 (J/mm³), and R_a = 0.213 (µm), the corresponding optimal input parameters to those response parameters are v_c = 149.91 (m/min), f_z = 0.044 (mm/tooth), a_p = 0.44 (mm), and r = 1.2 (mm).

Figure 12.

Data mining for finding the optimal process variable for the first scenario.

In the second scenario, multi-objective optimization problem that minimizes P_c and maximizes MRR while R_a is the predetermined constraint. Figure 13 shows the Pareto front generated by the NSGA-II based on the RSM model for the second scenario. Figure 13(a) and (b) only shows the Pareto point (optimal solution), and this demonstrates the variations in P_c and MRR according to the predefined R_a. When R_a increases, P_c and MRR increase accordingly. When R_a reaches a constraint value, the optimal solutions for P_c and MRR still exist but are not conducive to saving energy consumption. If MRR continues to increase to 4481.2 (mm³/min), P_c will reach 851.33 (W). For MRR, Point B at the right end of the Pareto front line (Figure 13(b)) is the best optimal point, but for cutting energy, point A at the end of the slope is the best optimal point (Figure 13(a)). However, the objective of second scenario is to increase energy efficiency, so the optimum selection point will not be selected in the ΔP_c₁ region. ΔP_c₁ is the magnitude that evaluates the effectiveness of optimization and the Pareto front and can be calculated as (851.33 – 752.24)/851.33 = 11.64%.

Figure 13.

Pareto front generated by NSGA-II based on RSM model for the second scenario.

In Figure 13(c), the variability relationship between P_c and MRR is shown adequately and accurately. The objective of the milling process is always machining with the highest MRR, and the smallest P_c but still achieving the desired surface quality. However, these two response parameters (P_c and MRR) conflict with each other so there should be a compromise to select the optimal solution. The Pareto point and feasible optimal solutions are exhibited in Figure 13(c). From this figure, the milling operator can select the optimal solution that suits the requirements. As in the first scenario, after selecting the desired technological responses based on the data mining in Figure 14, one is able to determine the input parameters corresponding to these technological responses. For example, when selecting an optimal point through Figure 13(c) with MRR = 3141.1 (mm³/min) and P_c = 685.67 (W), based on data mining, the optimal machining parameters can be determined as v_c = 137.44 (m/min), f_z = 0.086 (mm/tooth), a_p = 0.4 (mm), and r = 1.2 (mm).

Figure 14.

Data mining to find the optimal process variable for the second scenario.

A validation test is conducted based on the above data; the physical experimental results are shown in Table 9. From Table 9, it can be seen that the relative errors between the predicted value and the experimental value for R_a, P_c are 5.8% and 3.9%. The magnitude of relative errors is considered acceptable in machining. This again proves that the application of the RSM model combined with the NSGA-II optimal algorithm for this study is very reliable. Besides, this work also shows machining at an optimal value to compare two cutting conditions (nanofluid MQL and pure MQL) in terms of energy efficiency. The results indicate that machining under nanofluid MQL can improve energy efficiency compared to using pure MQL. The values of the effectiveness are shown in Table 10. The surface roughness reduces 8.4%, and Pc and Uc reduce 6.1% and 10.5%, respectively. It is clear that MQL with nanoparticle not only help to protect environment but also increase the machining quality and save machining energy.

Table 9.

Verification result at an optimal value.

	R_a	P_c	U_c
Predicted	0.390	685.67	0.259
Real experiment	0.414	713.21	0.248
Relative error (%)	–5.8	–3.9	4.3

Table 10.

Comparison between the two cutting condition (nanofluid MQL and pure MQL) at an optimal value.

	R_a	P_c	U_c
Nanofluid MQL	0.414	713.21	0.248
Pure MQL	0.452	759.45	0.277
Improvement (%)	8.4	6.1	10.5

MQL: minimum quantity lubrication.

Another contribution shown in Figure 13(c) is the evaluation of the improvement of the application of the NSGA-II algorithm based on the RSM model compared to the non-application of optimal algorithms. As seen in Figure 13(c), ΔP_c₂ is the deviation that shows the application and non-application of the optimal algorithm, at the same MRR = 1517.1 (mm³/min), the optimal solution for P_c is 732.64 (W), while the arbitrary un-optimized solution is 685.65 W. Therefore, the improvement in energy consumption is (732.64 – 685.65)/732.64 = 6.4%.

Conclusion

This paper addresses the modeling and optimization of machining parameters in the milling of new nickel-based alloy in terms of energy, productivity, and quality. The new point of this work is the combination of the study of milling of Inconel 800 and the study of cutting of hard-to-cut material in which the emerging MQL with carbon nanotube suspended lubrication. The use of MQL with MWCNT increases the tribology so that reduced the cutting force in specific cutting energy. In addition, this work applied a multi-objective optimization technique that facilitates the optimization process. Pareto plot and data mining methods are proposed in order to select the appropriate trade-off solution when solving the multi-objective optimization problem in metal cutting because the objectives are almost conflicting. The optimal front of Pareto plot helps the researcher visually determines the best solution in which the MRR is maximized; meanwhile, the cutting power is minimized with a predefined constraint of surface roughness. Some of the main conclusions drawn by this study are as follows:

The application of the multi-objective evolution algorithm (NSGA-II) in conjunction with RSM regression model is a robust approach for modeling the relationship between machining parameters and technological responses and for solving machining process parameter optimization. The optimal results indicated that the specific cutting energy could be reduced roughly by 20.2%, while energy consumption could be reduced by 6.4% compared to arbitrary cases without applying optimization.

The Pareto front plot and engineering data mining generated multi-objective optimization facilitate the selection of appropriate machining parameters to increase the efficient energy consumption, as well as maximize the material removal rates.

The application of nanofluid MQL with MWCNT and the optimization enhance the energy efficiency because the specific cutting energy is decreased and is in the low range (below 0.93 J/mm³). This is a significant contribution to environmentally conscious machining.

Nanofluid MQL with MWCNT is better than pure MQL when milling at the optimal point for the new nickel-based alloy. It not only protects the environment but also increases technological responses. The surface roughness, specific cutting energy, and cutting power can reduce 8.4%, 6.1%, and 10.5%, respectively.

The research model in this work can be effectively applied as a paradigm to study the optimization of metal cutting.

The future work will focus intensively on multi-objective optimization of Inconel-800 in which tool life and tool wear are included. In addition, the specific cutting energy map for milling of this kind of material should be built for a better understanding of this material in terms of cutting energy for the study on the energy efficiency of metal cutting of hard-to-cut material.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors acknowledge and thank the Ministry of Education of the Republic of China for their partial financial support of this study under Contract Number E0002B and the Ministry of Science and Technology of the Republic of China under Contract Number MOST 108-2622-E-992-009-CC3.

ORCID iDs

Ngoc-Chien Vu

Shyh-Chour Huang

References

Gupta

Pruncu

Mia

, et al. Machinability investigations of Inconel-800 super alloy under sustainable cooling conditions. Materials 2018; 11: E2088.

Gupta

Sood

. Machining comparison of aerospace materials considering minimum quantity cutting fluid: a clean and green approach. P I Mech Eng C J Mech 2017; 231: 1445–1464.

Singh

Gupta

Mia

, et al. Modeling and optimization of tool wear in MQL-assisted milling of Inconel 718 superalloy using evolutionary techniques Int J Adv Manuf Tech 2018; 97: 481–494.

Gupta

Jamil

Wang

, et al. Performance evaluation of vegetable oil-based nano-cutting fluids in environmentally friendly machining of inconel-800 alloy. Materials 2019; 12: 2792.

Malik

Pradhan

Mann

, et al. Subtractive versus hybrid manufacturing. Encyclopedia Renew Sust Mater 2019; 5: 474–502.

Venkatesan

Ramanujam

. Statistical approach for optimization of influencing parameters in laser assisted machining (LAM) of Inconel alloy. Measurement 2016; 89: 97–108.

Venkatesan

Ramanujam

Kuppan

. Investigation of machinability characteristics and chip morphology study in laser-assisted machining of Inconel 718. Int J Adv Manuf Tech 2017; 91: 3807–3821.

Gupta

Mia

Pruncu

, et al. Parametric optimization and process capability analysis for machining of nickel-based superalloy. Int J Adv Manuf Tech 2019; 102: 3995–4009.

Mia

Singh

Gupta

, et al. Influence of Ranque-Hilsch vortex tube and nitrogen gas assisted MQL in precision turning of Al 6061-T6. Precis Eng 2018; 53: 289–299.

10.

Borade

Kadam

. Comparison of main effect of vegetable oil and Al₂O₃ nanofluids used with MQL on surface roughness and temperature. Int J Mech Eng Tech 2016; 7: 203–213.

11.

Gupta

Sood

Sharma

. Optimization of machining parameters and cutting fluids during nano-fluid based minimum quantity lubrication turning of titanium alloy by using evolutionary techniques. J Clean Prod 2016; 135: 1276–1288.

12.

Hadi

Atefi

. Effect of minimum quantity lubrication with gamma-Al₂O₃ nanoparticles on surface roughness in milling AISI D3 steel. Indian J Sci Tech 2015; 8: 130–135.

13.

T-V

N-C

Nguyen

Q-M

. Optimization of cooling conditions and cutting parameters during hard milling of AISI H13 steel by using Taguchi method. In: Proceedings of the 2018 IEEE international conference on advanced manufacturing (ICAM), Yunlin, 16–18 November 2018, pp. 396–398. Piscataway, NJ: IEEE.

14.

Pradhan

Singh

Prakash

, et al. Investigation of machining characteristics of hard-to-machine Ti-6Al-4V-ELI alloy for biomedical applications. J Mater Res Tech 2019; 8: 4849–4862.

15.

Nouioua

Yallese

Khettabi

, et al. Investigation of the performance of the MQL, dry, and wet turning by response surface methodology (RSM) and artificial neural network (ANN). Int J Adv Manuf Tech 2017; 93: 2485–2504.

16.

Razak

Rahman

Noor

, et al. A review of minimum quantity lubricant on machining performance. In: Proceedings of the national conference in mechanical engineering research and postgraduate students, FKM Conference Hall, UMP, Kuantan, Malaysia, 26–27 May 2010, pp. 72–85.

17.

Sharma

Singh

Sørby

. A review on minimum quantity lubrication for machining processes. Mater Manuf Process 2015; 30: 935–953.

18.

Hsu

. Optimization of minimum quantity lubricant conditions and cutting parameters in hard milling of AISI H13 steel. Appl Sci 2016; 6: 83.

19.

Jang

D-y

Jung

Seok

. Modeling and parameter optimization for cutting energy reduction in MQL milling process. Int J Precis Eng Manuf-Green Tech 2016; 3: 5–12

20.

Debnath

Reddy

. Environmental friendly cutting fluids and cooling techniques in machining: a review. J Clean Prod 2014; 83: 33–47.

21.

Sharma

Tiwari

Dixit

. Effects of minimum quantity lubrication (MQL) in machining processes using conventional and nanofluid based cutting fluids: a comprehensive review. J Clean Prod 2016; 127: 1–18.

22.

Chol

Estman

. Enhancing thermal conductivity of fluids with nanoparticles. ASME Int Mech Eng Cong Expos 1995; 231: 99–106.

23.

Raja

Vijayan

Dineshkumar

, et al. Review on nanofluids characterization, heat transfer characteristics and applications. Renew Sustain Energ Rev 2016; 64: 163–173.

24.

Kakaç

Pramuanjaroenkij

. Review of convective heat transfer enhancement with nanofluids. Int J Heat Mass Transf 2009; 52: 3187–3196.

25.

Zhang

, et al. Heat transfer performance of MQL grinding with different nanofluids for Ni-based alloys using vegetable oil. J Clean Prod 2017; 154: 1–11.

26.

Ali

Babar

Shah

, et al. Preparation techniques of TiO₂ nanofluids and challenges: a review. Appl Sci 2018; 8: 587.

27.

Shahnazar

Bagheri

Abd Hamid

. Enhancing lubricant properties by nanoparticle additives. Int J Hydrog Energy 2016; 41: 3153–3170.

28.

Sundar

Sharma

Singh

, et al. Hybrid nanofluids preparation, thermal properties, heat transfer and friction factor—a review. Renew Sust Energy Rev 2017; 68: 185–198.

29.

Songmei

Xuebo

Guangyuan

, et al. A novel approach of applying copper nanoparticles in minimum quantity lubrication for milling of Ti-6Al-4V. Adv Prod Eng Manag 2017; 12: 139–150.

30.

Sayuti

Sarhan

Hamdi

. An investigation of optimum SiO₂ nanolubrication parameters in end milling of aerospace Al6061-T6 alloy. Int J Adv Manuf Tech 2013; 67: 833–849.

31.

Sharma

Tiwari

Singh

, et al. Tribological investigation of TiO₂ nanoparticle based cutting fluid in machining under minimum quantity lubrication (MQL). Mater Today Proc 2016; 3: 2155–2162.

32.

Sharma

Singh

Dixit

, et al. Characterization and experimental investigation of Al₂O₃ nanoparticle based cutting fluid in turning of AISI 1040 steel under minimum quantity lubrication (MQL). Mater Today Proc 2016; 3: 1899–1906.

33.

Wang

Liu

Song

, et al. Proper selection of cutting parameters and cutting tool angle to lower the specific cutting energy during high speed machining of 7050-T7451 aluminum alloy. J Clean Prod 2016; 129: 292–304.

34.

Park

H-S

Nguyen

T-T

Dang

X-P

. Multi-objective optimization of turning process of hardened material for energy efficiency. Int J Precis Eng Manuf 2016; 17: 1623–1631.

35.

Osman

Yılmaz

Ünver

HÖ

, et al. Slot milling of titanium alloy with hexagonal boron nitride and minimum quantity lubrication and multi-objective process optimization for energy efficiency. J Clean Prod 2020; 258: 120739.

36.

Tang

Liu

, et al. Optimising the machining time, deviation and energy consumption through a multi-objective feature sequencing approach. Energ Conver Manag 2018; 160: 126–140.

37.

Warsi

Ahmad

Jaffery

SHI

, et al. Development of specific cutting energy map for sustainable turning: a study of Al 6061 T6 from conventional to high cutting speeds2020; 106: 2949–2960.

38.

Campatelli

Lorenzini

Scippa

. Optimization of process parameters using a response surface method for minimizing power consumption in the milling of carbon steel. J Clean Prod 2014; 66: 309–316.

39.

Huang

Nguyen

. Multi-objective optimization of surface roughness and cutting forces in hard milling using Taguchi and response surface methodology. Key Eng Mater 2018; 773: 220–224.

40.

Selaimia

A-A

Yallese

Bensouilah

, et al. Modeling and optimization in dry face milling of X2CrNi18-9 austenitic stainless steel using RMS and desirability approach. Measurement 2017; 107: 53–67.

41.

Prakash

Singh

Pabla

, et al. Bio-inspired low elastic biodegradable Mg-Zn-Mn-Si-HA alloy fabricated by spark plasma sintering. Mater Manuf Process 2019; 34: 357–368.

42.

Camposeco Negrete

Nájera–C Miranda-Valenzuela

. Optimization of cutting parameters to minimize energy consumption during turning of AISI 1018 steel at constant material removal rate using robust design. Int J Adv Manuf Tech 2016; 83: 1341–1347.

43.

Dang

X-P

. Constrained multi-objective optimization of EDM process parameters using kriging model and particle swarm algorithm. Mater Manufact Process 2018; 33: 397–404.

44.

Yan

. Multi-objective optimization of milling parameters—the trade-offs between energy, production rate and cutting quality. J Clean Prod 2013; 52: 462–471.

45.

Nguyen

T-T

. Prediction and optimization of machining energy, surface roughness, and production rate in SKD61 milling. Measurement 2019; 136: 525–544.

46.

Pramanik

Islam

Basak

, et al. Optimizing dimensional accuracy of titanium alloy features produced by wire electrical discharge machining. Mater Manuf Process 2019; 34: 1083–1090.

47.

Singh

Prakash

, et al. Optimization and reliability analysis to improve surface quality and mechanical characteristics of heat-treated fused filament fabricated parts. Int J Adv Manuf Tech 2019; 102: 1521–1536.

48.

Huynh

Huang

. Optimization the first frequency modal shape of a tensural displacement amplifier employing flexure hinge by using Taguchi Method. J Phys Conf Series 2019; 1303: 012016.

49.

Gajalakshmi

Senthilkumar

Prabu

. Multi-response optimization of dry sliding wear parameters of AA6026 using hybrid gray relational analysis coupled with response surface method. Meas Control 2019; 52: 540–553.

50.

Ravichandran

Uthirapathy

. Measurement of error in computer numerical control machines and optimization using teaching–learning-based optimization algorithm. Meas Control 2019; 52: 929–937.

51.

Gülhan

Erbatur

. Kinematic arrangement optimization of a quadruped robot with genetic algorithms. Meas Control 2018; 51: 406–416.

52.

Kavitha

Venkumar

. A vibrant crossbreed social spider optimization with genetic algorithm tactic for flexible job shop scheduling problem. Meas Control 2020; 53: 93–103.

53.

Prakash

Kansal

Pabla

, et al. Multi-objective optimization of powder mixed electric discharge machining parameters for fabrication of biocompatible layer on β-Ti alloy using NSGA-II coupled with Taguchi based response surface methodology. J Mech Sci Tech 2016; 30: 4195–4204.

54.

Prakash

Kansal

Pabla

, et al. To optimize the surface roughness and microhardness of β-Ti alloy in PMEDM process using non-dominated sorting genetic algorithm-II. In: Proceedings of the 2015 2nd international conference on recent advances in engineering & computational sciences (RAECS), Chandigarh, India, 21–22 December 2015, pp. 1–6. Piscataway, NJ: IEEE.

55.

Iqbal

Ying

Saleem

, et al. Proposed particle swarm optimization technique for the wind turbine control system. Meas Control. Epub ahead of print 21 February 2020. DOI: 10.1177/0020294020902785.

56.

Prakash

Singh

, et al. Multi-objective particle swarm optimization of EDM parameters to deposit HA-coating on biodegradable Mg-alloy. Vacuum 2018; 158: 180–190.

57.

Prakash

Singh

, et al. Multi-objective optimization of MWCNT mixed electric discharge machining of Al–30SiC p MMC using particle swarm optimization. In: Sidhu

Bains

Zitoune

, et al. (eds) Futuristic composites. Singapore: Springer, 2018, pp. 145–164.

58.

Liu

Zhang

. Earth pressure prediction in sealed chamber of shield machine based on parallel least squares support vector machine optimized by cooperative particle swarm optimization. Meas Control 2019; 52: 758–764.

59.

Prakash

Singh

, et al. Multi-objective parametric appraisal of pulsed current gas tungsten arc welding process by using hybrid optimization algorithms. The Int J Adv Manuf Tech 2019; 101: 1107–1123.

60.

Rahmati

Sarhan

AAD

Sayuti

. Investigating the optimum molybdenum disulfide (MoS2) nanolubrication parameters in CNC milling of AL6061-T6 alloy. The Int J Adv Manuf Tech 2014; 70: 1143–1155.

61.

Vamsi Krishna

Srikant

Nageswara Rao

. Experimental investigation on the performance of nanoboric acid suspensions in SAE-40 and coconut oil during turning of AISI 1040 steel. Int J Mach Tool Manuf 2010; 50: 911–916.

62.

Sharmin

Gafur

Dhar

. Preparation and evaluation of a stable CNT-water based nano cutting fluid for machining hard-to-cut material. SN Appl Sci 2020; 2: 626

63.

Kuram

Ozcelik

Bayramoglu

, et al. Optimization of cutting fluids and cutting parameters during end milling by using D-optimal design of experiments. J Clean Prod 2013; 42: 159–166.

64.

Subramanian

Sakthivel

Sudhakaran

, et al. Modeling and analysis of surface roughness of AL7075-T6 in end milling process using response surface methodology. Arabian J Sci 2014; 39: 7299–7313.

65.

Nguyen

Hsu

. Surface roughness analysis in the hard milling of JIS SKD61 alloy steel. Appl Sci 2016; 6: 172.

66.

Berkani

Yallese

Boulanouar

, et al. Statistical analysis of AISI304 austenitic stainless steel machining using Ti (C, N)/Al₂O₃/TiN CVD coated carbide tool. Int J Ind Eng Comput 2015; 6: 539–552.