A comparative study on optimization of machining parameters by turning aerospace materials according to Taguchi method

The effects of cutting tool coating material and cutting speed on cutting forces and surface roughness were investigated by Taguchi experimental design. Main cutting force, Fz is considered as a criterion. The effects of machining parameters were investigated using Taguchi L18 orthogonal array. Optimal cutting conditions were determined using the signal-to-noise (S/N) ratio which is calculated for average surface roughness and cutting force according to the ‘‘the smaller is better’’ approach. Using results of analysis of variance (ANOVA) and signal-to-noise (S/N) ratio, effects of parameters on both average surface roughness and cutting forces were statistically investigated. It was observed that feed rate and cutting speed had higher effect on cutting force in Hastelloy X, while the feed rate and cutting tool had higher effect on cutting force in Inconel 625. According to average surface roughness the cutting tool and feed rate had higher effect in Hastelloy X and Inconel 625.


Introduction
Advanced materials, such as nickel-base and titanium alloys as well as composites are generally used at 650°C or higher temperatures at which high stresses occur and surface integrate required. These materials are widely used in industrial gas turbines, space vehicles, rocket engines, nuclear reactors, submarines, stream production plants petrochemical devices, hot tools and glass industries [1]. Inconel 625 has been used in aqueous corrosive environments due to its excellent overall corrosion resistance [2]. Inconel 625 (Alloy 625) is a nickel-based super alloy strengthened mainly by the solidsolution hardening of the refractory metals, niobium and molybdenum, in a nickel-chromium matrix. Alloy 625 was originally developed as a solid-solution strengthened material. It was determined that the alloy is hardenable [3][4][5][6]. Inconel 625 exhibits precipitation hardening mainly due to the precipitation of fine metastable phase (Ni3Nb) after annealing over a long period in the temperature range 550-850°C [4,5]. Moreover, various forms of carbides (MC, M6C and M23C6) can also precipitate depending upon the time and temperature of aging. Alloy 625 has extensive use in many industries for diverse applications over a wide temperature range from cryogenic conditions to ultra hot environments over 1000°C [6][7][8][9]. Hastelloy X and Inconel 625 is a nickel chromium-iron molybdenum alloy is developed for high temperature applications and it is derived from the strengthening particles, Ni2 (Mo, Cr), which formed after the two-step age-hardening heat treatment process. With face-centered cubic (FCC), Ni-Cr-Mo-W alloys, named as Hastelloy used for marine engineering, chemical and hydrocarbon processing equipment, valves, pumps, sensors and heat exchangers. Nickel-based super alloys have heat resistance, excellent mechanical properties, corrosion resistance and ability to operate in high temperature, attracting in nuclear industries [10,11]. Nickel-based alloys and super alloys are very difficult to process [12][13][14][15]. A nickel-based super alloy has generally chemical content 38-76% nickel (Ni), more than 27% chromium (Cr) and 20% cobalt (Co) [16]. Such materials having high corrosion resistance and high strength at high temperatures are used [12,13,[17][18][19][20][21] [16,22]. These alloys have excellent mechanical properties, workability and corrosion resistance in aviation and extensively in the chemical industry heaters, condensers, evaporator tubes, pipes mirrors. However, low thermal conductivity and high cutting strength is still considered as challenging [16,23].

Materials and method 2.1 Experiment specimens
Specimens of Hastelloy X and Inconel 625 which has an industrial usage, are prepared as the dimension of diameter Ø 25 · 40 mm then used for the experiments. The chemical composition and mechanical properties of specimens are given in Tables 1 and 2. These materials are hard to machine which make them suitable for high temperature applications.

Machine tool and measuring instrument of cutting forces
In the experimental study machining tests are carried out on JOHNFORD T35 industrial type CNC lathe maximum power of which is 10 kW and has revolution number between 50 and 3500 rev/min ( Figure 1). During dry cutting process, Kistler brand 9257 B-type three-component piezoelectric dynamometer under tool holder with the appropriate load amplifier is used for measuring three orthogonal cutting forces (F x , F y , F z ). This allows direct and continuous recording and simultaneous graphical visualization of the three cutting forces (Figures 2 and 3).

Cutting parameters, cutting tool and tool holder
The cutting speeds 65, 80, and 100 mm/rev were chosen by taking into consideration ISO 3685 standard as recommended by manufacturing companies. The depth of cut 1, 5 mm feed rate 0.10-0.15 mm/rev. During cutting process, the machining tests were conducted with three different cemented carbide tools namely Physical Vapor Deposition (PVD) coated with TiN/TiCN/TiN; Chemical Vapor Deposition (CVD) coated with T _ IN+AL2O3-T _ ICN+T _ IN; and WC/CO. The test specimens were chosen Ø 25 · 40 mm Properties of cutting tools and level of independent variables are given in Tables 3  and 4. Surtrasonic 3-P measuring equipment is used for the measurement of surface roughness. Measurement processes are carried out with three replications. For surface roughness on work-piece during machining, cut-off and sampling length are considered as 0.8 and 2.5 mm, respectively. Ambient temperature is 20 ± 1°C. The following details tool geometry CNMG inserts when mounted on the tool holder: (a) CNMG shape; (b) axial rake angle: 6°; (c) end relief angle: 5°; and (d) sharp cutting edge. The insert type CNMG 120404 with 75°approaching is mounted on PCLNR 2525 M 12 type tool holder. The levels for the determination of parameters estimated and actual test results S/N ratio and cutting values are given in Table 5. ANOVA results for the main cutting force (F z ) and surface roughness S/N ratio in Inconel 625 and Hastelloy X are given in Tables   Parameter in the determination of the maximum cutting force values for each level of the small S/N ratio determined and created new verification experiment was conducted according to the test combination. Tolerances specified for the product and quality in the design stage towards achieving the goal around the nominal value of each selected parameter to determine tolerance values. Product losses in the case where a different result from the target value by determining deviations calculated. Taguchi loss function, the expected target value and the deviation between the experimental values and the signal/noise (S/N) ratio is calculated by converting [24][25][26]. S/N ratio in calculating three different characteristics which are frequently used; nominal (face) value the better, smaller is better and bigger is better. In this study, the low surface roughness value, best performance will refer to the literature processed surfaces the lowest surface roughness values for the smaller the better S/N characteristic Due to the use in the analysis of at least the surface roughness and cutting forces for the smaller the better S/N characteristic is used. However, in experiments bigger the better S/N characteristic may be used [26,27].
The aim here is S/N ratio is to maximize. Thus assessment for each parameter the average S/N ratio and the largest S/N ratio with a level, is used to determine the best results. In this study, the low surface roughness and low cutting force value represents the best performance. Parameters for each level of the average S/N ratio by utilizing a graphical representation    of an optimal level for each parameter is determined. Accordingly, the parameters determined for each level of the S/N ratio is calculated using the estimated value. Estimated S/N ratio and output (surface roughness or cutting) value is used in calculating the formulas 4. The final step of the Taguchi experimental design process includes confirmation experiments [27,28]. Or this aim, the results of the experiments were compared with the predicted values with the Taguchi method and the error rates were obtained. S/N ratios g predict were predicted using the following model [27][28][29][30].
Here, g: The estimated S/N ratio. g m : Total average S/N ratio. g i : Parameter i at the level of the S/N ratio.
Here g predict is the main cutting force or Fz with regard to the S/N ratio.
Moreover, the optimum turning parameters were obtained for the performance characteristics using the Taguchi analysis. Where g m is the total mean of the S/N ratios, g i is the mean S/N ratio at the optimum level and k is the number of the main design parameters that significantly affect the performance characteristics. After predicting the S/N ratios other than experiments, the main cutting force or F z were calculated using the following equation. The final step of the Taguchi experimental design process includes confirmation experiments [28,29]. For this aim, the results of the experiments were compared with the predicted values with the Taguchi method and the error rates were obtained. S/N ratios were predicted using the following model [30]. In this research ''smallest is better'' was used since the minimum of the cutting force and surface roughness was intended. In the experiment, the S/N ratio can be calculated using the following equation: g is the number of replications and Yi is the measured characteristic.
Taguchi method, used to analyze and evaluate the numerical results for the orthogonal experimental design, the S/N ratio and ANOVA combining three tools such as the solution reaches [30][31][32][33].

Results of Taguchi analysis
Experiments conducted with two different cutting tool wear value obtained as a result of the L18 experimental design based on a total of 36 experiments were made orthogonal. L18 orthogonal design, in two levels, corresponding to 8 columns and 18 rows of cylindrical turning experiments (17 degrees of freedom) was formed. Cutting force and the surface roughness is small, as quality characteristics ''(S/N) SB, the smaller-better'' is selected [32,33]. The average surface roughness, the main cutting force data obtained in experiments and S/N ratios is given in Table 5. According to the data in Table 5, the lowest main cutting force at 100 m/min was found in Hastelloy X with KC 9240 insert as 538 N and in Inconel 625 as 483 N. In Table 5 the lowest average surface roughness was found with KC 9240 at 100 m/min in Hastelloy X as 0.755 lm and in Inconel 625 as 0.725 lm at 65 m/min. Determining the  Table 5.
Determining the minimum mean surface roughness values of the parameters for each level of the large S/N ratio determined and created new verification experiment was conducted according to the test combination. ANOVA results for the main cutting force (F z ), surface roughness and S/N ratio in Inconel 625 and Hastelloy X are provided in Tables 6-9.  Average S/N ratio Average S/N ratio

According to Fz(N)
According to Ra (μm) Figure 5. According to the level of machining parameters in Hastelloy X, cutting force (F z ), surface roughness Ra(lm) the signal-to-noise (S/N) ratio).

Results and conclusions
The experimental design described herein was used to develop a main cutting force and surface roughness prediction model roughness using analysis of Taguchi for turning Inconel 625 and Hastelloy X. Results of this experimental study can be summarized as follows: d It has seen that while feed rate (39.8%) and cutting tool (11.8%) has higher effect on cutting force in Inconel 625, the feed rate (65.99%) and cutting speed (11.14%) has higher effect on cutting force in Hastelloy X. While cutting tool (23.7%) and feed rate (16.5%) has higher effect on average surface roughness in Inconel 625, cutting tool (40.38%), and feed rate (33.15%) has higher effect on average surface roughness in Hastelloy X. d According to obtained experiments data, the lowest main cutting force has found in Hastelloy X with KC 9240 insert as 538 N and in Inconel 625 as 483 N both at 100 m/min. In the same KC 9240 insert, lowest average surface roughness has found at 100 m/min in Hastelloy X as 0.755 lm. And as 0.725 lm at 65 m/min. in Inconel 625. It was seen the effect of cutting tool on surface roughness has found higher on Hastelloy X and Inconel 625. d Taguchi orthogonal array arrangement, it has seen appropriate to analyzed the cutting force and average surface roughness defined in this article.