Multi-Optimization of Performance Parameters in Turn-Boring for Difficult-to-Cut Ti-6 Al-4 V

The present study propose an innovative turn-boring operation method and focuses on finding optimal turn-boring process parameters for Ti-6Al-4V by considering multiple performance characteristics using Taguchi orthogonal array with the grey relational analysis, the effect of machining variables such as, feed rate, depth of cut and cutting speed are optimized with considerations of multiple performance characteristics namely surface roughness, roundness error, material removal rate and power consumption the optimal values were found out from the Grey relational grade. The result of the Analysis of Variances (ANOVA) is shown that the most significant factor is cutting speed, followed by feed rate, radial depth of cut. Finally, confirmation tests were carried out to make a comparison between the experimental results. Experimental results have shown that machining performance in the turn-boring process can be improved effectively through this approach.


Introduction
Fine boring is one of the important machining aerospace and automotive parts that have to be efficient and high accuracy.It is an important task to select cutting parameters for achieving superior cutting performance.The evolution of boring machining operation properties using different parameters is a complex phenomenon.
There are many factors such as cutting speed, depth of cut, feed rate, insert material etc affecting the performance of boring machining operation resulting in different surface quality and accuracy.Venkatarao et al. [1] studied the effect of various input cutting parameters such as cutting speed, feed rate, and tool nose radius on tool life in boring for AISI 1040steel by analyzing surface roughness, amplitude of work piece vibration and volume of metal removed.Venkatarao et al. [2] using an artificial neural network to predict the cutting tool wear, surface roughness and vibration of the work piece in boring of AISI 316 steel.Chun et al. [3] using the response surface methodology to study the effect of the overhang, feed rate, and the depth of cut on machining errors in boring for AISI4140 steel.They found the depth of cut is the most significant parameters affecting the dimensional accuracy.Ti-6Al-4V is widely used in aerospace, automotive and biomedical industries because of its excellent high strength, good corrosion resistance, low thermal conductivity and Young's modulus However, difficult-to-machine and poor machinability [4][5], excessive tool wear and burr formation has always been a challenge in boring.
With the rapid development of technology in aerospace, automobile and biomedical industry, conventional machining operations can be not enough to satisfy machining quality, productivity and cost demands.To improve processing quality and varieties of metal compound machining methods has been developed.At this point, an innovative process such as turn-milling is an effective way of machine difficult-to-cut materials as a kind of compound machining.Schulz [5] proposed two different turn-milling operations, namely coaxial and orthogonal, and carried out experimental trials to identify cutting conditions (i.e.feed, speeds, depth of cut) that result in a good surface.
He has further concluded that short chips are formed by intermittent cutting process.[6] carried out a series of orthogonal turn-milling experiments and states that surface finish quality obtained by orthogonal turn-milling will be about 10 times better than those obtained by conventional turning.

Choudhury and Mangrulkar
Base on turn-milling method idea, we also proposed a new innovative turn-boring technology of both the boring tool cutter and workpiece rotation simultaneously.Due to the direction of rotation of the tool and the workpiece is reversed, resultant from intermittent cuts and chip breaking.It keeps low thermal stresses, lowers cutting forces and has useful material removal rates.It also can offer increased productivity for difficult-to-machine materials.So far, this compound processing machining method has not yet seen any study reports.
Even though a very few research works have been carried out to study the influence of boring parameters on different quality.However, CNC turn-boring have not seen the report, it is very necessary to establish optimal parametric combination with the intention of obtaining improved machined surface, accuracy and material removal rates.This study focused on the optimal turn-boring parameters considering the multiple qualities characteristics such as surface roughness, roundness and material remove rate using Taguchi based the Grey Relational Analysis (GRA).The grey relation analysis theory proposed by Deng [7] was used for solving the complex interrelationship among the multi-objective in various fields of manufacturing.It's is an effective approach to solve the multi-objective optimization.Recently, some researchers have effectively used this method for solving the intricate interrelationships between the multiple objectives in engineering manufacture with multiple responses.Kuram and Ozcelik [8] employed the Taguchi method and the Grey Relational Analysis to multi-optimize with multiple response outputs in the micro-milling.They studied the effects of spindle speed, feed per tooth and depth of cut on tool wear, force, and surface roughness.Yang et al. [9] applied the Taguchi method and the GRA to optimize the milling parameters such as the cutting speed, the feed rate, and the depth of cut for simultaneous optimization of the energy, production rate and cutting quality.Kant and Sangwan [10] using grey relational analysis to find the optimum values of machining parameters to achieve the minimum power consumption and surface roughness.Hwang et al. [11] investigated and optimized the high speed end milling of SKD61 Tool Steel using Taguchi methods with grey relational analysis.Recently, this method attracts more and more widely applied on mechanical machining and tribological engineering, including drilling [12][13][14], turning [15][16], milling [17][18], wire electrical discharge machining [19][20], and tribology [21].
As mentioned above, The GRA is an effective approach to solve the multi-objective optimization.Therefore, this study applied a Taguchi L27 orthogonal array to plan the experiments.The experimental design was organized for five parameters including cutting fluid concentration, cutting fluid temperature, feed rate, depth of cut and relative cutting speed of turn-boring with three levels for each factor.This study presented the multi-response optimization of turn-boring parameters to simultaneously minimize the surface roughness, roundness and maximize material remove rate using Taguchi based GRA in turn-boring of Ti-6Al-4V.

Experimental procedure
Turn-boring experiments were conducted on YAMAZAKI MAZAK INTEGREX 300-IV multi-tasking machine.The test materials were chosen as 15-5PH stainless steel.Its chemical composition and mechanical properties are shown in Tables 1 and 2, respectively.The schematic diagram of the experimental set-up is illustrated in Fig. 1.
Dimension of the workpiece was a length of 30mm outer diameter of 60mm and inner diameter of 32mm is shown in Fig. 2.The experimental conditions used to refer to a turn-boring operation are summarized in Tables 3. The cutting fluid properties are shown in table 4. Turn-boring parameters and their levels as showed in Table 5.
Taguchi's L27 orthogonal arrays are given in Table 6.The machining parameters are concentration of cutting fluid , temperature of cutting fluid , feed rate, depth of cut and cutting speed and the output response is surface roughness (Ra), roundness error (Er) and material removal rate (MRR), the surface roughness Ra is measured with surface roughness tester (MITUTOYO, MODEL SJ210), the roundness is measured with Coordinate Measuring Machine (DEA GLOBAL Silver SF 12.15.10).The material removal rate is calculated by using Eq. ( 1), as showed in Fig 3.
where n = 1000 / , Vc is relative Cutting Speed in m/min ,MRR is material removal rate in mm 3 /min.ap is Axial length of cut in mm.D2 is Finished Diameter in mm.D1 is Initial diameter in mm.n is machine speed in revolutions/minute.fr is machine feed rate in mm/rev.The power consumption is calculated by using Eq.(2) Where

Multi-response optimization with Grey relational analysis
Grey relational analysis is a statistical method to analyze the complex multi-response systems.Based on experimental data, the Grey relational analysis is utilized to convert the multi-response optimization into the single objective optimization of the grey relational grade.The aim of this study was to identify the optimal combination of turn-boring parameters that simultaneously minimize surface roughness, roundness and maximize the material removal rate in turn-boring of Ti-6Al-4V.Steps of grey relational analysis are as follows:

Grey relational generation
The first step of grey relational analysis is to normalize the experimental data depending on the type of performance response.Ra, Er, Pc and MRR are to normalize in the range of zero to one.This is called grey relational normalization.In the present study, as surface roughness, roundness and power consumption had to be minimized, the smaller-the-better model aims to obtain the minimum quality characteristics.MRR had to be maximized.
For the Lower-the-better criterion, the normalize data can be expressed as * ( ) = max ( ( )) − ( ( )) max (( ( )) − min ( ( )) ( For the Larger-the-better criterion, the normalize data can be expressed as * ( ) = ( ) − min ( ( )) max (( ( )) − min ( ( )) where * ( )is the value after grey relational generation (normalized value), and max ( ) and ( )are the largest and smallest values of ( ) for the kth response, respectively, k being 1 for surface roughness and 2 for roundness and 3 for material removal rate.The processed data after grey relational generation was given in Table 7.The normalized values are ranged between zero and one.Larger normalized results mean to the better performance and the best normalized result should be equal to 1.

grey relational coefficient
Grey relational coefficients denote the relationship between the ideal and the actual experimental results.Grey relational coefficient ( ) can be calculated as the following: Here, Δ ( ) = ‖ * ( ) − * ( )‖ , * ( ) is reference sequence and * ( ) ]) and is used to adjust the difference of the relational coefficient.Generally, the distinguishing coefficient is assumed as 0.5 to fit the practical requirements and the grey relational coefficient calculated using Eq. ( 4) was given in Table 8.

Calculate Grey Relational Grade (GRG)
The grey relational grade with Eq. ( 5) can be computed by averaging the grey relational coefficients to evaluate the multiple response as a single index as Here, n is the number of performance characteristics.The highest grey relational grade corresponds to the experimental value closest to the ideal normalized value.
Thus, higher grey relational grade shows that the corresponding parameter combination is closer to the optimal.According to Eq. ( 5), the GRG between the process parameters and the performance characteristics is calculated.Grey relational grade as shown in Table 8.

Optimal grey relational grade
Since higher GRG was desirable, the larger-the-better S/N quality characteristic was used to obtain the optimal combination for multi-response optimization approach for higher GRG is defined as follows: In Eq. ( 6), yi is the ith measured experimental results in a run/row and n explains the number of measurements in each test trial/row.S/N ratios of multiple quality characteristics were calculated by using Eq. ( 6) and were listed in Table 9.The level of a parameter with the highest S/N ratio gives the optimal level.As showed in Fig. 4.
The optimal turn-boring parameter setting for the multiple performance characteristic as A(level 1) -B(level 3) -C(level 1).Thus, the best parameters combination were feed rate of 0.08 mm/rev, radial depth of cut of 0.150 mm and cutting speed of 12 m/min, is the optimal process parameter combination.In addition, the difference between the maximum and minimum values for the GRG in Table 9 is 1.681 for the feed rate, 1.042 for radial depth of cut and 2.860 for the cutting speed.It reflects the impact level of the three process parameters on the performance characteristics.And the cutting speed has the most remarkable influence on the performance characteristics.

Analysis of variance (ANOVA)
The multi objectives are converted into a single object with the help of grey relational analysis.The significant contribution of each input parameter on the responses in turn-boring parameters is studied by using analysis of variance.The results of the ANOVA are given in Table 10.Feed rate, depth of cut and cutting speed influenced the multiple performance characteristics by 17.93%, 7.52% and 52.35%, respectively.
This indicates that the cutting speed is the most contributing factor.

Prediction of grey relation grade
After finding the suitable optimal parameters, it is necessary to predict the grey relational grade theoretically.The estimated grey relational grade of the optimal level of the design parameters combination can be calculated as: where is the GRG to predict the optimal machining parameters, is the total average GRG, is the average GRG at the optimal level, and k is the number of main design parameters that significantly affect the multiple performance characteristics.Table 11 shows the comparison results of the initial turn-boring and optimal turn-boring parameters.It was determined that there is a good agreement between estimated value and experimental value.It was found out that the improvement of grey relational grade from initial factor combination (A1-B2-C3) to the optimal factor combination (A1-B3-C1) was 0.301 and the percentage improvement in Grey relational grade with the multiple responses was 64.2%.

Experimental validation
In the last step, the obtained results are conducted to verify the optimized solution.
Table 11 compares the confirmation test results using the initial and optimal level combination of process parameters.It is obvious that Ra is decrease from 0.40µm to 0.23 µm, Er is decrease from 4.56µm to 3.21 µm, Pc f is decrease from 0.558347kw to 0.209670kw and MRR is increase from 0.001467(mm 3 /min) to 0.005869(mm 3 /min).
It indicates that the GRA algorithm can be used to improve the performance characteristics.

Conclusions
Ti-6Al-4V is a difficult to machine and costly material.Hence, optimization of machining parameters is necessary to obtain good quality surfaces.In this study, an innovative precision turn-boring has successfully demonstrated the application of the Taguchi-based grey relational analysis for multi-objective optimization of process parameters in turn-boring Ti-6Al-4V for achieving for simultaneous minimum surface roughness, roundness error, power consumption and maximum MRR.The main findings can be summarized as follows: 1.The optimum parameters for multiple optimization machine setting through grey relational analysis were feed rate of 0.08mm/rev, radial depth of cut of 0.150mm and cutting speed of 12 m/min.The percentage improvement in GRG = 64.2%.
Pc is the turn-boring power consumption in kw, fr is Feed Rate in mm/rev, Vc is relative Cutting Speed in m/min, dc is Radial depth of cut in mm and kc is Cutting impedance in N/mm 2 :1450

Table 5 .
Turn-boring parameters and their levels.

Table 6 .
Experiment design and real value.

Table 7 .
2. It was observed through ANOVA that the cutting speed has a dominant effect of almost 52.35% in contribution ratio, while feed rate has 19.73% and radial depth of cut has 7.52% influence on the surface roughness, roundness error and power consumption and material remove rate is considered simultaneously.3.The percentage improvement in GRG with the multiple responses was 64.20%.It is clearly shown that the performance indicators (surface roughness, roundness error, power consumption and material remove rate) are significantly improved in turn-boring for difficult-to-cut Ti-6Al-4V using the Taguchi-based grey rational analysis.Normalized sequence after grey relational generation.

Table 9 .
Response table for S/N ratio of average grey relational grade.Fig. 4 Main effects plot for grey relational grade

Table 10 .
ANOVA table for grey relational grade

Table 11 .
Comparison of the initial and optimized results.