Scholarly article on topic 'An Experimental Investigation of Machinability of Inconel 718 in Electrical Discharge Machining'

An Experimental Investigation of Machinability of Inconel 718 in Electrical Discharge Machining Academic research paper on "Materials engineering"

CC BY-NC-ND
0
0
Share paper
Academic journal
Procedia Materials Science
OECD Field of science
Keywords
{"Electrical Discharge Machining" / "Inconel 718" / "Response Surface Methodology" / "Multi Objective Particle SwarmOptimization ;"}

Abstract of research paper on Materials engineering, author of scientific article — Chinmaya P. Mohanty, Siba Shankar Mahapatra, Manas Ranjan Singh

Abstract This paper proposes an experimental investigation and optimization of the various machining parameters for the electrical discharge machining (EDM) processes on Inconel 718 super alloy using a multi objective particle swarm optimization (MOPSO) algorithm. A Box-Behnkin design of response surface methodology has been used to collect data for the study. The machining performances of the process are evaluated in terms of material removal rate (MRR) and surface quality which are functions of process variables such as open circuit voltage, discharge current, pulse-on-time, duty factor, flushing pressure and tool material. Mathematical model is developed relating responses with process variables. Finally, a MOPSO algorithm has been proposed for the multi objective optimization of the responses.

Academic research paper on topic "An Experimental Investigation of Machinability of Inconel 718 in Electrical Discharge Machining"

CrossMark

Available online at www.sciencedirect.com

ScienceDirect

Procedia Materials Science 6 (2014) 605 - 611

3rd International Conference on Materials Processing and Characterisation (ICMPC 2014)

An Experimental Investigation of Machinability of Inconel 718 in

Electrical Discharge Machining

Chinmaya P Mohanty*, Siba Shankar Mahapatra, Manas Ranjan Singh

Departement of mechanical Engineering, National Institute of Technology,Rourkela, Odisha,India, 769008

Abstract

This paper proposes an experimental investigation and optimization of the various machining parameters for the electrical discharge machining (EDM) processes on Inconel 718 super alloy using a multi objective particle swarm optimization (MOPSO) algorithm. A Box-Behnkin design of response surface methodology has been used to collect data for the study. The machining performances of the process are evaluated in terms of material removal rate (MRR) and surface quality which are functions of process variables such as open circuit voltage, discharge current, pulse-on-time, duty factor, flushing pressure and tool material. Mathematical model is developed relating responses with process variables. Finally, a MOPSO algorithm has been proposed for the multi objective optimization of the responses.

© 2014ElsevierLtd.Thisis anopen accessarticleunder the CC BY-NC-ND license (http://creativecommons.Org/licenses/by-nc-nd/3.0/).

Selection and peer review under responsibility of the Gokaraju Rangaraju Institute of Engineering and Technology (GRIET) Keywords: Electrical Discharge Machining;Inconel 718;Response Surface Methodology;Multi Objective Particle SwarmOptimization ;

1. Introduction

The non-conventional machining processes are more capable than conventional machining process owing to ease of machining of hard materials with complex shapes in the shortest span of time. Now-a-days, electrical discharge machining (EDM) is extensively used for machining of toughened and high strength to weight ratio conductive materials which are difficult enough to be machined by conventional machining processes. The process has many applications in manufacturing of dies and moulds in manufacturing industries and components in aerospace and automotive industries. Lee and Li (2001)have conducted an experimental study in which the effectiveness of the EDM process is evaluated in terms material removal rate (MRR), relative wear ratio (RWR) and surface roughness of tungsten carbide which are functions of process variables such as electrode material, polarity, discharge current,

* Corresponding author. Tel.: +919438480248; fax: +91-661-2462512. E-mail address: chinmaymohantymech@gmail.com

2211-8128 © 2014 Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

Selection and peer review under responsibility of the Gokaraju Rangaraju Institute of Engineering and Technology (GRIET) doi:10.1016/j.mspro.2014.07.075

open circuit voltage, pulse duration, pulse interval and flushing pressure. Habib (2009) has analyzed the effect of machining parameters such as current, gap voltage and pulse-on-time on MRR and TWR in EDM using response surface methodology where metal matrix composite Al/SiCp is machined with copper electrodes. Chattopadhyay et al. (2009) have used Taguchi's design of experiment (DOE) method to conduct experiment on rotary EDM using EN8 steel and copper as work piece-tool pair and proposed empirical relations between process responses and process variables such as peak current, pulse-on-time and rotational speed of tool electrode Dewangan and Biswas (2013) adopted for Taguchi experimental design for optimization of multiple responses, i.e., material removal rate (MRR) and tool wear rate (TWR) of electrical discharge machining (EDM) using AISI P20 tool steel as the work material and copper electrode. Das et al. (2003) have suggested an EDM simulation model using finite element for calculation of deformation, microstructure and residual stresses. Joshi and Pande (2009) have suggested a numeral model for EDM for precise and accurate prediction of process responses viz. material removal rate (MRR) and tool

wear rate (TWR) using finite element method (FEM)._

Nomenclature

AWw weight of material removed from work piece

T machining time

t duty factor in %

Ton pulse-on-time (^s)

V open circuit Voltage in Volt

Ipdischarge current (Amp)

Fp flushing pressure (bar)

Greek Symbol

pw density of work piece

Literature review reveals, though number of attempts have been made until now to enhance the accuracy, utility and productivity of the process, combination of response surface methodology (RSM) and multi objective particle swarm optimization(MOPSO) approach for obtaining optimal process variables for EDM on Inconel 718 alloy has not been attempted yet. It also shows only a few comparative studies have been reported until now to analyze the process responses with different tool material viz. brass, copper and graphite. Inconel 718, a super alloy of nickel and chromium finds extensive usage in aerospace and other related industries. The alloy finds wide range applications in manufacturing of components for liquid fuled rockets, rings and casings. The age-hardenable alloy is used in various formed sheet metal parts for aircraft, land-based gas turbine engines and cryogenic tank. It is also used in manufacturing of fasteners and instrumentation parts. To address this issue, the present research work proposes an experimental investigation on machinability of Inconel 718 alloy in EDM process in which the performance characteristics are measured in terms of material removal rate (MRR) and surface roughness (Ra) which are functions of process variables viz. open circuit voltage, current, pulse duration, duty factor, flushing pressure and electrode material. Analysis of variance (ANOVA) was conducted to identify the important process variables for the process. Finally, a multi-objective particle swarm optimization algorithm (MOPSO) has been proposed for the optimization of both the responses

2. Experimental strategy and material

The experimental architecture is planned as per response surface methodology. DOE is basically a scientific approach to successfully plan and perform experiments using statistics and is widely used to improve the quality of a products or processes with less experimental runs. Such approaches enable the user to define and study the effect of every single condition possible in an experiment where numerous factors are involved. RSM quantifies the relationship between the controllable input parameters and the obtained responses. The objective is to find a suitable approximation for the true functional relationship between independent variables and the response. Generally, a second-order model as given in Eq. 1.is employed in response surface methodology.

y = P0 +^PiXi +^PiiX2 +SSPijXiXj (1)

i=1 i=1 kj

where, y is the corresponding response for input variables X;'s, X; and X;Xj are the square and interaction terms of parameters respectively. p0, Pi, Pu and Py are the unknown regression coefficients and s is the error. Experiments are carried out in a die sinking CNC EDM machine (ECOWIN PS 50ZNC) with servo-head (constant gap) has been shown in Figure 1. Paraffin oil (specific gravity= 0.820) was used as dielectric fluid. Positive polarity for electrode and side flushing was used to conduct the experiments. The composition of Inconel 718 Ni+Co=(50-55)%, Cr=(17-21)%,Fe=(BALANCE), Nb+Ta=(4.75- 5.5)%,Mo=(2.8-3.3)%,Ti=(0.65-1.15)%,Al=(0.2-0.8)%. Some of the other properties are density=8.19 Kg/m3, melting point=1609 K, thermal conductivity=14.5W/m.K, Coefficient of thermal expansion=13.0 ^m/m°C at temperature (20-100 °C), Poisson's Ratio=0.27-0.3. Owing to sparks, a large amount of heat has to be dealt with EDM process. The tool should be of a good conductive material with high melting point to resist and dissipate the heat. Hence, commercially available copper, brass and graphite are considered as the electrode material in cylindrical shape of 13.5mm diameter. The EDM process is performed on Inconel 718 alloy having 8mm thickness and 10X11.5 mm2 rectangular work piece. The experiment is conducted as per Box-Behnken RSM design and initial-final weight of work piece and tool is noted down after each observation. Box-Behnken design has been preferred for the analysis because it performs non sequential experiments; it is having fewer design points. It is helpful in safe operating zone for the process as these designs do not have axial points. On the other hand, central composite designs have axial point outside the cube which may not be in the region of interest or may be impossible to run as they are beyond safe operating zone. There are 54 experimental runs to be performed in Box-Behnken RSM design with three levels of six factors and six center points. Each experiment is run for 30 minutes and table 1 shows the coding of the process variables. The layout of experimental runs with obtained responses is shown in table 2. Figure 2 shows the wok material Inconel 718 after machining.

ss 6 •••••••

I O <*"> <n O O 'O O o OnQ

ioèèô^

Fig. 1.Die sinking EDM machine (ECOWIN PS 50ZNC) The material removal rate (MRR) is calculated using the following equation 1000 x AWw

Fig. 2.Work material Inconel 718 after machining

Surface quality is measured by a portable surface roughness tester (Surftest SJ 210, Mitutoyo). Roughness measurements, in the transverse direction, on the work material are repeated five times and average of five readings of surface roughness values are noted down.For smooth experimental runs the process parameters are coded using the following equation

Coded Value (Z) =

X -X .

. -2--.... . (3)

where, Z is coded value (-1, 0, 1), X max and X min is maximum and minimum value of actual parameters and X is the actual value of corresponding parameter.

Table 1. Process parameters and their codes.

Process Parameters

Open circuit Voltage (V) in Volt Current( Ip) in Amp Pulse-on time(Ton) in ^s Duty Factor (t) in % Flushing Pressure (Fp) in bar Tool

Symbols Code

-1 0 1

A 70 80 90

B 3 5 7

C 100 200 300

D 80 85 90

E 0.2 0.3 0.4

F Brass Copper Graphite

Table 2.The box behnken design experimental strategy along with obtained responses

Sl. No. A B C D E F MRR (mm3/min) Surface Roughness (^m)

1 0 0 0 12.21 8.15

2 1 0 0 0 3.1 5.1

3 1 0 0 0 40.65 24.2

4 1 1 0 0 0 25.2 19.1

5 0 1 0 0 13.39 9.75

6 1 0 1 0 0 2.5 5.15

7 1 0 1 0 0 44.95 22.1

8 1 1 0 1 0 0 25.25 15.5

9 0 0 0 9.82 10

10 0 1 0 0 16.97 25.1

11 0 1 0 0 24.92 10.1

12 0 1 1 0 0 48.25 20.9

13 0 0 1 0 6.1 6.1

14 0 1 0 1 0 22.9 16.2

15 0 1 0 1 0 20.9 22.5

16 0 1 1 0 1 0 45.35 26.5

17 0 0 0 8.7 12.1

18 0 0 1 0 14.49 14.9

19 0 0 1 0 12.5 10.2

20 0 0 1 1 0 14.36 18.2

21 0 0 0 1 23.4 11.2

22 0 0 1 0 1 40.2 19.5

23 0 0 1 0 1 30.1 12.5

24 0 0 1 1 0 1 40.3 20.1

25 0 0 0 34.18 16.3

26 1 0 0 0 15.7 12.7

27 0 0 1 0 32.25 16.1

28 1 0 0 1 0 16.8 14.1

29 0 0 1 0 34.97 20.1

30 1 0 0 1 0 15.72 14.3

31 0 0 1 1 0 35.03 21.2

32 1 0 0 1 1 0 16.1 14.4

33 0 0 0 2.03 7.8

34 0 1 0 0 18.43 15.5

35 0 0 0 1 3.56 7.9

36 0 1 0 0 1 18.72 16.1

37 0 0 0 1 18.3 7.25

38 0 1 0 0 1 46.1 16.5

39 0 0 0 1 1 16.2 18.1

40 0 1 0 0 1 1 45 17.1

41 0 0 0 10.95 12.2

42 1 0 0 0 2.35 8.5

43 0 1 0 0 18.12 20.95

44 1 0 1 0 0 9.8 18.2

45 0 0 0 1 20.3 15.1

46 1 0 0 0 1 10.2 10.1

47 0 1 0 0 1 42.72 18.9

48 1 0 1 0 0 1 25.3 15.9

49 0 0 0 0 0 0 31.5 16.5

50 0 0 0 0 0 0 28.8 19.5

51 0 0 0 0 0 0 33.9 16.1

52 0 0 0 0 0 0 29.1 20.1

53 0 0 0 0 0 0 33.1 15.4

54 0 0 0 0 0 0 27.8 19.2

3. Results and discussion

The experimental observations are carried out as per the response surface methodology to analyze the effect of various important process parameters on the responses. Table 3 shows the analysis of variance (ANOVA) table for MRR after elimination of the insignificant process variables. It shows that the model is significant and voltage, current, pulse-on-time and tool are the significant process variables. Figure 3 shows the surface plot of MRR with current and tool. It shows that MRR value increases monotonically with increase in current with graphite and copper electrodes but increases slowly with the use of brass electrode. Material removal is higher, while machining with graphite electrode followed by copper and brass respectively. Similarly, from surface plot of MRR with voltage and pulse-on-time, it is observed that MRR increases with increase of voltage, reaches a maximum value and then decreases for low level of pulse-on-time. Similar trends have been also observed at higher values of pulse-on-time. Figure 4 shows the surface plot of surface roughness with current and tool material. It shows that surface quality deteriorates heavily with increases in current and with the use of graphite and copper electrodes.Graphite electrode exhibits the poorest performance with regard to the surface finish. Brass electrode at smaller values of discharge current produces finest surface quality. Surface quality deteriorates heavily with increase in pulse-on-time. Hence, smaller value of discharge current and pulse duration can be suggested subject to smaller material removal for finishing operation. The process model of the two responses obtained through regression analysis is given as below. MRR=30.91-7.15*A+11.03*B+7.10*C+0.63*D-0.13*E+9.34*F-1.89*A*B-0.88*A*C-1.33*A*F+2.98*B*C+0.66*B*E+3.13*B*F-0.32*C*D-1.14*C*E+2.64*C*F-0.63*E*F-5.72*A2-4.28*B2-2.23* C2-5.59*F2 (4)

SR=17.80-2.25*A+4.87*B+3.14*C+0.069*D+1.17*E+0.74*F-0.51*A*B-0.87*A*E-0.44*A*F-1.30*B*C-0.92*B*D-1.35*B*E-0.96*B*F+0.56*C*D+3.85* C* E-0.36*C*F-1.11*A2-2.06*B2+0.97*C2-1.00*D2+0.46*E2-2.93*F2(5)

The empirical relation between the process parameters and process responses established from the RSM analysis is used as objective function for solving the multi-objective particle swarm optimization (MOPSO) problem. The optimization model was run on MATLAB 13 platform in a Pentium IV desktop.

Fi g. 3 Surface plot of MRR with current and tool

Fig. 4 Surface plot of surface roughness with current and tool Table 3. ANOVA table for MRR

Source Sum of squares df Mean Square F- Value p-value Prob > F

Model 8553.83 20 427.69 33.38 < 0.0001 significant

A-Voltage 1228.37 1 1228.37 95.86 < 0.0001

B-Current 2920.30 1 2920.30 227.91 < 0.0001

C-Pulse-on-time 1210.12 1 1210.12 94.44 < 0.0001

D-Duty factor 9.39 1 9.39 0.73 0.3982

E-Flushing Pr. 0.43 1 0.43 0.033 0.8563

F-Tool 2092.72 1 2092.72 163.32 < 0.0001

AB 28.69 1 28.69 2.24 0.1441

AC 6.20 1 6.20 0.48 0.4917

AF 14.04 1 14.04 1.10 0.3027

BC 70.98 1 70.98 5.54 0.0247

BE 6.93 1 6.93 0.54 0.4673

BF 78.38 1 78.38 6.12 0.0187

CD 13.86 1 13.86 1.08 0.3059

CE 10.42 1 10.42 0.81 0.3737

CF 111.57 1 111.57 8.71 0.0058

EF 3.15 1 3.15 0.25 0.6233

AA2 381.10 1 381.10 29.74 < 0.0001

BA2 213.32 1 213.32 16.65 0.0003

CA2 52.88 1 52.88 4.13 0.0503

FA2 332.74 1 332.74 25.97 < 0.0001

Residual 422.85 33 12.81

Lack of Fit 378.70 28 13.52 1.53 0.3383 not significant

Pure Error 44.15 5 8.83 33.38

Cor Total 8976.67 53 427.69 95.86

Simulation study is carried out in MATLAB to demonstrate the potentiality of MOPSO algorithm. The initial population chosen for all the three algorithms is 100. The simulation parameters employed for MOPSO are as follows: the size of archive is 100, the inertia weight is 0.4 and both the cognitive parameters (c1 and c2) are taken as 2. This led to the development of Pareto-front as shown in Figure 5 generating optimal solution for the responses. A sample set of the optimal solution has been given in Table 4.

Fig. 5 Pareto-optimal front for objectives MRR and Surface roughness Table 1. Pareto Optimal sample solution set and corresponding variable settings

V in Volt Ip in (Amp) Ton in (^s) T in (%) Fp in bar Tool MRR(mm3/min) Surface Roughness(^m)

71.00955 6.228442 100 90 0.2 Graphite 57.949 23.17496

71.00955 6.032309 100 90 0.2 Graphite 56.6265 23.04081

70.07897 5.839783 100 90 0.2 Graphite 56.1591 22.79964

70.07897 5.755703 100 90 0.2 Graphite 55.654 22.70759

70.37349 5.611023 100 90 0.2 Graphite 54.5809 22.54878

70.74644 5.585821 100 90 0.2 Graphite 54.141 22.54035

70.1954 5.461992 100 90 0.2 Graphite 53.9165 22.32543

4. Conclusions

The proposed model shows the interactive and complex effects of various important process variables viz. open circuit voltage(V), discharge current(Ip), pulse-on-time(Ton) and tool material on responses justified through experimentation and analysis. The analysis of experimental observations revealed that tool material and discharge current and pulse-on-time are the important parameters in the machinability of Inconel 718 super alloy for both the responses. This research work offers an effective guideline to select optimum parameter settings for achieving the desired MRR, surface roughness during EDM die sinking of Inconel 718 alloy to the experimenter and practitioners. A multi objective particle swarm (MOPSO) algorithm has been proposed for optimization of the responses. The proposed model can be used for selecting ideal process states for achieving improved machining condition for Inconel 718 alloy while machining in EDM process.

References

Chattopadhyay, K. D., Verma, S., Satsangi, P. S. and Sharma, P. C., 2009. Development of empirical model for different process parameter during rotary electrical discharge machining of copper - steel (EN-8) system. Journal of Materials Processing and. Technology, 209, 14541465.

Das, S. Klotz, M. and Klocke, F. 2003. EDM simulation: finite element-based calculation of deformation, microstructure and residual

stresses,Journal of Materials Processing Technology, 142(2), 434-451. Dewangan, S. and Biswas, C.K. 2013. Optimisation of machining parameters using grey relation analysis for EDM with impulse flushing, International Journal of Mechatronics and Manufacturing Systems, 6(2), 144-158.

Habib, S.S., 2009. Study of the parameters in electrical discharge machining through response surface methodology approach, Applied

Mathematical Modeling, 33(12), 4397-4407. Joshi, S.N. and Pande, S.S., 2009. Development of an intelligent process model for EDM, International Journal of Advanced Manufacturing Technology, 45(3-4), 300-317.

Lee,S.H and Li, X. P.,2001. Study of the effect of machining parameters on the machining characteristics in electrical discharge machining of tungsten carbide,Journal of Material Process and Technology, 115(3), 344-358.