Open access peer-reviewed chapter

Optimization of Surface Roughness of D2 Steels in WEDM using ANN Technique

By Umesh K. Vates, N.K. Singh, B.P. Sharma and S. Sivarao

Submitted: September 13th 2018Reviewed: October 2nd 2018Published: May 31st 2019

DOI: 10.5772/intechopen.81816

Downloaded: 104

Abstract

Attempt has been made to investigate the experimental process and surface roughness (SR) optimization of cold working (high carbon and high chromium) hard die steel (D2) during wire electrical discharge machining processes (WEDM). It is very difficult to determine optimal cutting parameters for improving cutting performance has been reported. Wire electrical discharge machining process relies heavily on the operators’ technologies and experience because of their numerous and diverse range as using complicated cuts can made through difficult to machine electrically conductive components, WEDM process was developed to generate precise cutting on complicate, hard and difficult to machine materials. Tan-sigmoid and purlin transfer functional with bias based four layered back propagation artificial neural network (BPANN) approach have been used to investigate the effect of six independent parameters namely gap voltage (Vg), flush rate (Fr), Pulse on time (Ton), pulse off time (Toff), wire feed (Wf) and wire tension (Wt) over CLA value of surface roughness (Ra) along with corresponding material removal rate (MRR). A fractional factorial design of experiment of three level were employed to conduct 80 rows of experiment on (D2) steel with chrome coated copper alloy wire electrode. The predicted response, CLA values of SR and corresponding MRR were observed by the approach of BPANN from experimental (55 rows for training, 15 rows for validation and 10 for testing) data. Software instructed programme has been used individually for training, validation and testing in MATLAB 2010a to find the corresponding prediction output. Two fold cross over technique (TFCT) were used to developed distinguish (S1 and S2) models and also developed more models depending on numbers of neurons used in primary and secondary hidden layers. The model adequacy is very satisfactory as correlation coefficient (R2) is found to be 99.1% and adjusted (Radj.2) statistics is 98.5. It is found those spark time ON/OFF, wire feed rate, wire tension, gap voltage and flush rate and few of their interactions have significant effect on SR.

Keywords

  • WEDM
  • BPANN
  • SR
  • MRR
  • TFCT

1. Introduction

Wire electrical discharge machining is the metal removal process by means of repeated spark created between the wire electrode and work piece. It is considered as unique adaptation of the conventional EDM, which used an electrode to create the sparking within kerfs [11]. However, WEDM utilizes a continuously traveling chromium coated copper wire electrode ranging diameter 0.05–0.35 mm, which is capable to achieve very good sharpness of edge [4]. Very high temperature ranging 8000–10,000°C creates within the kerfs gap during machining, so that material removal may takes place by not only melting but directly vaporizations also. WEDM is used for the high precision machining to all type of electrically conductive metallic alloys, tool and die, graphite, and few ceramic and composite materials of any hardness which cannot be machined easily by conventional machining methods [1, 5].

Manufacturing processes (WEDM) has been chosen depending on the material characteristics and the type of responses required to be evaluating. The present study aimed to optimization of responses i.e. surface roughness with corresponding MRR of D2 steel by conducting 80 rows of experimental data using frictional factorial (26–2) design of experiment of five different set at three levels [3]. Four layered BPANN architecture has been used for modeling, where independent process variables are Vg, Fr Ton, Toff, Wf and Wt to get the précised and optimized values of responses Ra [6, 8, 10]. Best model S2 has been found on the basis of correlation coefficient (R2) between observed and predicted responses (SR) [12]. The response (SR) is expressed as the irregularities of material resulted from various machining operations. It is represented as ‘R a ’ symbol and used to be called center line arithmetic average roughness for the sampling length [2].

The optimum process parameters are much essential to achieve better surface finish with adequate material removal rate (MRR) or shrink of total machining time; lot of research attempts has been reported for modeling and investigation of WEDM process parameters [7], but sum of root mean square error (SRMSE) approach have been used to optimize the process parameters by taking 55 rows of training data [9].

2. Experimental setup

2.1 Selection of wire electrode and work piece

A chrome coated cylindrical pure copper wire electrode having 0.25 mm in diameter and high tensile strength were selected for conducting machining operation on 18 mm diameter of D2 steel rod to cut 5 mm thickness of disk using Electronica Maxicut, WEDM process. It is very clear that D2 is hard die steel and conducting material with high carbon and chromium content ( Table 1 ).

CSICrMoVHRCConductivity
1.50%0.30%12.00%0.80%0.90%5622 (W/mk)

Table 1.

Metallurgical component analysis: D2 steel.

The experiment has carried out on Wire Electrical Discharge Machine, model ELECTRONICA-MAXICUT, SLNO -250, (F:09:0002:01) having the facilities to hold the work piece within the place provided by the help of conductive fixture, so that they can complete the circuit between electrode and work piece. The spark is created depending upon gap voltage applied between the conductive work piece, electrode, and machining performance influence the major independent process parameter which selected for experiment as characteristics of screening test. Commercials grade of deionized water (density = 832 kg/m3) was used as dielectric fluid. 18 mm cylindrical rod of D2 steel was used as the work piece with negative polarity and the power supply has the provision to connect the 0.25 mm chromium coated pure copper tool electrode with positive polarity so that the material removal may takes place by influence of heat generated within kerfs due to applied voltage within it ( Figure 1 ).

Figure 1.

D2 steel machining using WEDM process.

The surface roughness Ra of the processed material have been measured precisely by using Surftest SJ-210 tester having center line average value (CLA), where least count of the equipment is 0.001 μm for the travel length of 0.85 mm ( Figure 2 ).

Figure 2.

Surftest SJ-210 (Mitutoyo).

2.2 Design of experiment and objective

Five different set of fractional factorial (26–2 = 16) experimental design have been selected at two levels, so that 80 rows of experimental data can be observed at three level of replication on D2 using WEDM. In this study the main aim to minimize the surface roughness of D2 on best possible maximum MRR during WEDM ( Table 2 ).

Factors/three level (coding)123
Gap voltage (Vg): (volt)306090
Flush rate (Fr): (L/min)468
Pulse on time (Ton): (μS)1.051.151.25
Pulse of time (Toff): (μS)130160190
Wire feed rate (Wf):(m/min)258
Wire tension (Wt): (g)300600900

Table 2.

Factors for screening test.

2.3 ANN architecture and training

The hit and trail method based on literature have been adapted to find 7 and 10 neurons in primary and secondary hidden layers respectively, which effects on the R-square statistics for best prediction modeling. Tan sigmoid activation (squashing) function used as the (infinite input to finite output range) learning capability by the controllable instructed programme in MATLAB 2010a. Steepest descent problem used for the training algorithm to train the multilayer network, where the values of gradient was smallest because of the small changes in weight and biases. p1, p2, p3,p4, p5 and p6 are the six input layer neurons and Oi is the single neurons in output layer, whereas I11-I17 and I21-I29 (7 neurons present in primary and 10 in secondary hidden layers) are the hidden layers ( Figure 3 ).

Figure 3.

Artificial neural network approach.

3. Experimentation

Two models (S1 and S2) were developed from 80 rows of experimental data performed of D2. Only training result of best performing model (S2) of 55 rows is presented here for achieving the aimed to optimization of influencing process parameters ( Figures 4 and 5 ).

Figure 4.

Predictions against observations of Ra for model-D2, S2, 7 N (training dataset).

Figure 5.

Predictions against observations of Ra for model-D2, S2, 7 N (validation dataset).

OPTIMIZATION OF PROCESS PARAMETER Ra: D2, 7 Neurons in hidden layer.

The best model needs to be predicted among Model-S1 and S2, in D2 steel. Effect of individual input parameters will be observed on the Ra ( Tables 3 6 ).

SNGap voltage (Vg)Flush rate (Fr)Spark time (TON)Spark time (TOFF)Wire feed (Wf)Wire tension (Wt)Surface roughness (Ra) Obs.Surface roughness (Ra) Pred.Material removal (MRR)Square of residuals
(V)(Lit./min)(μS)(μS)(m/min)(g)(μm)(μm)(mg/min)(μm)2
13041.0513023001.68581.68631022.5E−07
23041.0516026001.44521.4451921E−08
33041.1513056001.38841.37131330.0002924
43041.1516053001.46581.4428950.000529
53061.0513056001.38361.37881252.304E−05
63061.0516053001.52781.55531100.0007562
73061.1513023001.6761.6756971.6E−07
83061.1516026001.5641.4909950.0053436
96041.0513053001.17721.17541043.24E−06
106041.0516056001.20761.2083884.9E−07
116041.1513026001.2731.26631364.489E−05
126041.1516023001.34761.34551164.41E−06
136061.0513026001.33221.32771102.025E−05
146061.0516023001.15981.13711150.0005153
156061.1513053001.2481.19451180.0028623
163081.1516089001.51241.54221450.000888
173081.1519086001.3631.34821080.000219
183081.2516056002.12562.1282065.76E−06
193081.2519059001.67941.68231018.41E−06
209041.1516086001.10981.1096884E−08
219041.1519089001.10961.0952630.0002074
229041.2516059001.35721.36641078.464E-05
239041.2519056001.32181.3425880.0004285
249081.1516059001.22861.2292913.6E−07
259081.1519056001.11941.1062640.0001742
266061.1516056001.40381.40231552.25E−06
276081.0513059001.45921.4591624E−08
286081.0516056001.36011.34411390.000256
296081.2513026001.52081.53022028.836E−05
306081.2516029001.54351.55351680.0001
319061.0513056001.31271.3118788.1E−07
329061.0516059001.29731.3023722.5E−05
339061.2513029001.18231.18671171.936E−05
349061.2516026001.08321.08121054E−06
359081.0513029001.23961.2696890.0009
369081.0516026001.18381.1739819.801E−05
379081.2513056001.14131.1524920.0001232
389081.2516059001.11251.13641120.0005712
396061.0513026001.45361.45461281E−06
406061.0516029001.32081.34741140.0007076
419081.0513029001.13691.1423962.916E−05
429081.0516026001.09621.0905783.249E−05
439081.2513056001.15511.1551990
449081.2516059001.17231.1153740.003249
453041.1516023001.68131.66281120.0003422
463041.1519029001.57821.55771080.0004202
473041.2516089001.49351.52831630.001211
483041.2519083001.46581.46661556.4E−07
493061.1516089001.64021.63681211.156E−05
503061.1519083001.61281.60211320.0001145
513061.2516023001.63681.63541031.96E−06
523061.2519029001.56091.56681083.481E−05
536041.1516083001.21361.19451230.0003648
546041.1519089001.18711.18781284.9E−07
556041.2516029001.20361.20351481E−08
Average1.3654113.8

Table 3.

D2, S1, 7N, training data (combined parameters).

SNGap voltage (Vg)Flush rate (Fr)Spark ON time (TON)Spark OFF time (TOFF)Wire feed (Wf)Wire tension (Wt)Surface roughness (Ra) obs.Surface roughness (Ra) predicted.(Residual)2Material removal predicted (MRR)
(V)(Lit./min)(μS)(μS)(m/min)(g)(μm)(μm)(μm)2(mg/min)
13041.0513023001.68581.68632.5E−07105
23041.0516026001.44521.44511E−0895
33041.1513056001.38841.37130.0002924119
43041.1516053001.46581.44280.000529102
53061.0513056001.38361.37882.304E−05115
63061.0516053001.52781.55530.0007562116
73061.1513023001.6761.67561.6E−07114
83061.1516026001.5641.49090.0053436102
96041.0513053001.17721.17543.24E−06108
106041.0516056001.20761.20834.9E−0796
116041.1513026001.2731.26634.489E−05131
126041.1516023001.34761.34554.41E−06123
136061.0513026001.33221.32772.025E−05111
146061.0516023001.15981.13710.0005153117
156061.1513053001.2481.19450.0028623112
163081.1516089001.51241.54220.000888136
173081.1519086001.3631.34820.000219105
183081.2516056002.12562.1285.76E−06189
193081.2519059001.67941.68238.41E−0697
209041.1516086001.10981.10964E−0879
219041.1519089001.10961.09520.000207470
229041.2516059001.35721.36648.464E−05110
239041.2519056001.32181.34250.000428592
249081.1516059001.22861.22923.6E−07101
259081.1519056001.11941.10620.000174269
266061.1516056001.40381.40232.25E−06153
276081.0513059001.45921.4594E−08158
286081.0516056001.36011.34410.000256143
296081.2513026001.52081.53028.836E−05208
306081.2516029001.54351.55350.0001163
319061.0513056001.31271.31188.1E−0774
329061.0516059001.29731.30232.5E−0593
339061.2513029001.18231.18671.936E−05122
349061.2516026001.08321.08124E−06111
359081.0513029001.23961.26960.000997
369081.0516026001.18381.17399.801E−0586
379081.2513056001.14131.15240.000123281
389081.2516059001.11251.13640.0005712106
396061.0513026001.45361.45461E−06135
406061.0516029001.32081.34740.0007076111
419081.0513029001.13691.14232.916E−0596
429081.0516026001.09621.09053.249E−0574
439081.2513056001.15511.1551094
449081.2516059001.17231.11530.00324988
453041.1516023001.68131.66280.0003422117
463041.1519029001.57821.55770.0004202100
473041.2516089001.49351.52830.001211158
483041.2519083001.46581.46666.4E−07163
493061.1516089001.64021.63681.156E−05115
503061.1519083001.61281.60210.0001145141
513061.2516023001.63681.63541.96E−06112
523061.2519029001.56091.56683.481E−05109
536041.1516083001.21361.19450.0003648123
546041.1519089001.18711.18784.9E−07125
556041.2516029001.20361.20351E−08144
Average1.36540.002642114.8

Table 4.

Training data for model: S2, Ra, N7, D2 steel.

SNGap voltage (Vg)Flush rate (Fr)Spark time (TON)Spark time (TOFF)Wire feed (Wf)Wire tension (Wt)Surface roughness (Ra)Material removal (MRR)Square of residuals
(Ra)
(V)(Lit./min)(μS)(μS)(m/min)(g)(μm)(mg/min)(μm)2
Vg
23041.0516026001.4452921E−08
73061.1513023001.676971.6E−07
276081.0513059001.45921624E−08
546041.1519089001.18711284.9E−07
209041.1516086001.1098884E−08
439081.2513056001.1551990
Fr
13041.0513023001.68581022.5E−07
556041.2516029001.20361481E−08
73061.1513023001.676971.6E−07
319061.0513056001.3127788.1E−07
276081.0513059001.45921624E−08
439081.2513056001.1551990
Ton
419081.0513029001.1369962.916E−05
429081.0516026001.0962783.249E−05
546041.1519089001.18711284.9E−07
209041.1516086001.1098884E−08
556041.2516029001.20361481E−08
439081.2513056001.1551990
Toff
73061.1513023001.676971.6E−07
276081.0513059001.45921624E−08
556041.2516029001.20361481E−08
369081.0516026001.1838819.801E−05
219041.1519089001.1096630.0002074
546041.1519089001.18711284.9E−07
Wf
349061.2516026001.08321054E−06
23041.0516026001.4452921E−08
439081.2513056001.1551990
319061.0513056001.3127788.1E−07
483041.2519083001.46581556.4E−07
163081.1516089001.51241450.000888
Wt
73061.1513023001.6760971.6E−07
13041.0513023001.68581022.5E−07
23041.0516026001.4452921E−08
319061.0513056001.3127788.1E−07
546041.1519089001.18711284.9E−07
219041.1519089001.1096630.0002074

Table 5.

D2, S2, 7N, training data (individual parameters corresponding to least square of residuals).

Correlation coefficient (R2): Training data of D2 steel (best performing model S2) using 7 and 10 neurons in primary and secondary hidden layers.

MaterialModelR2 valueEquation of lines (correlation between obs. and pred. values of Ra)Average predicted Ra (μm)Root mean square error (μm)Percentage RMSE (%)Average % RMSE RMSE
D2S1 Training0.983y = 1.005x - 0.0101.38640.0034010.24530.8129
0.7353
S1 Validation0.967y = 1.067x - 0.0901.30080.010770.8279
S1, Testing Testing0.963y = 0.879x + 0.1541.40160.019141.3655
S2 set, Training Training0.991y = 1.004x - 0.0081.36540.0026420.19340.3865
S2 validation0.988y = 0.984x + 0.0281.38880.0070150.5051
S2, testing testing0.979y = 1.006x - 0.0061.42320.0065650.4612

Table 6.

Summary of R2 values of training validation and testing data: 7 N in 1st and 10 N in 2nd L, Ra.

4. Optimization of process parameters

It is evident from Table 3 , that each independent influencing input parameter has corresponding values of their square of residuals at each three levels. Two values at each level (2 × 3 = 6 rows) has been taken for each inputs, where lowest possible square of residuals are available, to draw the Figure 6(a–f) .

Figure 6.

(a–f) 3D scattered plot between Ra vs. MRR vs. individual independent parameter.

5. Result

Figure 6 (a–f) shows the relations between individual influencing parameters (Vg, Fr, Ton, Toff, Wf and Wt) to their optimized response, surface roughness (Ra) with corresponding values of MRR. Table 5 also indicates that unique values of each influencing parameters (corresponding to its serial numbers of Table 5 ) gives optimum responses, which has been highlighted.

Again experiment has been conducted on D2 steel using WEDM by setting the individual optimum parametric combinations (Vg, Fr, Ton, Toff, Wf and Wt) as 90 (V), 8 (Lit./min), 1.05 (μS), 190 (μS), 2 (m/min) and 900 (g) respectively and found the values of Ra = 0.9638 (μm) at MRR = 105 (mg/min) ( Table 7 ).

SNGap voltage (Vg)Flush rate (Fr)Spark time (TON)Spark time (TOFF)Wire feed (Wf)Wire tension (Wt)Surface roughness (Ra) obs.Surface roughness (Ra) predicted(Zero residual)2Material removal predicted (MRR)
(V)(Lit./min)(μS)(μS)(m/min)(g)(μm)(μm)(μm)2(mg/min)
209041.1516086001.10981.10964.00E−0879
439081.2513056001.15511.1551094
429081.0516026001.09621.09053.25E−0574
546041.1519089001.18711.18784.90E−07125
349061.2516026001.08321.08124.00E−06111
546041.1519089001.18711.18784.90E−07125

Table 7.

Best parametric combination with their possible responses.

6. Conclusion

It has been concluded that the best fitted model (S2) for material removal rate and surface roughness of D2 steel has been achieved by artificial neural network using WEDM. From best modeled training data, optimum parametric combinations (Vg, Fr, Ton, Toff, Wf and Wt) observed as 90 V, 8 Lit./min, 1.05 μS, 190 μS, 2 m/min and 900 g respectively and found the values of Ra = 0.9638 μm at MRR = 105 mg/min, whereas the average Ra = 1.3654 μm at MRR = 114.8 mg/min. It has been concluded that ANN modeling technique is best fitted for surface roughness prediction and able to successfully minimize (SR) is 29.41% with 8.53% decreases the MRR from its average values on D2 steel using BPANN under WEDM. Such combinations may be applied for industrial application, where it is needed.

© 2019 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution 3.0 License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

How to cite and reference

Link to this chapter Copy to clipboard

Cite this chapter Copy to clipboard

Umesh K. Vates, N.K. Singh, B.P. Sharma and S. Sivarao (May 31st 2019). Optimization of Surface Roughness of D2 Steels in WEDM using ANN Technique, Applied Surface Science, Gurrappa Injeti, IntechOpen, DOI: 10.5772/intechopen.81816. Available from:

chapter statistics

104total chapter downloads

More statistics for editors and authors

Login to your personal dashboard for more detailed statistics on your publications.

Access personal reporting

Related Content

This Book

Next chapter

Synthesis and Nonlinear Optical Studies on Organic Compounds in Laser-Deposited Films

By Maria Marinescu

Related Book

First chapter

Advances in Aerodynamic Design of Gas Turbines Compressors

By Ernesto Benini

We are IntechOpen, the world's leading publisher of Open Access books. Built by scientists, for scientists. Our readership spans scientists, professors, researchers, librarians, and students, as well as business professionals. We share our knowledge and peer-reveiwed research papers with libraries, scientific and engineering societies, and also work with corporate R&D departments and government entities.

More About Us