Effect of Electrical Discharge Machining and Shot Blast Peening Parameters on Fatigue Life of AISI D 2 Die Steel

The present paper deals with studying the effect of electrical discharge machining (EDM) and shot blast peening parameters on work piece fatigue lives using copper and graphite electrodes. Response surface methodology (RSM) and the design of experiment (DOE) were used to plan and design the experimental work matrices for two EDM groups of experiments using kerosene dielectric alone, while the second was treated by the shot blast peening processes after EDM machining. To verify the experimental results, the analysis of variance (ANOVA) was used to predict the EDM models for high carbon high chromium AISI D2 die steel. The work piece fatigue lives in terms of safety factors after EDM models were developed by FEM using ANSYS 15.0 software. The results appeared that the experimental fatigue safety factors (at 10 6 cycles) decreased by (11 %) after EDM using copper electrodes compared with as-received material and this value is higher by (3.35 %) when using graphite electrodes. The fatigue strength at the same number of cycles was (0.88) and (0.84) times the fatigue strength of asreceived material for copper and graphite electrodes respectively. While fatigue strength and safety factors increased after EDM when increasing shot peening time, at the higher shot peening time is by (19.1 %) when using copper electrodes and by (23.26 %) when using graphite electrodes.


INTRODUCTION
Electrical discharge machining is one of the most successful, practical and profitable nonconventional machining processes for machining newly developed high strength alloys and creating complex shapes within the parts and assemblies in the manufacturing industry that cannot be done by conventional machines, Murray, et al., 2014 andMajhi, et al., 2013, with high degree of dimensional accuracy and economical cost of production, Prabhu, and Vinayagam, 2008.
EDM technique was progressed due to the growing application of EDM process and the challenges being faced by the modern manufacturing industries.New developments in the field of material science have led to new engineering materials that are hard, precise and difficult-tomachining metallic materials, composite materials, Sundaram, and Rajurkar, 2011, and Klocke, et al., 2012, and high tech ceramics, having good mechanical properties and thermal characteristics as well as sufficient electrical conductivity so that they can readily be machined by spark erosion Gu, et al., 2012 andJahan,et al., 2012.AISI D2 die steel is recommended for tools requiring very high wear resistance, combined with moderate toughness (shock-resistance).This grade of tool steel was chosen because of its wide range of application in tooling and manufacturing sections Atefi, et al., 2012 andMajhi, et al., 2014.EDM components are commonly applied in high temperature, high-stress, and highfatigue-load environments.Under such conditions, the cracks on the machined surface act as stress raisers and lead to a considerable reduction in the fatigue life of the component.Although a post-machining treatment can be performed to remove the recast layer to ensure the mechanical integrity of the component, this adds to the time and expense of the manufacturing operation.Accordingly, the current study conducts an experimental investigation of the economic and quick shot blast peening process to identify the optimal EDM machining parameters which suppress the formation of cracks in the recast layer for the longest lives under different fatigue loads.
Shot blast peening uses hard smooth hard steel balls with high velocities to yield a plastic deformation on the work piece surface layer.During the shot peening process, each piece of shot that strikes the material acts as a tiny peening hammer, imparting to the surface a small indentation or dimple.Shot peening is the most economical and practical method of ensuring surface residual compressive stresses.Compressive stresses are beneficial in increasing the fatigue strength, the wear resistance, endurance limit, the corrosion fatigue and to obtain better surface hardness and quality.Shot peening significantly improves the poor fatigue performance after EDM Stráský, et al., 2013 andDmowska, et al., 2012.The improvements of the fatigue strength, the wear resistance, endurance limit by induced residual compressive stress are the main aims of using the shot blast peening processes.The present paper concerns with studying, analyzing the effects of EDM and shot blast peening on fatigue life for AISI D2 die steel and developing numerical models for verifying the fatigue tests results by using the response surface methodology (RSM) and the finite element method (FEM) with ANSYS version 15.0 software.

EXPERIMENTAL WORK
The work piece specimens were prepared with dimensions 89.9x30x4.25 mm, according to requirement of the plain bending fatigue testing machine type Avery 7305, as shown in Fig. 1.The specimens for chemical composition and mechanical properties tests were prepared on the bases of ASTM-77 steel standard for mechanical testing of steel products ASTM A370, 1977.The specimens' dimensions and shape for fatigue tests is shown in Fig. 2. Two groups were fabricated for fatigue tests, where the second experimental group was used for shot blast peening processes.
Two types of electrode materials were selected (Copper and Graphite).The electrodes were manufactured with a square cross-section of 24 mm and 30 mm lengths, with a quantity of 24 pieces for each type.The work pieces after EDM machining with the used copper and graphite electrodes are depicted in Fig. 3.The prepared electrodes were polished and examined for chemical composition properties.The average values of chemical composition of the selected work piece material and the equivalent values given according to ASTM A 681-76 standard specification for alloy and die steels ASTM A681, 1976, are listed in Table 1.The results of tensile test and Rockwell hardness tests are given in Table 2.The chemical compositions of the copper electrodes are listed in Table 3.
The main EDM selected parameters include the gap voltage V p (140V), the pulse on time duration period T on (40 and 120 µs), the pulse off time duration period T off (14 and 40 µs), the duty factor (ƞ =75%), and the pulse current I P (8 and 22 A).Two side dielectric flashing with a pressure = 0.73 bar (10.3 psi).
The shot blast peening treatment processes were done on the drum type blast wheel (impeller) shot blasting machine shown in Fig. 4 for experimental group (2), which is similar to group (1) in all EDM parameters used the kerosene dielectric alone.The experiments were divided into three subgroups.The first subgroup includes the specimens numbers (1, 4, 7, 10, 12, 15, 18 and 20), used a shooting time of (30) minutes.The second subgroup includes the specimens numbers (2, 5, 8, 11,13,16,19 and 21), used a shooting time of (45) minutes, while the third subgroup which includes the specimens numbers (3, 6, 9, 14 and 17), used a shooting time of (60) minutes.
In this work, ( 22) experiments were done for each group using the ACRA CNC-EB series EDM / Taiwan which is shown in Fig. 5, where a new set of work piece and electrode was used in each experiment.The first (11) experiments were conducted by using the copper electrodes, while the last (11) experiments were done by using the graphite electrodes.The selected specimens and both electrodes materials were prepared after grinding, polishing processes for obtaining better fatigue examining characteristics.

THE INFLUENCE OF EDM PARAMETERS ON SURFACE ROUGHNESS CHARACTERISTICS
The influence of EDM parameters on the surface roughness characteristics for each work piece and each electrode (copper and graphite electrodes) was done before and after EDM machining and after the shot peening surface treatments by using the portable surface roughness tester.Fig. 6, shows that SR values increase with increasing the pulse current and pulse on duration.The use of graphite electrodes gives SR values less (better) than using the copper electrodes because their higher thermal and electrical conductivity produce a uniform value of discharge energy at lower pulse current and time, works to minimize the defects resulting from increased discharge energy, such as electromechanical pits and decay formation which keep the producing surfaces with higher quality and fine roughness.Fig. 7 shows the influence of EDM and shot peening parameters on the work pieces surface roughness (SR) indicates that the SR values are reduced with lower values of pulse current, pulse on duration time and longer shooting time.Increasing the pulse current and time producing high thermal energy generated that causes high melting with cooling accelerated cycles causing an increase in hardness and thus the lack of effect of shot blast peening process on the surface roughness.It is also noted that the surface roughness when using copper electrodes is higher than that of graphite electrodes due to high electrical resistivity of copper, which helps to generate high spark energy.When using lower values of pulse current and times, considerably less energy is generated and that will soften the metal causing a significant effect of the shot peening so improved surface roughness.

MODELING AND SIMULATION FATIGUE LIFE USING FEM
In this ANSYS fatigue analysis, the Von-Mises stress theory was used to compare against the experimental stress value.Fatigue strength factor is a modification factor to account the differences between the components in service from the as tested conditions.
The Multiphysics, static structural models domain loads, include the environment temperature, the fixing supported and the loading force.Setting the fatigue strength factor (Kf), which is equal to (1) and (0.72) for flat as received specimens and for EDM machining work pieces, respectively Shigley, and Mischke, 2006.The experimental fatigue results for both groups after EDM and shot blast peening processes are given in Table 4 and 5, respectively.
The experimental average values of fatigue strength at ( cycles) and the experimental and numerical fatigue safety factor values for groups (1) and ( 2) are given in Table 6 and 7, respectively, where the fatigue safety factor values were calculated as the ratio of fatigue strength at ( cycles) of the any experimental result with respect to the fatigue strength at ( cycles) of the as received material which is equal to (270 MPa).
The S/N fatigue strength obtained at ( cycles) curves after EDM machining are shown in Fig. 8 and 9 using pulse current (8 A) and (22 A), respectively.These figures show that, copper electrodes gave fatigue life values higher than graphite electrodes, and fatigue life increasing with decreasing the pulse current and increasing the pulse on duration time.While, the fatigue lives values for experimental group (2) are increasing with the decrease of pulse current values and pulse on duration time and the increase of blast shot peening time and graphite electrodes gave fatigue life values higher than copper electrodes.
Three level factorial response surface methodology (RSM) and the design expert 9.0 software were used to analyze the obtained fatigue safety factor for each two experimental parametric subgroup.The (ANOVA) technique was used to analyze the significance of EDM process and the shot blast peening parameters, where the F-test ratio is calculated for a 95% level of confidence.The inversion model obeys the least squares theory Lawson C. L et al, 1974, Kariya T. and Kurata H., 1975.The ANOVA function then runs in order to assess the results for group (1) experiments using the copper and graphite electrodes and by using the inverse forward transform for two factorial models given in Table 8.The Model F-value of 8.35 implies the model is significant.The lower the p-value, the more significant in the results expected.In terms of statistical significance, it is often suggested that when the p value is more than 0.05, it is corresponding to a 5% confidence.Values of "Prob > F" less than 0.0500 indicate model terms are significant.In this case A, B, C are significant model terms.
Table 9 shows the ANOVA analysis for group (2) experiments using copper and graphite electrodes after EDM machining and shot blast peening with linear reduced partial sum of squares transform model.The model F-value of 18.76 implies the model is significant.In this case A, B, C, D are significant model terms of estimated regression obtained as shown in Table (7).
The maximum fatigue life and safety factor obtained by the FEM and ANSYS solutions and simulations using the copper and graphite electrodes at the pulse current (8 A) and pulse on time (120 µs) are given in Fig. 10 and 11 for group (1) and (2) using the copper and graphite electrodes at the same current value, the lower pulse on time (40 µs), respectively, and longer shot time for experimental group (2).Each of these tables shows two simulation figures for each of input parameters EDM sub-group.The right figures represent the numerical modeled fatigue safety factor.The figures in the left show the fatigue life model simulation and the fatigue strength at ( cycles), which were obtained from the S/N curve of each experimental subgroup, the input EDM process parameters and the model loading force.
The final predicted empirical equation of fatigue strength (at 106 cycles) for actual factors obtained after EDM machining by using of copper electrodes for group (1) is: Fatigue strength at 10 6 cycles = +239.03571-1.28571*Pulse current +0.087500 * Pulse on time (T on ) (1) And, when using graphite electrodes is: Fatigue strength at 10 6 cycles = +228.53571-1.28571*Pulse current +0.087500 * Pulse on time (T on ) (2) For experimental group (2), the final predicted empirical equation after EDM machining and shot blast peening processes using copper electrodes is: The analysis of results for fatigue safety factor for both experimental groups using the copper and graphite electrodes are shown in Fig. 12 and 13, respectively.While, the fatigue stresses at ( cycles) are shown in Fig. 12and 13, respectively.
Fig. 12 shows the fatigue safety factor analysis for group (1) using the copper electrodes, where the fatigue safety factor values are increasing with the decrease of pulse current values and the increase the pulse on duration time, reaching the maximum value as (0.85), experimentally (0.89) compared with the fatigue safety factor for as received material, which is equal to one, at a current value of (8 A) and a pulse time of (120 µs).Whereas, when using the graphite electrodes, the fatigue safety factor values reached the maximum value as (0.80), experimentally (0.86) at the same input current and time on period, as shown in Fig. 13.This means that the use of copper electrodes and the kerosene dielectric alone gives higher fatigue safety factor values by (3.35 %) when compared with the use of graphite electrodes.Fig. 14 shows the analysis of fatigue strength at ( cycles) using the copper electrodes, where these fatigue stresses values are increasing with the decrease of pulse current values and the increase of pulse on duration time, reaching the maximum value as (240 MPa) at a current value of (8 A) and pulse on time (120 µs).When using the graphite electrodes, these fatigue stresses values reached the maximum value as (232 MPa) at the same input current and time on period, as shown in Fig. 15.This means that the use of copper electrodes and the kerosene dielectric alone gives higher fatigue stresses (at 106 cycles) values by (3.45 %) when compared with the use of graphite electrodes and a pulse time of (120 µs).
These values of strength are equal to the ratios (0.88) and (0.84) for copper and graphite electrodes, respectively compared with the fatigue stresses (at cycles) for the as received material, which equal to one.The high fatigue safety factors and fatigue stresses (at cycles) levels obtained when using the copper electrodes are because the copper material has higher electrical resistivity and lower conductivity which produced lower heat discharges energy at the gap between the electrode and the work piece, especially with longer period of pulse on time, where the plasma channels are better arranged and then less unwanted metallurgical changing with brittle carbides formation will occur with less defects and lower white layer thickness.And, all these factors are strengthening the work piece against fatigue failure and then longer lives were obtained.
The using of graphite electrodes also produced higher unwanted carbides due to high heat formation and the carbon particles migration to the work piece as well as the carbon particles in the kerosene dielectric, where these brittle carbides especially in die steel grade with high carbon, high chromium and other added elements tend to form carbides, and then lower fatigue lives will be obtained.
The fatigue safety factor for experimental group (2) values using the copper electrodes are increasing with the decrease of pulse current values and pulse on duration time and the increase of blast shot peening time, reaching the maximum value as (1.22), experimentally (1.05) at a current value of (8 A), a pulse time of (40 µs) and longer shot time (60 min.).While, when using the graphite electrodes, the fatigue safety factor values reached the maximum value as (1.29), experimentally (1.06) at the same input current, pulse on time period and shot time, as shown in Fig. 12 and 13.This means that after EDM and shot blast peening processes, the use of graphite electrodes and the kerosene dielectric alone gives higher fatigue safety factor values by (0.95 %) when compared with the use of copper electrodes and higher by (19.10 %) and (23.26 %) when compared with the results of group (1) without using the shot blast peening and using the copper and graphite electrodes, respectively.Although the graphite electrode generates thermal energy more than that of copper, it works with the longer pulse time on annealing the work piece surface and on reducing the creation of martensitic structure, and that will lead to increasing the fatigue life.
The fatigue stresses at ( cycles) analysis using the copper electrodes for experimental group (2) are increasing with the decrease of pulse current, the pulse on duration time and the blast shot time values, reaching the maximum value as (284 MPa) at a current value of (8 A), pulse on time (40 µs) and blast shot time (60 min.).Whereas, when using the graphite electrodes, these fatigue stresses values reached the maximum value as (287 MPa) at the same input current and time on period time, as shown in Fig. 14 and 15.
The reason of obtaining higher fatigue safety factor is because the use of low pulse current generates lower thermal energy, which cannot work to make large metallurgical changes in the crystalline structure of the work piece surface.Also, the abrasion process of EDM machining cannot accomplish its work completely due to the high amount of thermal energy necessary for melting the surface layer of work piece, and thus the abrasive phenomenon will work with less abilities required to remove the surface layers as well as the lack of interactions required for the generation of new carbides due to low level of energy generated.
This means that the use of graphite electrodes and the kerosene dielectric alone after EDM and blast shot peening processes fatigue stresses (at 10 6 cycles) gives higher values by only (0.35 %) when compared with the use of copper electrodes and yields a higher fatigue life than the situation when working without shot peening processes by (19.58 %) and (23.71 %) using the copper and graphite electrodes, respectively.
The values of these stresses are equal to the ratios (1.05) and (1.06) for copper and graphite electrodes, respectively compared with the fatigue stresses at ( cycles) for the as received material.The high fatigue safety factor and fatigue stress (at cycles) levels obtained when using graphite and copper electrodes are because the lower levels of current and pulse on time period produced a lower heat discharges energy at the gap between the electrode and the work piece.This means that less unwanted metallurgical changing with brittle carbides formation will be obtained due to lower level of carbon particles migration from the electrode to the work piece and also less defects and lower white layer thickness.
And, all these factors are strengthening the work piece against fatigue failure and then longer lives were obtained with the use of high effective techniques of shot blast peening, which is working on the conversion of tensile surface residual stresses to high level of compressive residual stresses and produced a new strength surfaces with preventing of micro cracks and other surface defects, especially at these low levels of input parameters.The work pieces surfaces are still soft, and a good surface hardening operation by the shot blast peening was gained, consequently a higher fatigue lives will be obtained.

CONCLUSIONS
1-The fatigue safety factor after EDM compared with as-received material and fatigue strength are increased with the decrease of pulse current and increase of pulse on time, except when using the shot blast peening or graphite mixing powder, with decrease pulse on time.
2-The experimental fatigue safety factors and fatigue stresses after EDM and kerosene dielectric alone reached (0.89) using copper electrodes, which is higher by (3.35 %) when using graphite electrodes.
3-The fatigue stresses at ( cycles) are equal to the ratios (0.88) and (0.84) for copper and graphite electrodes, respectively compared with as received material, which equal to one, and reached the maximum value at a current value of (8 A) and pulse on time (120 µs).The use of copper electrodes gives higher fatigue stresses by (3.45 %) when compared with the use of graphite.
4-The fatigue safety factor and fatigue stresses after EDM and shot blast peening increased when using graphite electrodes, which increased by (0.95 %) compared with copper electrodes and higher by (19.10%) and (23.26%) when comparing with working without shot blast peening using copper and graphite electrodes, respectively.
5-A higher fatigue life were obtained than the situation when working without shot peening processes by (19.58 %) and (23.71 %) using the copper and graphite electrodes, respectively.
6-All fatigue stresses at ( cycles) for the as received material ratio are close to those results of fatigue safety factors for the same input parameters, and this proves the accuracy of EDM and PMEDM models developed by FEM using ANSYS software.
 Majhi, S. K., Pradhan, M. K., and Soni, H., 2013, Optimization of EDM parameters using integrated approach of RSM, GRA and Entropy method.Bal.No. - Stráský, et al., 2013, worked on multi-method characterization of combined surface treatment of Ti-6Al-4V alloy for biomedical use after EDM, acid etching and shot peening.Shot peening significantly improves poor fatigue performance after EDM.Dmowska, et al., 2014, presents the results of the influence of EDM parameters on surface layer properties.It was proved that the application of the roto-peen after the EDM resulted in lowering roughness height up to 70%.Havlikova, et al., 2014, presented an approach of surface treatment of electric discharge machining, chemical milling (etching) and shot peening resulting in significantly improves the favorable mechanical properties.A considerable amount of work has been reported on the measurement of EDM performance using various designs of experiments (DOE) techniques especially utilizing the (RSM).Mehdi et al., 2015, used response surface methodology (RSM) to analyze the effect of EDM parameters for machining Al-Mg-2Si composite material on microstructure.The results show that voltage and current, and pulse on time are the most significant factors.Santoki, and Ashwin, 2015, studied the recent developments and effect of machining parameters on performance parameters in EDM.Sabareesaan, et al., 2015, developed a prediction model for material removal rate (MRR) for electrical discharge machining of Inconel X750 by RSM using Minitab software.

Figure 1 .
Figure 1.The Avery Denison plain bending fatigue testing machine type 7305, England.

Figure 2 .Figure 3 .
Figure 2. The specimen dimensions and shape for fatigue tests

Figure 6 .Figure 7 .Figure 8 .
Figure 6.The 3D graph models for the effect of EDM parameters on surface rouphness (SR) for experimental group (1)

Figure 9 .
Figure 9.The S/N curves for both experimental groups after EDM and shot blast peening, using pulse current (22 A).

Figure 12 .Figure 13 .Figure 14 .Figure 15 .
Figure 10.The FEM fatigue life and safety factor Models for copper and graphite electrodes for experimental group (1).

Table 2 .
The mechanical properties of the selected materials.

Table 4 .
The experimental fatigue life results for experimental group (1) after EDM machining.

Table 5 .
The experimental fatigue life results for experimental group (2) after EDM machining and shot blast peening processes.

Table 6 .
The experimental average values of fatigue stress at ( cycles) and fatigue safety factor for group (1) after EDM machining.

Table 7 .
The experimental average values of fatigue stress at ( cycles) and fatigue safety factor for group (2) after EDM and shot blast peening processes.

Table 8 .
The (ANOVA) table for the EDM machining input and response factors for group (1) experiments

Table 9 .
The (ANOVA)table for the EDM machining and shot blast peening inputs and response factors for group (2) experiments