1 Introduction

High Strength Low Alloy (HSLA) are widely used steels in the automotive industry and structural components, where weight-saving is desirable, mainly due to its high strength-to-weight ratio [1, 2]. These steels are characterized by their specific mechanical properties such as weldability, corrosion, and abrasion resistance [3, 4]. Pieces made of HSLA steels become difficult to machine by conventional cutting methods to get the final shape. When conventional methods are used, the operation cost related to wear on tools increases. In order to overcome this limitation, it is proposed to use electric discharge machining (EDM) on HSLA steels [5].

The EDM process has the advantage that, regardless of workpiece hardness, it usually minimizes distortions and vibration problems during the machining. EDM is a thermal process for machining of conductive materials [6]; therefore, it is reported [7, 8] that the machined surface is modified due to the high temperature that produces microstructural and micro hardness changes, element distribution modification, heat affected zone (HAZ), as well as thermal and tensile residual stresses. All of these effects are not desirable on automotive HSLA steel components.

It is important to mention that EDM is based on spark discharges, because each of these discharges generates a plasma channel with high temperature (above of 8000 K) and high energy density capable of melting and vaporizing the workpiece material, but also from the tool electrode [9], in which some of electrode material surface is eroded and deposited onto the workpiece surface [7] producing surface modification. Moreover, the decomposition of dielectric is another way that the workpiece surface becomes contaminated with elements such as C, O, and H [10].

The amount of material removed by a sequence of discharges between the workpiece and electrode generates a re-solidified material layer (known as recast or white layer) on the machined surface. In the literature [11], it is reported that the recast layer has complex chemical products that increase surface hardness due to the gradient transfer of elements from the electrode and the pyrolytic carbon from the dielectric [12]. The increase in hardness promotes a decrease in the ductility of the surfaces and micro cracks appears in the subsurface region. In addition, EDM discharge energy has a direct relationship between surface finish and mechanical properties on the machined surface [13].

In this context, this research aims to study the effect of parameters on the microstructures due to thermal effects that are present during the removal of material, by the EDM process, in order to achieve a better understanding of the parametric selection during machining of high-technology components.

2 Experimental

An experimental design 2K was designed and conducted in random order. The factors of the experiment were selected, including discharge current intensity, duty cycle, and auxiliary capacitance. The different experimental conditions as well as their initial levels are shown in Table 17.1, resulting in sixteen runs (with a replica). The aim of this experimental set is to obtain information about the perforations integrity, measuring the material removal rate (MRR), and the ratio between exit diameter and entrance diameter, as a measure of machining precision. The “MINITAB” statistical software was utilized to design and analyze experimental data.

Table 17.1 Experimental conditions (factors) and their levels for the initial set of experiments
Full size table

The parameters are chosen based on reports in literature [14, 15], and they were the independent controllable EDM parameters. As it is reported, current has a great influence on the MRR because current increases the discharge energy; the duty cycle is the percentage fraction between the pulse on-time and signal period, which also can influence the discharge energy. Moreover, the capacitance is used to stabilize the shape of pulse during the process.

Different machining tests were performed on the SODICK K1C EDM drilling system. SAE945X HSLA steel plates were used for drilling, while a brass electrode tube was chosen with a circular cross section of 0.800 mm diameter and “VITOL” as dielectric medium. The pulse frequency was fixed at 25 kHz; the gap voltage was fixed to 22 V. The resulting specimens were characterized by Philips scanning electron microscope (SEM). The base metal was also characterized by optical microscopy. After the experiment, the workpiece was examined with optical microscope OLYMPUS STM6.

Prior to drilling, two plates of steel were cut at dimensions of 47 × 24 mm. Both plates were prepared with a grinded cross section face, in order to put them face to face and ensure full contact, the two pieces were held together in a bench vice. Drills were done just in the joint between the plates, and every hole was separated by 2 mm from the next (see Fig. 17.1).

Fig. 17.1
figure 1

Schematic diagram of EDM setup

Full size image

The machining time was measured by two observers using stopwatches, one observer starts to take time when the electrode begins sparking at the top of workpiece and stops when the electrode is visible out of the workpiece; the second observer starts to take time when the voltmeter shows an increase in applied voltage and stops when the voltmeter shows significant decrease in variation. The average of both sampled times was taken as the real machining time. To estimate the MRR, the entrance and exit diameters of the hole were measured, while the volume was calculated using the truncated cone formula; then it was divided by machining time, which provides the removal of material as a function of time.

These plates were subsequently separated and then evaluated by microscope technique, after metallographic preparation using a standard polish technique. An ultrasonic cleaner “Kendal HB-S-49DHT” in each step was used. Nital 2 % was used for chemical etching to reveal the ferrite boundary.

3 Results and Discussion

3.1 Characterization of Material and Influence of Parameter on Perforation Integrity

A specimen of HSLA steel, in as-received condition, was analyzed by optical emission spectroscopy (OES) to determine chemical compositions (see Table 17.2), and the microstructure of HSLA steel was evaluated by optical microscopy (Fig. 17.2). Phase distributions of pearlite, martensite, and ferrite embedded in a banded microstructure [16] are exhibited in Fig. 17.2 due to the direction of the rolling process. This initial microstructural condition is crucial to analyze the effects of parameters on the drilling of steel because of the characteristics in chemical composition and hardness of the phases.

Fig. 17.2
figure 2

Micrograph of SAE 945X SHLA steel as-received, (a ) Microstructural banding and (b ) Closed up image showing ferrite (F), pearlite (P), and martensite (M)

Full size image
Table 17.2 Chemical composition of SAE 945X HSLA by OES (wt%)
Full size table

Likewise, Table 17.3 shows the results of the experiments after drilling the HSLA steel. The first output parameter is machining time, then with the time machining the MRR was calculated. The value of the ratio between the exit diameter and the entrance diameter (ϕ ext/ϕ ent) means that as the value approaches one, the perforation has a cylindrical shape.

Table 17.3 Experimental design and replica results
Full size table

The regression coefficients in coded form are given in Table 17.4, which indicates the individual and interaction effects of discharge current, duty cycle, and capacitance. The predictors with significant contributions in mathematical models are identified with their P values less than 0.05.

Table 17.4 Regression coefficients of experimental design
Full size table

Figure 17.3 shows the main effects and interaction effect for MRR. As can be seen in Fig. 17.3a, the discharge current has a steeper slope, which indicates a factor that has a greater impact on the output measure in Fig. 17.3b. Discharge current (I), capacitance (C), and the interaction of current-duty cycle (I-DC), current-capacitance (I-C) are the significant parameters that promote material removal. The higher the level of current and capacitance, the more material is removed (Fig. 17.3). This behavior is achieved because MRR is proportional to the amount of energy supplied during each pulse [10]. The latter agrees with literature [17], where the discharge energy (W e in Joules) is influenced by the current intensity (I e in Amperes), discharge voltage (U e in volts), and time machining (t in seconds);

$$\displaystyle{ W_{e} =\int _{ 0}^{t_{\mathrm{on}} }U_{e}\left (t\right )I_{e}\left (t\right )dt }$$
(17.1)
Fig. 17.3
figure 3

Main effects plot for MRR (left) and, interaction effects plot for MRR (right)

Full size image

Equation (17.1) shows the discharge energy as a function of time; this behavior explains the increasing of MRR when the current and duty cycle are increased. In literature [14], it is reported that the increasing of energy discharge causes the plasma channel to expand, resulting in more material being melted. Thus, more material is removed taking less time to drill.

The relationship ϕ ext/ϕ ent is only influenced by the interaction of duty cycle and capacitance. Figure 17.4 shows a steeper slope in duty cycle and capacitance. It appears that when the lower level for duty cycle approaches one, the perforation becomes more of a perfect cylinder. In the case of capacitance, it is the opposite result that holds true; the higher value of capacitance results closer to one. This phenomenon can be attributed to the stability of plasma arc formed in each spark; when short pulse off-time is not possible, remove the melted material, causing subsequent sparks to be unsteady, which increases the possibility the next discharge occurs in the same place, promoting the erosion in only one place. The more duty cycle increases, the more energy discharges, which produces more electrode erosion, so that the electrode loses its original dimensions [14].

Fig. 17.4
figure 4

Main effects plot for ϕ ext/ϕ ent (left) and interaction effects plot for ϕ ext/ϕ ent (right)

Full size image

3.2 Characterization of Holes Drilled by EDM

In order to investigate the role that the parameters of EDM process play on the microstructure, selected samples with and without burr were inspected. Figure 17.5 shows holes made in the HSLA steel by EDM process under the parameters that are shown in the Table 17.1. As it can be seen, Fig. 17.5b shows melted material on the top. By contrast, Fig. 17.5a shows a hole without melted material. In addition, the melted material showed segregation at the top and edge, which is demonstrated and explained in detail in the following paragraph.

Fig. 17.5
figure 5

SEM image of perforations made by EDM process; (a ) 15 A, 50 % duty cycle,0.02 μF capacitance; (b ) 35 A, 50 % duty cycle, 0.02 μF capacitance

Full size image

In order to study the thermal effect using the parameters of the experimental design, these specimens were observed by SEM. Figure 17.6a–e shows the burr on the hole of the specimen as well as chemical composition by EDX. Unsurprisingly, the high current parameter value used in the experimental design can increase the temperature that melts the copper electrode in a large quantity and deposits onto the workpiece. This result is consistent with the Cu-rich zones detected in the EDX spectrum (Fig. 17.6b). Others elements such as Mn and Si from the steel workpiece were detected at the top and edges of the hole (Fig. 17.6c, d). The content of oxygen is associated to the oxidation of iron (Fig. 17.6e). These findings indicate that at high temperature, segregation in the base metal may occur at rapid rate due to high pressurized dielectric flow, therefore, changing the mechanical behavior in these zones and leading to phase transformations.

Fig. 17.6
figure 6

SEM image and EDX of the hole using 35 amperes, 50 % duty cycle, 0.02 μF capacitance. (a ) Burr formation; (b ) Cu; (c ) Mn, (d ) Si, and (e ) O

Full size image

The ɛ-Cu precipitates can destroy the austenite interphase boundary by the martensitic transformation during quench as well as influence on the kinetics of pro-eutectoid cementite precipitation [18]. Likewise, segregation of Cu can increase the susceptibility to hydrogen cracking, which is particularly related to the steel composition and kind of nonmetallic inclusions (type, size, and morphology) and the material ability to accommodate hydrogen. Reports in the literature [19] suggest that copper affects hydrogen uptake; above 0.2 %, Cu susceptibility is reduced. Moreover, it reduces the concentration of hydrogen at the surface but does not alter the threshold of hydrogen concentration for cracking. Over 1 % Mn increases the susceptibility of cracking, but the cooling rate can remove the detrimental Mn effect [20, 21].

Inspections near to the martensitic transformation zones show, in detail, the presence of microcracks as well as pro-eutectoid cementite precipitates in the boundary with a small content of copper analyzed by EDX (Fig. 17.7a, b). It is postulated that copper as a substitutional atom inside the lattice of martensite can lead a plastic strain so that it exceeds the yield stress of steel and begins microcrack growth.

Fig. 17.7
figure 7

SEM images of samples with 35 amperes, 70 % Duty cycle, 0.02 μF Capacitance. (a ) Microcracks inside martensite and EDX and (b ) Pro-eutectoid cementite precipitates and EDX on the boundary

Full size image

Hashimoto and Kunieda [22] reported that a surface temperature can reach up to 5000 K during the EDM process. The combination of high temperature, which vaporize and/or melt material of both electrode and workpiece, and the rate of cooling (due to high pressurized dielectric fluid) contributes in a wide heat affected zone (HAZ) that is associated with the transformations from austenite to martensite as well as the deposition of molten copper from the electrode onto the base metal, when the following parameters are used 35 A, 50 % duty cycle, 0.02 μF. This finding agrees with other reports in the literature [23–25] that suggest the deposited copper is caused by trapping of the molten electrode in the recast layer on the workpiece surface.

On the contrary, the foremost difference between the low current value and the high current value conditions is that segregation clusters of Cu, Mn, and Si do not present at the top and edge of the specimen without melted material (Fig. 17.8a–e). This condition shows a better homogenous chemical composition than at high conditions of machining.

Fig. 17.8
figure 8

SEM image and EDX of the hole using 15 amperes, 50 % duty cycle, 0.02 μF capacitance according to experimental design. (a ) Melted material at the top; (b ) Cu; (c ) Mn; (d ) Si; and (e ) O

Full size image

Figure 17.9 shows a deeper HAZ with approximately 35 μm, and it exhibits phase transformation to martensite at the edge of perforation. This occurs when 35 A are used (50 % duty cycle, 0.02 μF capacitance. Under a minor current level (15 A, 50 % duty cycle, 0.02 μF), HAZ is about 14 μm.

Fig. 17.9
figure 9

SEM image of the edge of perforation at (a ) 15 A, 50 % duty cycle, 0.02 μF capacitance (b ) 35 A, 50 % duty cycle, 0.02 μF capacitance

Full size image

4 Conclusions

From the statistical analysis, it can be inferred that the discharge current intensity has a greater influence on the MRR, while capacitance has a reduced influence on the increase of MRR. Regarding the integrity of the perforations, the most influential parameters are the capacitance and duty cycle; to produce holes without taper, both low duty cycle and high capacitance are required.

With high current levels, trapping of copper particles in the recast layer of the perforations are found, producing changes in material properties. In addition, the change of ferrite-pearlite phases to martensite is promoted.

The transformation to martensite phase in the presence of segregated copper causes the formation of microcracks, which could reduce the corrosion, mechanical resistance, and fatigue resistance of the EDM’ed surface. On the other hand, it is postulated that copper as substitutional atom inside lattice of martensite can lead a plastic strain and increase the feasibility of microcracks and the same fragile nature of martensite.

Different levels of discharge energy produce a HAZ near the subsurface region, which has some variation in thickness along the EDM’ed surface. The discharged energy increases the plasma temperature, which affects a deeper zone. The HAZ is about 15 μm at 15 A discharging current, but it is more than double to 36 μm at 35 A. Also, in this strip, martensite phase and thickening of cementite plates precipitation are formed.

After completion of statistical analysis, this work suggests that EDM is a good choice for machining high-strength steels at high MRR, when the machining parameters are carefully combined. However, some secondary effects must be taken into account, namely: surface finish and geometric precision.

If reliability of the EDM’s surface is a concern, then a subsequent process could be required in order to remove damaged zones. Since the holes are small, internal honing is not suitable, but electrochemical polishing should be considered as a good alternative to improve the surface finish quality.