1 Introduction

Recently, wire electrical discharge machining (WEDM) is a popular nontraditional process method in many manufacturing industries, such as mold, tool, automobile, and aerospace. WEDM can not only machine any conductive material regardless of hardness but also some workpiece with intricate shape. Besides, there is weak or nonexistent cutting stress in the subsurface material because WEDM is a noncontact machining process, and the workpiece material is removed by way of melting, vaporizing, and thermal explosion due to ultrahigh discharge energy density [1,2,3,4]. But, because WEDM is a thermal-removing process, recast layer and heat-affected zone are inevitably formed which have a direct effect to the use performance and service life of workpiece. Moreover, with the substantial growth of manufacturing requirement, studying and improving the machining efficiency and surface quality of WEDM are always research hotspots.

1.1 Literature review

A considerable amount of researches about the machining performance of WEDM had been carried out which mainly included experiment study, parameters optimization, an expert system, and a magnetic-assisted method. As for experiment study, Huang et al. [5], Ravindranadh et al. [6], and Puri et al. [7] carried out lots of experiments to investigate the relationships between process parameters and machining characteristics. Several mathematical models of technological indexes (such as MRR, Ra, kerf width, crack density) had been fitted out by the method of regression analysis. Zhang et al. [8] focused on exploring the influences of machining parameters on surface integrity in cutting nanocomposite ceramic process. It was pointed out that the way of removing ceramics included melting, evaporating, and thermal exploding. About parameter optimization, Mahapatra et al. [9], Kuriakose et al. [10], Yuan et al. [11], and Mukherjee et al. [12] employed several optimization algorithms to search the optimal parameter combination to obtain fast cutting speed, good surface roughness, and high geometrical precision simultaneously, such as nonlinear regression analysis, nondominated sorting genetic algorithm, Gaussian process regression, and nontraditional optimization algorithms. As described in their studies, the method of parameter optimization could significantly improve machining efficiency and surface quality. However, it was a black box model which depended on a mass of repeated cutting experiment. The function mechanism of the relationship between process parameters and machining performance had not been revealed. As regards expert system, Katz et al. [13], Huang et al. [14], and Lautre et al. [15] utilized the method of neural network to establish several expert systems for maintenance and fault diagnosis which was time saving in knowledge acquisition and with the ability of self-learning, and they were very convenient for operators to maintain machines well. However, they were not really expert systems which could adaptively control machining parameters for various machining conditions. It is necessary to take a good deal of time to build an expert system which is suitable to any machining condition because the optimal machining parameters vary with workpiece material, workpiece thickness, and other factors. Liao et al. [16] proposed the concept of specific discharge energy (SDE) to analyze the relationship between process parameters and the machining performance of different materials. It was pointed out that the materials with close SDE could perform similar machining characteristics. Yet, it was very hard to find two materials with close SDE in practical process. The mechanism of SDE of different materials had not been clarified in their study. In regard to magnetic-assisted method, Lin et al. [17, 18] added magnetic force to expel the removed debris out in EDM, and the machining characteristics (such as machining efficiency, electrode wear rate, and surface roughness) had been effectively improved. Zhang et al. [19] and Wang et al. [20] proposed a hybrid technique for enhancing the MRR, surface roughness, and surface crack density based on a magnetic-assisted method and ultrasonic vibration in WEDM. It was found that the machining performance could be effectively improved under the appropriate process parameters. However, there were few researches about the effect of magnetic field on the workpiece subsurface quality of ferromagnetic material in WEDM.

1.2 Outline of our work

The goal of this paper is to present two measures for improving the machining characteristics based on specific discharge energy and magnetic field–assisted method. A set of a single-factor experiment is implemented to analyze the effects of carbon content and discharge parameters on machining performance. Besides, the concept of SDE is proposed to reveal the fundamental influence of workpiece material properties on the MRR and to obtain the optimal machining feed rate. Moreover, the surface quality is investigated in terms of surface roughness, surface microstructure, microhardness, and recast layer. Finally, the verified experiment is carried out to evaluate the improvement effects of the two proposed methods on the machining characteristics.

2 Experiment

2.1 Material and machine tool

In this study, five different models of carbon steels, which are widely used in many manufacturing industries, are chosen as workpiece material. Their chemical compositions are as shown in Table 1. According to previous researches [21], the removing process of WEDM is very complex and violent because thousands of discharge sparks occur in one second. Workpiece material is eclipsed by way of melting and evaporating. Hence, the removing mechanism of WEDM is widely considered as thermo-physical removing [22, 23]. Then, several thermo-physical properties are chosen as representative factors, as shown in Table 2.

Table 1 The chemical compositions of five models of carbon steels
Full size table
Table 2 The physical properties of five models of carbon steel
Full size table

Besides, the cutting experiments are carried out on the independently developed 5-axis dedicated machine tool for wire electrical discharge machining, as indicated in Fig. 1. It mainly consists of a CNC system, power supply, lathe bed, 5-axis motion platform, machining region, wire winding system, and cooling system. The power supply is a high-performance and high-frequency pulse generator. The configuration parameters of this WEDM machine tool are listed in Table 3.

Fig. 1
figure 1

The physical map of WEDM machine tool

Full size image
Table 3 The configuration parameters of WEDM machine tool
Full size table

2.2 Experiment design and result

In this study, a set of a single-factor experiment is implemented to analyze the influences of carbon content and discharge parameters on machining characteristics. The discharge parameters are composed of gap peak voltage (SV, unit: V), discharge peak current (I, unit: A) and pulse-on time (ton, unit: μs). The experiment design and result are indicated in Table 4. The other parameters are as follows: pulse-off time is 10 μs, wire tension is 15 N, wire speed is 0.15 m/s, dielectric pressure is 0.8 MPa, workpiece thickness is 20 mm, and feed speed is 5 mm/min. The MRR can be calculated by Eq. 1.

$$ \mathrm{MRR}=\frac{LH}{t} $$
(1)

where L is the cutting length, H is the workpiece thickness, and t is the machining time (unit: min).

Table 4 The design and result of single-factor experiment
Full size table

3 Discussion

In this section, the influence trends of carbon content and discharge parameters on machining performance are obtained according to experiment data. The concept of specific discharge energy is presented to clarify the variation trend of the MRR. In addition, the experiments of microstructure observation and microhardness measurement are implemented to reveal the variation principle of RLT.

3.1 Material removal rate

3.1.1 Influence trends

Figures 2, 3, and 4 show the effects of carbon content and discharge parameters on the MRR. As shown in Fig. 2, it can be found that the MRR increases with the increase of pulse-on time. The MRR obtained by 12-μs pulse-on time is about 1.9 times faster than that by 6 μs when workpiece material is with the same carbon content. Besides, it can be observed from Fig. 3 that the MRR is approximately linearly related to discharge peak current. The growth rates with discharge peak current of the MRR of different carbon steels are roughly the same. In addition, as indicated in Fig. 4, it can be seen that the MRR has a positive correlation with gap peak voltage. It is common knowledge that the free electrons escape from electrode surface and collide the workpiece surface at a very high velocity due to ultrahigh electrical field strength. Most of electron kinetic energy is converted to thermal energy because of the huge mass difference between atom and electron. The effective discharge energy for removing workpiece material can be obtained by Eq. 2 [24, 25]. Above three influence relationships between the MRR and discharge parameters can be explained by Eq. 2. The higher gap peak voltage, the larger discharge peak current, and the longer pulse-on time can produce more discharge energy for removing workpiece material.

$$ Q=\eta \times \mathrm{SV}\times I\times \frac{t_{\mathrm{on}}}{t_{\mathrm{on}}+{t}_{\mathrm{off}}} $$
(2)

where Q is the energy absorbed by workpiece, η is the energy coefficient for workpiece, SV is the gap voltage, I is discharge current, ton is pulse-on time, and toff is pulse-off time.

Fig. 2
figure 2

The influence of carbon content and pulse-on time on material removal rate

Full size image
Fig. 3
figure 3

The influence of carbon content and discharge peak current on material removal rate

Full size image
Fig. 4
figure 4

The influence of carbon content and gap peak voltage on material removal rate

Full size image

Besides, in Figs. 2, 3, and 4, there is one thing in common that the MRR of carbon steel with low carbon content is faster than that with high carbon content. The MRR of 10# steel is about 20% faster than that of the 60# steel under the same discharge parameters.

3.1.2 Specific discharge energy

From Subsection 2.1, it can be obtained that the mechanism of WEDM is thermal-removing process. The MRR of different materials varies with the physical properties of workpiece under the same discharge parameters. Hence, the concept of specific discharge energy (SDE, unit: mm3/J) is proposed to synthetically describe the relationships of the MRR between discharge parameters and material properties. Specific discharge energy is defined as the ratio of removed material volume to discharge energy, and its mathematical expression is expressed in Eq. 3, where, d is wire diameter.

$$ \mathrm{SDE}=\frac{\mathrm{MRR}\times d}{Q} $$
(3)

On the basis of Table 4, the specific discharge energy of different carbon steels can be worked out, as shown in Table 5. It can be found that the specific discharge energy of carbon steel with lower carbon content is larger than that of carbon steel with higher carbon content. Besides, the specific discharge energy of the same model of carbon steel under different discharge parameters is also roughly the same. More specifically, the specific discharge energy slowly increases with the increase of discharge energy. This fact may be attributed to that, within a certain range, high discharge energy can generate great discharge explosive force to remove workpiece material.

Table 5 Specific discharge energy of different carbon steels
Full size table

Besides, the relationship between specific discharge energy and material physical properties can be acquired, as indicated in Table 6. Moreover, the mathematical model between specific discharge energy and material physical properties is worked out by the method of polynomial fitting, as shown in Eq. 4. From Table 6 and Eq. 4, it can be found that the specific discharge energy is closely related with material physical properties. In addition, this mathematical model has high fitting precision which can be used to predict the optimal machining feed rate and establish an expert system.

$$ \mathrm{SDE}=101.068-0.00498096\rho -0.031889{T}_{\mathrm{m}}-0.0305418{C}_{\mathrm{p}}+0.0685124k $$
(4)
Table 6 The relationship between specific discharge energy and material physical properties
Full size table

According to Eq.4, the MRR increases with the increase of heat conduction coefficient and the decreasing of melting point, density, and specific heat capacity. It is easy to understand that removing workpiece material with high melting point and larger density needs the more thermal energy. Moreover, in WEDM, the discharge energy density in single pulse discharge process is extremely high because the period time of single pulse ranges from hundreds of nanoseconds to tens of microseconds. The maximal temperature in discharge gap can reach up to 5000–10,000 °C. Workpiece material is removed by way of melting and evaporating. Then, combining the physical properties of workpiece material with the removal mechanism of WEDM, the more workpiece material can be removed when workpiece material is with higher heat conduction and lower specific heat capacity.

According to the concept of SDE, the model of the MRR can be simplified as Eq.5. Furthermore, on the basis of Table 4, the regression equations between process parameters and the MRR of different workpiece materials can be obtained, as shown in Eq.6.

$$ \mathrm{MRR}=\mathrm{SDE}\times \left(a+b\times \mathrm{SV}+c\times I+d\times {T}_{\mathrm{on}}\right) $$
(5)
$$ \Big\{{\displaystyle \begin{array}{l}{\mathrm{MRR}}_{10\#}=-445.82+5.89\mathrm{SV}+21.35I+15.45{T}_{\mathrm{on}}\\ {}{\mathrm{MRR}}_{20\#}=-395.61+5.22\mathrm{SV}+18.94I+13.71{T}_{\mathrm{on}}\\ {}{\mathrm{MRR}}_{35\#}=-373.72+4.93\mathrm{SV}+17.89I+12.95{T}_{\mathrm{on}}\\ {}{\mathrm{MRR}}_{60\#}=-338.22+4.47\mathrm{SV}+16.19I+11.72{T}_{\mathrm{on}}\end{array}} $$
(6)

Combining Eq.5 with Eq.6, the empirical model of the MRR can be expressed in Eq.7.

$$ \mathrm{MRR}=\mathrm{SDE}\times \left(-132.43+1.75\mathrm{SV}+6.34I+4.59{T}_{\mathrm{on}}\right) $$
(7)

In this subsection, a group of verified experiment is implemented to evaluate the precision of the empirical model of the MRR. The 45# steel is selected as workpiece material. The density is 7813 kg/m3, the melting point is 1515 °C, the specific heat capacity is 472 J/kg/K, and the heat conduction coefficient is 47.68 W/m/K. On the basis of Eq.4, the SDE of 45# steel 2.6908 × 10−3 mm3/J. Besides, according to the experiment design of Table 4, the design and results of verified experiment are listed in Table 7. The other parameters are as follows: pulse-off time is 10 μs, wire tension is 15 N, wire speed is 0.15 m/s, dielectric pressure is 0.8 MPa, and workpiece thickness is 20 mm. Moreover, the predicted value of the MRR can be obtained based on the empirical model of the MRR.

Table 7 The experiment and predicted data of the MRR in verified experiment
Full size table

From Table 7, it can be found that the relative error between experiment and predicted value of the MRR are less than 5%. In the other words, this empirical model of the MRR has high precision and reliability.

3.2 Surface quality

Except for material removal rate, the surface quality of workpiece is another important factor to evaluate the machining performance of WEDM. However, recast layer and heat-affected zone (HAZ) is inevitably formed in the subsurface material due to the thermal-removing process of WEDM. In this subsection, the experiments of microstructure observation and microhardness measurement are implemented to illustrate the change of subsurface material. Then, the influence of carbon content and discharge parameters on surface quality is also investigated.

3.2.1 Surface microstructure and microhardness

Surface microstructure

According to metallic material science [26, 27], these five models of carbon steels are hypo-eutectoid steel whose carbon content ranges from 0.0218 to 0.77%. The bulk material of hypo-eutectoid steel consists of ferrite (α-Fe) and pearlite (P). Pearlite (P) is a blend of ferrite (α-Fe) and cementite (Fe3C). The content of pearlite (P) increases with the increase of carbon content, and the exact value of pearlite (P) content can be calculated by lever law. Besides, ferrite (α-Fe) is with low hardness and good plasticity. Pearlite (P) has good comprehensive mechanical properties.

Figure 5 represents the surface microstructure of the five models of carbon steels. The discharge parameters are as follows: gap peak voltage is 45 V, discharge peak current is 10 A, and pulse-on time is 10 μs. It can be clearly seen from Fig. 5 that (1) in subsurface material, along with thickness direction, the material structure is as follows: recast layer, heat-affected zone, and bulk material. (2) In the bulk material, the content of ferrite (α-Fe) and the grain size of ferrite (α-Fe) obviously decrease as the raise of carbon content. (3) There is one thing in common in Fig. 5 (a–e), that is, the grain size of bulk material is larger than that of heat-affected zone. The grain size of heat-affected zone is obviously larger than that of recast layer. (4) Under a light microscope, the image of recast layer is blacker than heat-affected zone and bulk material.

Fig. 5
figure 5

The microstructure of five models of carbon steels by SV 45 V, I 10 A, ton 10 μs a 10#. b 20#. c 35#. d 45#. e 60#

Full size image

In WEDM, most of gas material sprays to discharge gap at very high speed because of high pressure differential, and this part of material can be almost entirely washed out by flushing dielectric. But, only a part of liquid material can be expelled out by flushing dielectric because the discharge gap is very narrow, and the rest will be recrystallized and adhered on workpiece surface again. This recrystallized material is the so-called recast layer. Moreover, a part of subsurface material will be heated to the temperature which is higher than the austenitizing temperature and lower than molting point, and this part of material cannot be removed. But, its phase has been changed due to the fast cooling effect of flushing dielectric, and the material in this zone is named heat-affected zone.

Microhardness

Figure 6 indicates the subsurface material microhardness of the five models of carbon steels. The discharge parameters are as follows: gap peak voltage is 45 V, discharge peak current is 10 A, and pulse-on time is 10 μs. It can be found that the microhardness of the five models of carbon steel has a similar variation trend with the depth below the machining surface. More specifically, the maximal hardness is obtained at the area near machining surface, and it drops rapidly firstly and then increases slowly along with the depth below the machining surface, and finally remains stable at a certain value. On the basis of the formation mechanism of recast layer and heat-affected zone, the main structures of recast layer are martensite and secondary cementite. Besides, there is a good deal of over tempered martensite mixing in heat-affected zone. It is well-known that the hardness of martensite and secondary cementite is higher than pearlite (P) and ferrite (α-Fe). However, the brittleness of secondary cementite is also significantly higher than that of pearlite (P) and ferrite (α-Fe). Moreover, the hardness of over tempered martensite is lower than that of pearlite (P). Furthermore, it can be seen that the maximal value and equilibrium value of carbon steels are positively associated with carbon content. This fact corresponds with the characteristic of carbon steels. In essence, as for hypo-eutectoid steel, the content of pearlite (P) increases with the increase of carbon content. The hardness of pearlite (P) is obviously higher than that of ferrite (α-Fe).

Fig. 6
figure 6

The subsurface material microhardness of five models of carbon steels by SV 45 V, I 10 A, ton 10 μs

Full size image

3.2.2 Surface roughness and recast layer thickness

Figures 7, 8, and 9 suggest the effects of carbon content and discharge parameters on surface roughness and recast layer thickness. It can be observed that (1) the variation trends of surface roughness along with discharge parameters are similar to those of recast layer thickness. This is because surface roughness is the macro-reflection of recast layer thickness. (2) Surface roughness and recast layer thickness show an increasing trend with the increase of discharge peak current and pulse-on time. This fact is attributed to that discharge energy is decided by discharge peak current and pulse-on time. The higher discharge energy will heat the more workpiece material to melting point or boiling point. According to the formation mechanism of recast layer, the higher surface roughness and the thicker recast layer will be produced under higher discharge energy. (3) Surface roughness and recast layer thickness drop with the increase of gap peak voltage. This fact seems to be opposite to the relationship between discharge energy and recast layer thickness. In reality, gap peak voltage is not only a key factor of discharge energy but also a determining factor of cutting kerf width [28,29,30], high gap peak voltage can break down wide cutting kerf. This can effectively improve dielectric circulation, and more discharge debris can be washed out by flushing dielectric.

Fig. 7
figure 7

The influences of carbon content and pulse-on time on surface roughness and recast layer thickness. a Ra. b RLT

Full size image
Fig. 8
figure 8

The influences of carbon content and discharge peak current on surface roughness and recast layer thickness. a Ra. b RLT

Full size image
Fig. 9
figure 9

The influences of carbon content and gap peak voltage on surface roughness and recast layer thickness. a Ra. b RLT

Full size image

Moreover, surface roughness and recast layer thickness firstly increase and then decrease with carbon content. The recast layer thickness of the 35# steel is larger than that of other carbon steels. On one hand, from the angle of melting point, it can be found from Table 6 that the melting point of carbon steel drops along with carbon content. Obviously, the material with low melting point will easily form a recast layer. The recast layer thickness of workpiece increases with the carbon content of carbon steel. On the other hand, from the angle of material removal rate, the material removal rate has a decreasing trend with the rise of carbon content. The faster material removal rate means the more workpiece material is removed by way of evaporating under the same discharge parameters [31,32,33]. As we know, just a small part of evaporated material will form a recast layer because the jet velocity of gas material is very fast. Hence, the recast layer thickness of workpiece decreases with the carbon content of carbon steel. Synthetically considering these two angles, the maximal recast layer thickness of workpiece could be achieved at the 35# steel.

4 Verified experiment of two methods for improving machining characteristics

According to previous researches [34, 35], the recast layer and heat-affected zone have a harmful effect on the application of workpiece by WEDM due to its high brittle and easy-to-form microcrack. So, they should be reduced as possible in the practical machining process. In this section, two methods are proposed to improve machining efficiency and surface quality. The 45# steel is chosen as workpiece material which is one of the most widely used steels in manufacturing industry.

4.1 The optimal machining feed rate according to SDE

In WEDM, there is the achievable material removal rate under certain discharge parameters. In this subsection, the cutting experiments under different machining feed rates are carried out to investigate the application of the empirical model of the MRR. The experiment design and result are shown in Table 8. The MRR and the optimal machining feed rate (Fr) can be calculated by Eqs.7 and 8, respectively. The other two machining feed rates are 0.5*Fr and 2*Fr. The other parameters are as follows: pulse-off time is 10 μs, wire tension is 15 N, wire speed is 0.15 m/s, dielectric pressure is 0.8 MPa, and workpiece thickness is 20 mm.

$$ \mathrm{Fr}=\frac{\mathrm{MRR}}{H} $$
(8)
Table 8 The experiment design and result under different machining feed rates
Full size table

It can be found from Table 8 that (1) the optimal machining feed rate (Fr) can realize the shortest machining time followed by 2*Fr and 0.5*Fr. (2) The optimal machining feed rate (Fr) can obtain the lowest surface roughness followed by 0.5*0 and 2*Fr. This fact is attributed to that (a) when the machining feed rate is 2*Fr, the wire electrode will frequently go back because of short circuit. This phenomenon can result in the next discharge point occurring on the same position. (b) When the machining feed rate is 0.5*Fr, the wire electrode lags behind because the gap voltage is higher than the reference gap voltage. This phenomenon can lead to secondary discharge. These two cases will decrease the actual machining speed and surface roughness.

4.2 Magnetic-assisted trim cutting

From our large number of experiments, it is found that magnetic-assisted method can effectively improve the machining performance in cutting paramagnetic material process. However, this method has no significant effect on machining characteristics in rough cutting ferromagnetic material process. Besides, we also find that the magnetic field–assisted method can obviously improve the machining characteristics in trim cutting ferromagnetic material process when the magnetic field device is appropriately set. The practical map and schematic map of magnetic field–assisted cutting are shown in Fig. 10.

Fig. 10
figure 10

The practical and schematic map of magnetic field–assisted trim cutting. a Practical map. b Schematic map

Full size image

As we know, in discharge-removing process, the melting material will be rapidly solidified to debris due to the cooling effect of flushing dielectric. The flushing effect is not enough to wash all debris out because the discharge gap is very narrow. Then, a part of discharge debris will be left in discharge gap or adhere on machining surface again. This phenomenon can promote the formation of recast layer and result in abnormal discharge state.

When magnetic field is added to assist cutting ferromagnetic material as in Fig. 10b, the discharge debris will be subjected to three types of forces, including Lorentz force, magnetic field attractive force, and flushing dielectric thrust. More specifically, the thrust force from flushing dielectric is in the downward direction. The Lorentz force is in a direction parallel to the thrust from flushing dielectric when the magnetic field generator is set as Fig. 10b. Besides, the direction of magnetic field attractive force is perpendicular to machining surface. It has a negative effect on debris expelling out. Moreover, the magnetic field attractive force obeys to Eq.9, and it is proportional to the derivative of a magnetic induction vector with respect to displacement. In trim cutting ferromagnetic material, magnetic field attractive force can be ignored because of the continuity of ferromagnetic material. But, it is bad for machining in rough cutting ferromagnetic material due to the phenomenon of magnetic focusing.

$$ {F}_{\mathrm{M}}=\frac{\lambda VM}{\mu_0{\mu}_r}\frac{dM}{dx} $$
(9)

where λ is volume susceptibility, V is volume of ferromagnetic material, M is magnetic induction intensity, μ0 is the permeability of vacuum, and μ0 is relative permeability.

Figure 11 depicts the microstructure of the 45# steel by trim cutting without or with magnetic assisted. The machining parameters is as follows: gap peak voltage is 50 V, discharge peak current is 4 A, pulse-on time is 6 μs, and pulse-off time is 10 μs, feed rate is 0.1 mm, and magnetic induction intensity is 0.15 T. Figure 12 suggests the microhardness of the 45# steel under three processing methods. It can be found that: (1) the machining surface by trim cutting is obviously smoother than that by rough cutting. (2) The method of trim cutting can significantly decrease recast layer, and there exists still a certain thickness (about 35 μm) of heat-affected zone in subsurface material when magnetic field is not applied to assist machining. (3) The method of magnetic-assisted trim cutting can further decrease heat-affected zone. The thickness of heat-affected zone can be reduced to less than 20 μm.

Fig. 11
figure 11

The microstructure of 45# steel by trim cutting without or with magnetic assisted. a Without. b With

Full size image
Fig. 12
figure 12

The microhardness of the 45# steel under three processing methods

Full size image

5 Conclusions

This paper investigates and improves machining efficiency and surface quality based on specific discharge energy and magnetic-assisted method. From the above theoretical and experiment analyses, the following conclusions can be drawn:

  1. (1)

    According to the single-factor experiment, the effects of carbon content and discharge parameters on the MRR, Ra, and RLT have been acquired. More specifically, the MRR increases with the decrease of carbon content and the increase of discharge peak current, gap peak voltage, and pulse-on time. Besides, Ra and RLT firstly increase and then decrease with the increase of carbon content, and they have a positive correlation with discharge peak current and pulse-on time.

  2. (2)

    The concept of specific discharge energy (SDE) is proposed to reveal the fundamental influence of workpiece material properties on the MRR. The mathematical model between SDE and the physical properties of workpiece material is also fitted out with high precision.

  3. (3)

    Verified experiment data show that, comparing with other machining feed rates, the optimal machining feed rate according to SDE can realize the fastest actual machining speed and the lowest surface roughness simultaneously. In addition, the method of magnetic-assisted trim cutting can almost eliminate recast layer and reduce heat-affected zone to less than 20 μm.