Plas Ma Radius Prediction

Download as pdf or txt
Download as pdf or txt
You are on page 1of 16

The International Journal of Advanced Manufacturing Technology (2023) 126:2429–2444

https://doi.org/10.1007/s00170-023-11292-6

ORIGINAL ARTICLE

Theoretical and experimental analysis of plasma radius expansion


model in EDM: a comprehensive study
Emmanouil L. Papazoglou1 · Panagiotis Karmiris‑Obratański1,2   · Nikolaos E. Karkalos1 ·
Muthuramalingam Thangaraj3 · Angelos P. Markopoulos1

Received: 10 January 2023 / Accepted: 17 March 2023 / Published online: 22 March 2023
© The Author(s) 2023

Abstract
Electrical Discharge Machining (EDM) is an established non-conventional process, which is particularly efficient for the
processing of hard-to-cut materials, in order to obtain high dimensional accuracy and surface integrity. However, in order to
determine the appropriate parameters for machining novel materials, it is necessary to investigate the EDM process in depth,
both by experiments and numerical models, taking into consideration the fundamental physical phenomena occurring during
this process and be able to predict the surface morphology and microstructural alterations under various conditions. One
of the challenging issues of EDM simulation models that still remain open is the representation of the evolution of plasma
channel radius, for which various approaches have been proposed such as a linear, power law, or a more complex piecewise
relation, in respect to time. Thus, in this work, the effect of different relations for the plasma channel radius evolution on
energy absorption coefficient, plasma flushing efficiency (PFE), and crater morphology is compared under various condi-
tions with a numerical model, which is also compared to experimental data. The results indicate that the energy absorption
coefficient is dependent on the plasma column radius function, as slower growth of plasma channel leads to lower absorption
coefficient and more efficient material removal, whereas a lower variation and different trends under different conditions
were observed regarding PFE values, in respect to the power law exponent. Finally, the crater dimensions were shown to be
consistently more narrow and deeper with higher exponents; thus, based on actual observations of indicative craters, it was
revealed that the appropriate values for the exponent of the power law plasma radius function are below 0.25.

Keywords  EDM · FEM · Plasma channel radius · Power law function · PFE

Nomenclature hmelt Artificial heat transfer coefficient that is utilized


Cp Material’s specific capacity (J/kg K) to calculate the amount of energy that results in
fw The percentage of the spark energy that it is the melting of the material on the eroding front
absorbed by the workpiece (–) (W/(m2 K))
hdiel Heat transfer coefficient (W/(m2 K)) IP Pulse-on current (A)
k Material’s thermal conductivity (W/mK)
LH Material’s latent heat of melting (J/kg)
n Exponent coefficient that is correlated with the
plasma channel growth rate (–)
Q Volumetric heat source (W/m3)
* Panagiotis Karmiris‑Obratański qconv Convection boundary heat flux (W/m2)
[email protected]
qmelt Artificial, specifically defined heat flux that con-
1
School of Mechanical Engineering, Laboratory trols the temperature on the molten front (W/m2)
of Manufacturing Technology, National Technical University qpl (r,t) Plasma channel boundary heat flux (W/m2)
of Athens, Athens, Greece qrad Radiation boundary heat flux (W/m2)
2
Faculty of Mechanical Engineering and Robotics, r The distance from the center of the plasma chan-
Department of Manufacturing Systems, AGH University nel (m)
of Science and Technology, Krakow, Poland
R0pl Maximum plasma radius (m)
3
Department of Mechatronics Engineering, SRM Institute
of Science and Technology, Chennai 603203, India
Rpl (t) The plasma radius over time (m)

13
Vol.:(0123456789)

2430 The International Journal of Advanced Manufacturing Technology (2023) 126:2429–2444

T Temperature (K) the construction of various parts for the automotive,


t Time (s) aerospace, and biomedical industries [12]. Although
Tamb Ambient temperature (K) this process can be successfully carried out in industrial
Tdiel Dielectric fluid temperature (K) practice, its application for advanced or novel materials
Ton Pulse-on time (s) is not straightforward and requires optimization based
Tph.ch. Material’s melting point (K) on both the efficiency of the process, indicated by
υmelt The velocity of the crater’s wall due to the mate- material removal rate (MRR) and tool wear ratio (TWR),
rial erosion (m/s) and on the surface integrity, indicated by surface
V Close circuit voltage (V) roughness or modification of the microstructure [11].
ε Emissivity coefficient of steel (–) However, due to obvious difficulties, both in directly
ρ Material’s density (kg/m3) monitoring the process and determining the procedure
σ(t) The standard deviation of the Gaussian distribu- of the creation of heat affected zone (HAZ) and white
tion (m) layer (WL), reliable simulation models are required in
order to provide a deeper understanding of the complex
phenomena occurring in this process. In fact, modeling
1 Introduction of EDM process is not a trivial task, as the occurrence
of complex thermophysical phenomena should be taken
EDM is one of the most popular non-conventional into consideration. Although simulation models for EDM
machining processes, which involves the use of a large have been developed for several decades, there are issues
number of spark discharges in order to remove material which remain unresolved still nowadays, and thus, EDM
from the workpiece surface. The discharges occur as remains not entirely understood or explained by a universal
a result of a sufficient voltage difference between an model. Furthermore, it is inevitable to adopt reasonable
electrode acting as cathode and the workpiece acting simplifications for the models in order to perform
as anode, which gives rise to a plasma channel of high simulations in a feasible timeframe, given the limitations
energy density [1]. Subsequently, high temperatures of the available computational power, but the accuracy
occur in the gap between anode and cathode, melting or of the simulations should be ensured by optimizing the
ablating the workpiece material [2, 3]. The EDM process computational model parameters based on experimental
is schematically presented. Thus, this process is particularly observations (Fig. 1).
effective during processing of hard-to-cut materials, such As the thermal effects are dominant in EDM, most simu-
as steels [4, 5], titanium [6–8], or nickel alloys [9], as it lation models focus on modeling the effect of heat input
does not involve mechanical contact between a tool and the into the workpiece by using a heat source with a Gaussian
workpiece, and it can process every electrically conductive distribution [13, 14]. This heat source represents a single
material, regardless of its mechanical strength [10]. Various spark, and heat transfer is mainly considered to take place
parameters can affect the efficiency of EDM, such as the by conduction [15, 16]. Then, the calculated MRR result-
pulse-on current, pulse-on time, machining voltage, duty ing from the action of the spark is compared to the experi-
factor, and thermophysical properties of the workpiece mental [13, 17]. Other important issues with the models are
material. Due to the large number of parameters, researchers the determination of energy distribution between electrode
often conduct a considerable amount of experiments for and workpiece, which cannot be considered constant, but
each material and perform optimization studies [11]. dependent on process parameters [18–21], and the deter-
EDM process has been extensively studied during mination of the PFE, which is relevant to the percentage
the last decades, as it has been proven important for of molten material removed from the workpiece by every

Fig. 1  Schematic presentation
of the EDM process

13
The International Journal of Advanced Manufacturing Technology (2023) 126:2429–2444 2431

spark. For PFE, the simplification that it is equal to 100% is determined a formula for the spark column expansion,
definitely unacceptable [13, 21], as it contradicts the forma- based on a differential equation describing spark volume
tion of WL, which is evident by many experimental works. increase and also took into account variable heat transfer
Especially for the plasma column, although some time, depending on the distance of points from the center
authors neglected the variability of its spatial dimension, of the spark. Chu et al. [32] presented a comprehensive
most authors take into account the plasma channel radius methodology for the calculation of plasma column radius,
by semi-empirical formulas or even adopt time-dependent temperature, and pressure including various components,
relations, as it was shown that the variation of the plasma such as a model for breakdown in the dielectric medium,
channel radius can directly affect the simulation results, which takes into consideration the nucleation of bubbles
especially for shorter discharge times [22]. Semi-empirical and production of discharge and models for the initial stages
formulas usually correlate the radius with discharge power of plasma column formation and expansion stage until the
and pulse-on time [2, 23, 24] or pulse-on time and discharge end of each discharge. Dhanik and Joshi [33] also developed
current, but formulas using gap voltage exist as well [25], a model for plasma column, including the nucleation and
although other parameters, such as properties of the elec- growth of bubbles at the first stage and appropriate fluid
trode and dielectric, polarity, and workpiece material, can dynamics and heat transfer models for the heating stage.
also affect it [20]. These formulas are developed based on Shabgard et  al. [34] used a differential equation for
the experimental results after fitting procedure, while direct the calculation of plasma column radius, which took into
measurements of the plasma column dimension, e.g., using consideration the contribution of electric field created by
high-speed cameras and spectrometers, are rare [26, 27]. charged particles to the acceleration of plasma channel, the
These formulas are considerably popular, can be used for contribution of magnetic field created by the movement
a specific range of conditions, and can provide fairly good of the charged particles, the contributions of internal and
approximation of the dimensions of the plasma column, external pressure of plasma channel, and the surface tension
something that is fundamental for the application of heat in the interface between plasma channel and dielectric.
source in the model. However, phenomena such as the rapid Especially for magnetic field-assisted EDM process,
expansion of the plasma column during the first microsec- Shabgard et al. [35] also used a slightly modified model
onds after its creation can only be taken into consideration of their original one, including the effect of the additional
by time-dependent formulas. magnetic field in the differential equation of plasma
Regarding the modeling of plasma column expansion, expansion. Gholipoor et al. [36] compared the results of the
apart from the obvious difficulties in observing this model presented by Shabgard et al. [34] and showed that the
phenomenon directly, there is also a lack of established calculated plasma radius values are slightly higher to those
models for plasma dynamics describing the different states of semi-empirical models.
of plasma column during EDM, i.e., creation, expansion, However, in the majority of relevant works, due to com-
and collapse [28]. Only a few authors have adopted putational power limits, a simpler approach was followed
comprehensive approaches relevant to the time evolution using a single or a piecewise function to model the time
of plasma column. For example, Mujumdar et  al. [28] dependence of plasma column radius. In specific, the fol-
attempted to create a model for plasma column taking into lowing types of functions have been already proposed in
account chemical phenomena, apart from electrical, and the relevant literature: linear, power law, combined power
thermal ones. Using this model, they directly predicted law with semi-empirical and piecewise. For the linear func-
heat flux to the workpiece and electrode under various tion, it can be noted that only a few works have used it,
conditions. Eubank et al. [29] developed a model for the such as the work of Singh and Ghosh [37], who analyzed
calculation of plasma radius, temperature, and pressure plasma formation by a fluid mechanics equation, in order to
evolution using a fluid mechanics equation, an energy establish a thermoelectric model for the calculation of the
balance equation, a radiation equation, and an equation of electrostatic force on the electrode and the stress distribution
state, whereas thermophysical properties were calculated on the workpiece during a discharge. In their work, plasma
by taking into consideration some fundamental reactions radius was assumed to vary linearly with time for small dis-
occurring during the process. Pandey and Jilani [30] charge durations.
adopted an iterative procedure for the determination of On the other hand, the power law function is by far the
plasma channel radius based on the assumption that cathode most popular in the relevant literature. In these works, the
spot temperature is constant during the pulse-on time and assumption made is relevant to the continuous expansion
equal to the boiling point temperature of the electrode of plasma column radius during the discharge time [30].
material. Results were obtained for various electrode Izquierdo et al. [38] argued that modeling of plasma radius
materials, and this approach was considered sufficiently with a power law function including a constant term is in
accurate compared to experimental data. Zhang et al. [31] line with the common assumption that the expansion of

13

2432 The International Journal of Advanced Manufacturing Technology (2023) 126:2429–2444

plasma channel is abrupt for a few microseconds and then accuracy of some of the models was also limited only to
stabilizes. Similar approaches were adopted by Shao and cases with low or high values of discharge current.
Rajurkar [16] and Guo et al. [17]. Kliuev et al. [39] used Liu and Guo [50, 51] developed a plasma radius growth
a similar function with the addition of a second constant model with a piecewise function employing two empirical
term which could be determined by an optimization proce- constants. The first expression, which was a power law func-
dure. Schneider et al. [40] also employed a function with tion in respect to time, was considered valid up to a certain
a second constant term, which was dependent on maxi- critical time, and then the plasma radius was considered con-
mum spark radius, initial radius, and discharge duration. stant. The estimation of empirical constants and critical time
In every case where power law was adopted, the exponent was based on a considerable amount of experimental data
has values below 1.0. In many of these works [16, 17], the [51]. This model reflects some experimental observations,
exponent of the power law function is 0.75; however, it which indicated that plasma is not continuously expand-
is worth noting that an exponent with a value around 0.2 ing during the entire pulse duration [51]. Li et al. [52] used
has also been suggested in the works of Revaz [41] and also a piecewise function, but the function employed for the
Perez [42] based on various experimental observations, in expansion stage was linear in respect to time.
contrast to the works which used larger exponents. More Natsu et al. [27] performed observations with high-
specifically, Revaz et al. [41] commented that the exponent speed camera and confirmed that the expansion of plasma
value of 0.75 corresponds only at the very beginning of is completed after a few microseconds after the breakdown
the discharge and determined that the most suitable value of the dielectric. Thus, they proposed a piecewise model,
for the exponent was close to 0.2. termed as first stage expansion model (FSEM), and com-
Assarzadeh and Ghoreishi [43, 44] noted that the dis- pared it to a power law model which was inferior to the
charge channel radius is dependent on various factors and FSEM regarding the crater dimensions. Izquierdo et al.
utilized a combined semi-empirical/power law relation [53] also compared power law functions to piecewise ones
for the growth of discharge channel radius, including an (FSEM), including a linear part for an extremely brief time
additional term dependent on discharge current. In this period and a constant part. The simulation results were
relation, time was varied between 0 and Ton value, after compared to experimental ones regarding MRR and sur-
which the plasma channel is assumed to collapse, and this face roughness indicators. At first, it was shown that a
relation was considered accurate up to a specific spark FSEM model with an optimal value for the final plasma
energy level. Similar approaches were used by Guo et al. radius can provide an excellent estimation for MRR and
[45] and Xie et al. [46]. Vishwakarma et al. [47] also intro- acceptable error for surface roughness. Thus, they consid-
duced a relation which included other terms rather than ered it necessary to include the first stage of rapid expan-
time, such as the discharge length and discharge power, sion in their model. However, the use of an improper final
as well as empirical constants. The latter approach was value for plasma radius leads to a much higher error level
also mentioned by Kumar et al. [20] and Kansal et al. [48]. for FSAE models, compared to power law ones.
Rajhi et al. [49] conducted a comprehensive comparison From the analysis of works relevant to the modeling
regarding the use of constant equivalent plasma column of plasma column radius in the relevant literature, it can
radius functions, time-dependent radius function, and be observed that there is a lack of a comprehensive com-
time-dependent radius function with a semi-empirical parison of the results produced between different plasma
term. The models which employed a constant radius, even radius expansion functions for various different conditions
if different semi-empirical functions were used, showed based on experimental validation of both crater dimen-
similar results both regarding the highest temperature and sions and microstructure observations. Moreover, the vari-
the time when it was reached, as well as similar tempera- ation of PFE values in respect to different plasma radius
ture variation in respect to time. For the time-dependent expansion functions has also not been considered yet in the
models, although the time to reach the maximum tempera- literature. Thus, in the present work, a numerical model for
ture was similar, significant differences were obtained for the prediction of crater morphology, HAZ, and WL forma-
the temperature profiles. Regarding crater morphology, it tion during EDM is employed in order to comprehensively
was shown that both crater diameter and depth varied for investigate the effect of the plasma column radius expan-
different radius models, and especially for crater diam- sion on the results of the EDM simulations. Three different
eter, considerably different results could be obtained. types of functions are compared, namely, constant, linear,
One model showed shallow crater morphology, whereas and power law functions, and their effect on the geometry
the others predicted a more hemispherically shaped one. of produced craters, energy absorption coefficient, and
Finally, MRR prediction was also found to be dependent PFE is analyzed, based also on actual microscope obser-
on the radius model, with some models overestimating vations of craters produced during EDM of 60CrMoV18-5
MRR and others underestimating its values, whereas the steel under various conditions.

13
The International Journal of Advanced Manufacturing Technology (2023) 126:2429–2444 2433

2 Materials and methods the heat losses due to convection and radiation, and finally
an artificial, specifically defined heat flux that controls the
For the current study, a FEM model was employed to temperature on the molten front. Thus, the boundary heat
simulate the material ablation during a single spark. Tak- flux term of Eq. 1 can be expressed as
ing into the account the fuzzy and inherent chaotic nature
of the process in microscale, the adoption of a “typical-
−k ⋅ ∇T = qpl (r, t) + qconv + qrad + qmelt (2)
average spark” approach is reasonable, justified, and up to The heat flux due to the plasma channel is modeled by
a point necessary. This choice is in line with the literature a Gaussian boundary heat source that it is mathematically
where Klocke et al. [6] deduced that the material removal described as
from a continuous discharge process cannot be studied in
a representative way by single-discharge experiments. The
2
1 − r 2
qpl (r, t) = V ∙ IP ∙ fw ∙ e 2𝜎(t)
(3)
current modeling methodology has already been validated 2𝜋𝜎(t)2
regarding its adequacy to simulate the material removal
mechanism during EDM [54]; thus, in the current study, with V the close circuit voltage in V, I P the pulse-on
and keeping in mind that the main aim is to define how the current in A, fw the percentage of the spark energy that it is
boundary conditions regarding the temporal evolution of absorbed by the workpiece, r the distance from the center
the plasma radius affect the simulation results, the FEM of the plasma channel in m, and σ(t) the standard devia-
model will be discussed only in brief, emphasizing on the tion of the Gaussian distribution that can be expressed as
main topic of the study. 1
The general model assumptions and necessary simpli- 𝜎(t) = R (t) (4)
3 pl
fications are:
with Rpl(t) the plasma radius over time in m.
• All the sparks are assumed to be identical As it is already mentioned in introduction and based on
• Conduction is considered as the only mechanism of the relevant literature review, the estimation of the plasma
heat transfer into the material volume [6] radius consists of an area of active research. It has been
• A Gaussian boundary heat source is utilized to model deduced that among the different semi-empirical equations
the heat flux from the plasma channel to the workpiece for the definition of the plasma radius depending on the
[21, 55] machining parameters, a suitable relationship that can also
• The percentage of the spark energy that it is absorbed be combined with deformed geometry feature is [2, 54]
by the workpiece depends on the machining parameters
R0pl = 0.85 ⋅ 10−3 ⋅ IP0.48 ⋅ Ton
0.35
(5)
• The PFE also depends on the machining parameters
• Deformed geometry feature is employed to model and
with R0pl the maximum radius that the plasma channel
simulate the material erosion mechanism
obtains during the discharge in m and Ton the pulse-on time
• A homogenous and isotropic material with tempera-
in s. Thus, although through Eq. 5 the maximum plasma
ture-dependent thermophysical properties is considered
radius can be estimated, the temporal evolution of the
plasma channel radius is not clear, and hence, it is also not
clear how the temporal growth of the plasma channel
2.1 Governing equations and boundary conditions
affects the obtained results from the simulation. Hence, in
of the FEM model
order for this topic to be fully and in-depth investigated,
in the current study, four different approaches for the
The Fourier’s law that describes the heat transfer due to
plasma radius growth are tested. More specifically, the first
conduction is expressed as
and most common approach is that the plasma channel has
𝜌 ⋅ Cp ⋅ 𝜕T a constant radius for the whole discharge duration, and
}
+ ∇q = Q 𝜕T
𝜕t ⇒ 𝜌 ⋅ Cp ⋅ − ∇(k ⋅ ∇T) = Q thus
q = −k ⋅ ∇T 𝜕t
(1)
Rpl (t) = R0pl (6)
with T the temperature in K, ρ the material density in
[kg/m3], Cp the material specific capacity in J/kg K, k the The three other approaches assume an exponential
material thermal conductivity in W/mK, and Q a volumet- growth of the plasma radius over time but with different
ric heat source (or a heat sink) in W/m3. growth rates, and hence, three different exponents are
For the current model, four boundary heat sources are adopted and tested. The time-dependent plasma radius
considered, namely, the heat flux from the plasma channel, can be expressed as

13

2434 The International Journal of Advanced Manufacturing Technology (2023) 126:2429–2444

method, two distinct process phases are regarded: the 1st


( )n
t
Rpl (t) = R0pl (7) of the material removal and the 2nd of the WL formation.
Ton
During the 1st phase, it is considered that the material on
with n the different exponent that is correlated with the the eroding front remains just over the melting temperature,
plasma channel growth rate which has the following val- whilst the whole molten material is removed. In the 2nd
ues: 0.05, 0.15, and 0.25. These values were chosen in phase, no material removal occurs, the material overheats
accordance with the relevant literature which indicates that (i.e., reaches temperatures above its melting point), and thus,
an appropriate value for the exponent of the power law is the WL is gradually formed. The 1st phase of the material
around 0.2 [35, 36]. However, given that this subject was erosion can be modeled and simulated by employing an arti-
not thoroughly investigated, a wider range of exponents was ficial, specifically defined, boundary heat flux:
adopted. {
The plasma radius as function of time for the four differ-
( ) 0 for T ≤ Tph.ch.
qmelt = hmelt Tph.ch. − T withhmelt
ent plasma radius models is depicted in Fig. 2. It is deemed 109 for T > Tph.ch.
that these four models can adequately describe the different (10)
approaches that have been suggested in the relevant literature with qmelt the heat flux that results in the material melt-
regarding the plasma channel growth during the discharge ing on the eroding front in W/m2 and Tph.ch. the material’s
time. melting point in K. The hmelt (W/m2 K) represents the arti-
ficial heat transfer coefficient that is utilized to calculate the
The heat losses due to convection and radiation are respec- amount of energy that results in the melting of the material
tively calculated as on the eroding front. The qmelt is used in Eq. 11 to define
the material erosion rate, i.e., the eroding velocity of the
(8)
( )
qconv = hdiel T − Tdiel crater’s wall:
qmelt
𝜐melt = ( ( (11)
qrad = 𝜀 T 4 − Tamb (9)
4
( ) ) )
𝜌 Cp T − Tph.ch. + LH
with qconv and qrad the heat fluxes due to convection and with υmelt the velocity of the crater’s wall due to the mate-
radiation respectively in W/m2, hdiel the heat transfer coef- rial erosion in m/s and LH the material’s latent heat of melt-
ficient between the workpiece and the dielectric fluid with ing in J/kg.
value ­105 (W/(m2 K)) [56], and ε = 0.75 the typical emis- During the 2nd phase, zero erosion is considered, and
sivity coefficient of steel. Finally, Tamb and the Tdiel are the thus
ambient temperature and the dielectric temperature respec-
tively that are both considered equal to 293.15 K [K]. hmelt = 0 ⇒ 𝜐melt = 0 (12)
The material melting rate and the material removal rate
The time point for which the hmelt is zeroed (and hence the
(i.e., the erosion rate) are directly correlated through the
υmelt is also zeroed) is when the simulation crater’s volume
energy that is required in order for the eroding front to
equals to the experimentally defined mean crater’s volume.
switch over from solid into liquid. In the current modeling

Fig. 2  The plasma radius as


function of time for the four dif-
ferent adopted approaches

13
The International Journal of Advanced Manufacturing Technology (2023) 126:2429–2444 2435

The fine tuning of the energy absorption coefficient fw, CNC EDM machine, by employing a rectangle copper elec-
which is depending on the machining parameters (see Eq. 3), trode of nominal dimensions 14 × 20 mm. The workpieces
is based on the experimental results about the average white were slices of CALMAX steel, while synthetic hydrocar-
layer thickness (AWLT). More specifically, the simula- bon oil (kerosene) was utilized as dielectric fluid that was
tion results must not only be in line with the experimental properly channeled into the working tank through a flushing
regarding the crater’s volume, but also predict and calculate nozzle with nominal pressure of 0.7 MPa. The close circuit
a similar AWLT. Thus, the fw is defined in such a way that voltage and the duty factor were kept constant at 30 V, and
at the end of the pulse-on time, the formed crater has the 50% respectively, while a 0.5-mm nominal depth of cut was
foreseen volume and the overheated area has the width of the defined. Finally, the machined surfaces were observed by
AWLT. The AWLT is experimentally measured by dividing using a Keyence VHX-7000 optical microscope. For the
the area of the WL with its respective length. The AWLT measuring of the crater dimensions, proper software was
of the simulation results is calculated in a similar way. In employed, i.e., Adobe Photoshop® and ImageJ. The mate-
Fig. 3, the aforementioned measurements for experimental rial thermophysical properties that were used in simulations
and simulation results are depicted. are listed in Table 1, whilst the aforementioned machining
Finally, the PFE, and based on its definition, is calcu- conditions are listed in brief in Table 2.
lated by assuming that during the discharge the entire molten The selection of CALMAX steel for the validation of the
material is removed by the workpiece. Hence, to estimate the model was done primarily based on the CALMAX proper-
PFE, the material erosion rate (i.e., the υmelt) is not zeroed ties and secondary because of the prior authors’ knowhow in
during the pulse duration, and thus, an optimal maximum machining CALMAX with EDM. Namely, CALMAX steel
crater volume is defined. The PFE results from the quotient is characterized by high toughness, good wear resistance, and
of the real crater volume that has been previously calculated polishability, while it is widely utilized in molding and die
to the optimal one. applications. Moreover, EDM is a process that is commonly
The simulations were carried out for four different employed for the machining of CALMAX steel and other
machining parameters’ combinations, while the model tool steels. In addition, authors have carried out an extensive
validation and fine tuning was done based on a previously experimental study regarding the machining of CALMAX
conducted extensive study regarding the machining of steel with EDM, allowing them to have a more comprehen-
60CrMoV18-5 (CALMAX) steel with EDM [11]. More sive view of the relevant topic and the specific behavior of
specifically, the experiments were conducted in an Agie- CALMAX steel during its machining with EDM. More spe-
Charmilles Roboform 350 Sp industrial-type die-sinking cifically, by selecting these specific machining parameters, a

Fig. 3  Measurement and
calculation of the AWLT for
a experimental results and b
simulation results

13

2436 The International Journal of Advanced Manufacturing Technology (2023) 126:2429–2444

wide range of per pulse energies is covered, i.e., from 1.92 boundaries do not affect the obtained results. Triangular mesh
mJ per pulse up to 51 mJ per pulse. This range of machin- of variable density was employed, with a finer discretization on
ing parameters can be reasonably considered representative the top boundary, where the plasma heat source acts and on the
for CALMAX steel since the MRR and the Ra between the vertical boundary that handles the geometry deformation due
aforementioned machining parameters combinations have to material erosion. Finally, a time-dependent fully coupled
been increased approximately by 1096% and 310%, respec- solver, with an absolute tolerance of 0.1 and a relative tolerance
tively. Thus, and keeping in mind that during the planning of 0.01, was utilized for the numerical solution of the model.
of the EDM process, except of the MRR, a number of other
parameters have also to be taken into consideration (e.g., sur-
face quality, WL thickness, surface roughness, tool wear), this 3 Results and discussion
range of machining parameters can be considered realistic,
representative, and with practical value as well. The simulation results for the energy absorption coefficient
Regarding the FEM model parameters, axial symmetry was are listed in Table 3.
used; hence, 2D axisymmetric models were solved. The control Concerning the energy absorption coefficient fw and as
volume height and width were set equal to R0pl and 1.2·R0pl it follows by the obtained data that are plotted in Fig. 4, it
respectively, ensuring an efficient control volume where has an expected behavior (see the average values in Fig. 4)

Table 1  Material properties Material thermophysical properties of 60CrMoV18-5 (CALMAX) Steel


Thermophysical property @Temperature (K) Value

Melting point (K) – 1750


Latent heat of melting (J/kg) 1750 270000
Density < 1750 7664.66–0.1951147∙T–8.666967E-
5∙T2
Specific heat (J/kg K) 293 455
473 525
679 608
Thermal conductivity (W/mK) 473 27
673 32

Table 2  Machining parameters
Machining parameters
No. IP (A) Ton (μs)

1 5 12.8 The machining was carried out with hydrocarbon oil (kerosene) as dielectric fluid and with a copper electrode. For
2 9 25 the specific machining conditions, the close circuit voltage was kept constant at 30 V, and the duty factor was also
constant at 0.5.
3 13 50
4 17 100

Table 3  Simulation results for the energy absorption coefficient


Energy absorption coefficient fw
No. IP (A) Ton (μs) Constant Exponential plasma radius Exponential plasma radius Exponential plasma
plasma radius with 0.05 coefficient with 0.15 coefficient radius with 0.25 coef-
ficient

1 5 12.8 0.072 0.070 0.066 0.064


2 9 25 0.068 0.065 0.061 0.058
3 13 50 0.164 0.156 0.147 0.146
4 17 100 0.119 0.113 0.105 0.102

13
The International Journal of Advanced Manufacturing Technology (2023) 126:2429–2444 2437

Fig. 4  Workpiece power absorp-


tion coefficient in respect to the
machining parameters and the
plasma heat source model

in respect to the different machining parameters. More spe- combination (i.e., difference between the constant plasma
cifically, for low pulse-on current and pulse-on time, low channel radius model and the exponential model with expo-
coefficient values are calculated, while for the higher pulse- nent value of 0.25) is less than 15%, with an average dif-
on current and pulse-on time, the absorption is almost dou- ference of 12.8%. This is reasonable keeping in mind that
bled. This kind of dependence and behavior of the absorp- in fundamental level of physical mechanisms, the whole
tion coefficient from the machining parameter combination process is governed by a power balance; thus, the required
is typical in EDM, and it has been previously recorded for power to melt a specific volume of material cannot signifi-
other alloys and materials as well [1]. In regard to the influ- cantly change, since a similar power source and the same
ence of the plasma model on the absorption coefficient, it physical mechanisms are utilized. Hence, it can be said as a
is deduced that the slower the plasma channel growth is rule of thumb that although the obtained results show a high
considered, the lower absorption coefficient is calculated. level of consistency, the slower the plasma channel expan-
The higher fw is defined for the constant plasma chan- sion is considered by each plasma channel radius model, the
nel radius, while the lower is defined for the exponential lower absorption coefficient is expected to be defined.
model with exponent value of 0.25. This can be reasonably The PFE is directly related to the absorption coefficient,
attributed to the more efficient material removal for higher but with inverse trend in respect to the machining condi-
power densities. The slower plasma channel growth results tions combination. An increase of the fw results in a decrease
in an increased average power density, and thus, the same in PFE and vice versa, implying that the material removal
material removal can be achieved with a lower power, i.e., mechanism becomes less efficient as the workpiece absorbs
lower absorption coefficient. However, it is important to be a higher portion of the spark power. This can also be sup-
noticed that the difference between the highest and the low- ported by the fact that, in general, more intense machining
est fw calculated values for the same machining parameters parameters lead to an increased AWLT, a result of the partial

Fig. 5  Plasma flushing effi-


ciency in respect to the machin-
ing parameters and the plasma
heat source model

13

2438 The International Journal of Advanced Manufacturing Technology (2023) 126:2429–2444

only removal of the molten material from the crater’s area. acts more focused, resulting in the increase of the crater’s
These phenomena and mechanisms have been extensively depth and the simultaneous decrease of the crater’s width.
discussed in some previous studies [1, 11]. Regarding the Additionally, the higher power concentration for the slower
impact of the plasma channel radius model on the calculated plasma channel radius growth and the consequent formation
PFE, and as can be seen from the data in Τable 3 and the of a deep and narrow crater creates a “self-amplified” mech-
respective plot of Fig. 5, the plasma channel radius model anism that further increases the depth of the crater, since
has a minor and vague impact on the calculated PFE, as the discharge energy is “trapped” and acts on the tapered
the observed trends are different for different experimen- crater area.
tal cases. In general, the deviation between the lowest and The aforementioned self-amplified mechanism is con-
the highest PFE values for the same machining parameters’ firmed by the crater’s dimension ratio (i.e., the width to
combination is low, with a maximum deviation of 12.4% for depth ratio) that is presented in Fig. 6. As the plasma chan-
the 13 A and 50 μs and an overall mean value of 8.8%. Thus, nel radius growth rate decreases (i.e., exponential model
as a general conclusion, it can be deduced that the plasma with higher exponent), the crater’s dimension ratio has an
channel radius model does not have a significant effect on exponential increase leading in a drilling-type of crater. This
the calculation of PFE, which is mainly affected and corre- is also clearly inferred by observing the crater geometry for
lated with the absorption coefficient and the AWLT. pulse-on current value of 17 A, pulse-on time of 100 μs, and
The plasma channel radius model has a clear impact on for the different plasma channel models that are presented in
the formed crater geometry. Based on the obtained data that Fig. 7. The craters’ morphology gradually changes from a
are listed in Table 4, it is easily concluded that the slower the shallow and wide crater for constant plasma channel radius
plasma channel growth is considered, a narrower and deeper to a deep and narrow crater for the exponential plasma radius
crater is formed. This is reasonable and can be attributed to with an exponent of 0.25.
the higher average power concentration during the discharge Hence, a reasonable question is raised, regarding the
simulation. Taking into consideration that the slower plasma limit that the exponent may have in order for the simula-
channel growth does not result in a sizeable decrease of the tion results to realistically model the EDM process and
power absorption coefficient, a comparable amount of power the underlying physical mechanisms. Keeping in mind the

Table 4  Simulation results for the plasma flushing efficiency


Plasma flushing efficiency
No. IP (A) Ton (μs) Constant Exponential plasma radius Exponential plasma radius Exponential plasma
plasma radius with 0.05 coefficient with 0.15 coefficient radius with 0.25 coef-
ficient

1 5 12.8 0.055 0.056 0.059 0.059


2 9 25 0.087 0.092 0.094 0.096
3 13 50 0.062 0.066 0.070 0.070
4 17 100 0.084 0.090 0.088 0.087

Fig. 6  Crater dimension ratio


in respect of the machining
parameters and the plasma heat
source model

13
The International Journal of Advanced Manufacturing Technology (2023) 126:2429–2444 2439

Fig. 7  The different crater geometries for IP equal to 17 A and Ton equal to 100 μs for a constant plasma radius, b exponential plasma radius with
0.05 coefficient, c exponential plasma radius with 0.15 coefficient, and d exponential plasma radius with 0.25 coefficient

exponential increase of the crater’s dimension ratio for since an extremely deep and narrow crater is formed. Thus,
higher values of the exponent in the exponential model, it the argument that only exponents less than or close to 0.25
can be justified that the exponent must typically have low should be employed is justified, confirming similar results
value. Higher values of the exponent will result in nonre- in the relevant literature [35, 36].
alistic crater geometries. In order to prove this argument, Finally, to further support the above conclusion and to
although the predicted crater dimensions in the aforemen- provide a robust validation of the model, experimental data
tioned cases seem realistic, an extreme case with a power regarding the craters’ characteristics were utilized. In Fig. 9,
law function with an exponent value of 1.0 for the plasma the machined surfaces for the different machining param-
channel radius was also simulated. In Fig. 8, the crater eters’ combinations are presented. Although it is challeng-
geometry for exponent equal to 1 (i.e., linear increase of ing to define a single crater since the process is chaotic and
the plasma channel radius) is presented, for the lowest dis- there is a superposition of the craters, in Fig. 9, a number of
charge energy (i.e., 5 A and 12.8 μs). Clearly, the morphol- distinct craters were defined. The geometrical characteristics
ogy of the crater is not representative of the EDM process, of those craters are in line with the simulation results that are

Fig. 8  Crater geometry for


IP 5A and Ton 12.8μs and for
exponential plasma radius with
a coefficient value of 1.0

13

2440 The International Journal of Advanced Manufacturing Technology (2023) 126:2429–2444

Fig. 9  Experimental measurements of craters’ dimensions for machining parameters a 5 A–12.8 μs, b 9 A–25 μs, c 13A–50 μs, and d 17 A–100
μs

Table 5  Simulation results for Crater dimensions


the crater dimensions
No. IP (A) Ton (μs) Constant plasma Exponential Exponential Exponential
radius plasma radius plasma radius plasma radius
with 0.05 coef- with 0.15 coef- with 0.25 coef-
ficient ficient ficient
Width Depth Width Depth Width Depth Width Depth

1 5 12.8 32.76 1.30 32.46 1.41 31.6 1.88 31.22 3.34


2 9 25 49.24 2.10 48.08 2.25 46.82 2.88 45.8 4.71
3 13 50 111.06 5.76 108.68 6.61 105.88 9.98 104.68 18.43
4 17 100 130.9 6.08 127.56 6.73 123.54 9.18 122.12 15.62

listed in Table 5. The width and depth of the craters gradu- and form the craters rim are not considered. Hence, a small
ally increase for higher per pulse energies, remaining though and acceptable deviation is expected between the experimen-
always in the range of the simulation values. It is important tal and the simulation results. Thus, as a general conclusion
to notice that the modeling and simulation regards a typical and taking into the account the experimental results as well,
average crater, while the fluid mechanics phenomena that it can be deduced that the presented models for the plasma
take place due to the plasma channel gradients of pressure channel radius are in the realistic range of exponents, and

13
The International Journal of Advanced Manufacturing Technology (2023) 126:2429–2444 2441

Table 6  Comparison between simulation and experimental results for the craters’ width
Craters’ width in μm
No. Constant Exponential plasma Exponential plasma Exponential plasma Average Exper. measured Deviation between
plasma radius with 0.05 radius with 0.15 radius with 0.25 simula- simulation and
radius coefficient coefficient coefficient tion experimental

1 32.76 32.46 31.6 31.22 32.01 38±5 15.8%


2 49.24 48.08 46.82 45.8 47.49 49±9 2.1%
3 111.06 108.68 105.88 104.68 107.58 121±13 10.8%
4 130.9 127.56 123.54 122.12 126.03 134±12 5.7%

they also can be utilized for the further study of the process will occur. Moreover, and although the energy absorption
as well as the crater formation mechanisms, by adopting all coefficient, the plasma flushing efficiency, and the craters’
of the presented plasma channel models under a statistically dimensions as absolute values are expected to differ in
driven distribution. respect to the different alloys, they are also expected to fol-
In Table 6, a comparison between the simulation craters’ low common trends in respect of the machining parameters.
width and the experimentally measured is presented. At first, Thus, in short, it can be said that the deduced conclusions
it has to be noticed that the experimental measuring of the should be considered as a robust guideline and methodology
craters’ width is challenging since craters overlap is taking for further and in-depth study and investigation of the EDM
place, and thus, the definition of a single crater geometry process, and secondarily and with the required caution to be
is very difficult. Additionally, the stochastic and chaotic adopted as absolute values.
nature of the process results to a noticeable deviation in the
craters’ width. However, despite these inherent difficulties
and based on the results of Table 6, it is deduced that the
simulation results are in line with the experimental ones, 4 Conclusions
having a less than 16% deviation. Thus, and as an overall
conclusion, it can be said that the predicted craters are of In this work, a 2D axisymmetrical FE numerical model
comparable width with the experimentally measured ones, was used to determine the suitability of various functions
while the deviation in the values of the experimental results regarding the modeling of plasma column radius expansion
suggests and implies the feasibility of the appropriate utiliza- during die-sinking EDM process. More specifically, for the
tion of different plasma models under a statistically driven modeling of expanding plasma column radius, power law
distribution for more realistic and reliable results. functions, with three different exponents, were compared
At this point, it has also to be mentioned and clarified to a constant radius function. The model was based on the
that the deduced conclusions have both a quantitatively and experimental MRR and WL formation and comparison with
a qualitative interpretation. More specifically, the exact val- experimental results under various discharge current and
ues of the energy absorption coefficient, the plasma flush- pulse-on time values allowed deducing several interesting
ing efficiency, and the craters’ dimensions are defined for conclusions.
the specific alloy (CALMAX steel) and the aforementioned Regarding the values of energy absorption coefficient, it
machining conditions and parameters. It has been proved was found that when a power-law function with high expo-
and established in the relevant literature that different alloys nent values is adopted for the plasma column radius, lower
have different behavior during the EDM process; thus, these values are expected, indicating that when the plasma col-
indexes have to be calculated and determined case by case. umn growth rate is slower, material removal is more efficient
In general, alloys that belong to the same class or have simi- due to higher average power density. Moreover, although the
lar machinability by EDM are expected to have close and difference between the energy absorption coefficient values
comparable behavior; nevertheless, any arbitrary generaliza- determined with constant function, which corresponds to a
tion has to be avoided, especially regarding entirely different power-law function with an exponent of zero and power-law
material classes (e.g., between ferrous alloys and titanium function with the largest exponent, was less than 15%, the
alloys). On the other hand, there are some fundamental and trend of absorption coefficient decrease with exponent value
basic qualitative conclusions that have a far more general increase was clear.
applicability. Namely, and as it has already been mentioned, PFE values showed a minimal variation, namely, below
the exponent in plasma radius calculation formula must have 15%, with power-law exponent increase, and different trends
a low value; otherwise, nonrealistic geometries and results were observed under different process conditions. Thus, it

13

2442 The International Journal of Advanced Manufacturing Technology (2023) 126:2429–2444

can be concluded that the choice of power-law exponent has 3. Jameson EC (2001) Electrical discharge machining. Society of
generally negligible effect on the PFE. Manufacturing Engineers, Michigan
4. Fattahi S, Ullah AS (2022) Optimization of dry electrical discharge
Crater dimensions change considerably with different machining of stainless steel using big data analytics. Procedia CIRP
power-law exponents for the plasma column radius expan- 112:316–321. https://​doi.​org/​10.​1016/j.​procir.​2022.​09.​004
sion function. More specifically, the higher the exponent is, 5. Sen R, Paul S, choudhuri B (2022) Investigation on wire electrical
the lower the width and the higher the depth of the crater discharge machining of AISI 304 stainless steel. Mater Today Proc
62:1210-1214. https://​doi.​org/​10.​1016/j.​matpr.​2022.​04.​458
becomes, with an exponential growth of the depth/width 6. Klocke F, Mohammadnejad M, Holsten M et al (2018) A com-
ratio being also observed, which is unrealistic. These obser- parative study of polarity-related effects in single discharge EDM
vations were attributed to the higher average power con- of titanium and iron alloys. Procedia CIRP 68:52–57. https://​doi.​
centration when the growth rate of plasma column radius is org/​10.​1016/j.​procir.​2017.​12.​021
7. Kumar Baroi B, Kar S, Kumar Patowari P (2018) Electric dis-
lower. In the case of exponent value of 1, implying a linear charge machining of titanium grade 2 alloy and its parametric
increase of plasma column radius, the results are similar study. Mater Today Proc 5:5004–5011. https://​doi.​org/​10.​1016/j.​
to those of drilling, with a high value of depth/width ratio, matpr.​2017.​12.​078
which seems unreasonable. 8. Akıncıoğlu S (2022) Taguchi optimization of multiple perfor-
mance characteristics in the electrical discharge machining of
Based on observations of crater morphology from actual the TIGR2. Facta Univ Ser Mech Eng 20:237. https://​doi.​org/​
experiments, it was found that the average dimensions of the 10.​22190/​FUME2​01230​028A
craters are similar to those predicted from the simulations, 9. NAS E, AKINCIOĞLU S (2019) Kriyojenik İşlem Görmüş
something that further proves that the choice of exponent Nikel Esaslı Süperalaşımın Elektro-Erozyon İşleme Performansı
Optimizasyonu. Acad Platf J Eng Sci 7:1–1. https://​doi.​org/​10.​
values below 0.25 is reasonable. This suggestion is particu- 21541/​apjes.​412042
larly useful for further investigations of the complex, multi- 1 0. Fonda P, Wang Z, Yamazaki K, Akutsu Y (2008) A funda-
physics phenomena occurring during EDM. mental study on Ti-6Al-4V’s thermal and electrical proper-
ties and their relation to EDM productivity. J Mater Process
Technol 202:583–589. https://​d oi.​o rg/​1 0.​1 016/j.​j matp​r otec.​
Author contribution  M.L.P., P.K.O., and N.E.K. carried out the experi- 2007.​0 9.​0 60
ments and formal analysis of the original draft, while M.T. and A.P.M. 11. Karmiris-Obratański P, Papazoglou EL, Leszczyńska-Madej B
as supervisors discussed the methods, conclusions, and provide the et al (2022) An optimalization study on the surface texture and
final review and editing. Finally, all the authors reviewed the paper. machining parameters of 60CrMoV18-5 steel by EDM. Materi-
als (Basel) 15:3559. https://​doi.​org/​10.​3390/​ma151​03559
12. Abbas NM, Solomon DG, Bahari F (2007) A review on current
Declarations  research trends in electrical discharge machining (EDM). Int
J Mach Tool Manuf 47:1214–1228. https://​doi.​org/​10.​1016/j.​
Ethical approval  Not applicable. ijmac​htools.​2006.​08.​026
13. Jithin S, Raut A, Bhandarkar UV, Joshi SS (2018) FE modeling
Conflict of interest  The authors declare no competing interests. for single spark in EDM considering plasma flushing efficiency.
Procedia Manuf 26:617–628. https://​doi.​org/​10.​1016/j.​promfg.​
Open Access  This article is licensed under a Creative Commons Attri- 2018.​07.​072
bution 4.0 International License, which permits use, sharing, adapta- 14. Tang J, Yang X (2016) A thermo-hydraulic modeling for the
tion, distribution and reproduction in any medium or format, as long formation process of the discharge crater in EDM. Procedia
as you give appropriate credit to the original author(s) and the source, CIRP 42:685–690. https://​doi.​org/​10.​1016/j.​procir.​2016.​02.​302
provide a link to the Creative Commons licence, and indicate if changes 15. Shabgard MR, Seyedzavvar M, Oliaei SB, Ivanov A (2011) A numer-
were made. The images or other third party material in this article are ical method for predicting depth of heat affected zone in EDM pro-
included in the article's Creative Commons licence, unless indicated cess for AISI H13 tool steel. J Sci Ind Res (India) 70:493–499
otherwise in a credit line to the material. If material is not included in 16. Shao B, Rajurkar KP (2015) Modelling of the crater formation
the article's Creative Commons licence and your intended use is not in micro-EDM. Procedia CIRP 33:376–381. https://​doi.​org/​10.​
permitted by statutory regulation or exceeds the permitted use, you will 1016/j.​procir.​2015.​06.​085
need to obtain permission directly from the copyright holder. To view a 17. Guo YB, Klink A, Klocke F (2013) Multiscale modeling of sink-
copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. ing-EDM with Gaussian heat flux via user subroutine. Procedia
CIRP 6:438–443. https://​doi.​org/​10.​1016/j.​procir.​2013.​03.​047
18. Khatter CP, Pandey OP (2009) Analysis of energy distribution
during electric discharge machining of tungsten carbide. Tribol
References Mater Surf Interfaces 3:2–15. https://​doi.​org/​10.​1179/​17515​
8308X​383198
19. Oßwald K, Schneider S, Hensgen L et al (2017) Experimental
1. Papazoglou EL, Karmiris-Obratański P, Leszczyńska-Madej B, investigation of energy distribution in continuous sinking EDM.
Markopoulos AP (2021) A study on electrical discharge machin- CIRP J Manuf Sci Technol 19:36–43. https://​doi.​org/​10.​1016/j.​
ing of titanium grade2 with experimental and theoretical analysis. cirpj.​2017.​04.​006
Sci Rep 11:1–21. https://​doi.​org/​10.​1038/​s41598-​021-​88534-8 20. Kumar A, Bagal DK, Maity KP (2014) Numerical modeling of wire
2. Jahan MP (2015) Electrical discharge machining (EDM): types, electrical discharge machining of super alloy inconel 718. Procedia
technologies and applications. Nova Science Publishers, New York Eng 97:1512–1523. https://​doi.​org/​10.​1016/j.​proeng.​2014.​12.​435

13
The International Journal of Advanced Manufacturing Technology (2023) 126:2429–2444 2443

21. Shabgard M, Ahmadi R, Seyedzavvar M, SNB O (2013) Math- 38. Izquierdo B, Sánchez JA, Plaza S et al (2009) A numerical model
ematical and numerical modeling of the effect of input-parameters of the EDM process considering the effect of multiple discharges.
on the flushing efficiency of plasma channel in EDM process. Int Int J Mach Tools Manuf 49:220–229. https://​doi.​org/​10.​1016/j.​ijmac​
J Mach Tools Manuf 65:79–87. https://​doi.​org/​10.​1016/j.​ijmac​ htools.​2008.​11.​003
htools.​2012.​10.​004 39. Kliuev M, Florio K, Akbari M, Wegener K (2019) Influence of
22. Zhang Y, Liu Y, Shen Y et al (2014) A novel method of determin- energy fraction in EDM drilling of Inconel 718 by statistical analy-
ing energy distribution and plasma diameter of EDM. Int J Heat sis and finite element crater-modelling. J Manuf Process 40:84–93.
Mass Transf 75:425–432. https://d​ oi.o​ rg/1​ 0.1​ 016/j.i​ jheat​ masst​ rans​ https://​doi.​org/​10.​1016/j.​jmapro.​2019.​03.​002
fer.​2014.​03.​082 40. Schneider S, Herrig T, Klink A, Bergs T (2022) Modeling of the
23. Singh H (2012) Experimental study of distribution of energy dur- temperature field induced during electrical discharge machining.
ing EDM process for utilization in thermal models. Int J Heat CIRP J Manuf Sci Technol 38:650–659. https://​doi.​org/​10.​1016/j.​
Mass Transf 55:5053–5064. https://​doi.​org/​10.​1016/j.​ijhea​tmass​ cirpj.​2022.​05.​012
trans​fer.​2012.​05.​004 41. Revaz B, Witz G, Flükiger R (2005) Properties of the plasma channel
24. Ming W, Zhang S, Zhang G et al (2022) Progress in modeling of in liquid discharges inferred from cathode local temperature measure-
electrical discharge machining process. Int J Heat Mass Transf ments. J Appl Phys 98:1–7. https://​doi.​org/​10.​1063/1.​21374​60
187. https://​doi.​org/​10.​1016/j.​ijhea​tmass​trans​fer.​2022.​122563 42. Pérez R, Carron J, Rappaz M et al (2007) Measurement and metal-
25. Padhi BN, Choudhury SK, Janakarajan R (2021) Influence of energy lurgical modeling of the thermal impact of EDM discharges on steel.
parameters on plasma radius, energy fraction and plasma flushing 15th Internat Conf Electromachining, pp 17–22
efficiency for a single discharge in EDM. Preprint. https://​doi.​org/​ 43. Assarzadeh S, Ghoreishi M (2017) Prediction of root mean square
10.​21203/​rs.3.​rs-​362696/​v1 surface roughness in low discharge energy die-sinking EDM process
26. Qin L, Liu C, Zhang R et al (2022) Characterization of plasma dis- considering the effects of successive discharges and plasma flushing
charge channel in EDM using photography and spectroscopy. Proce- efficiency. J Manuf Process 30:502–515. https://​doi.​org/​10.​1016/j.​
dia CIRP 113:184–189. https://​doi.​org/​10.​1016/j.​procir.​2022.​09.​129 jmapro.​2017.​10.​012
27. Kojima A, Natsu W, Kunieda M (2008) Spectroscopic measure- 44. Assarzadeh S, Ghoreishi M (2017) Electro-thermal-based finite ele-
ment of arc plasma diameter in EDM. CIRP Ann - Manuf Technol ment simulation and experimental validation of material removal
57:203–207. https://​doi.​org/​10.​1016/j.​cirp.​2008.​03.​097 in static gap singlespark die-sinking electro-discharge machining
28. Mujumdar S, Curreli D, Kapoor SG, Ruzic D (2014) A model of process. Proc Inst Mech Eng Part B J Eng Manuf 231:28–47. https://​
micro electro-discharge machining plasma discharge in deionized doi.​org/​10.​1177/​09544​05415​572661
water. J Manuf Sci Eng Trans ASME 136:1–12. https://​doi.​org/​10.​ 45. Guo C, Wang Y, Wang Z, Xie B (2014) A coupled thermal-structural
1115/1.​40262​98 finite element analysis of a single pulse micro-EDM process. Key
29. Eubank PT, Patel MR, Barrufet MA, Bozkurt B (1993) Theoretical Eng Mater 579-580:86–90. https://​doi.​org/​10.​4028/​www.​scien​tific.​
models of the electrical discharge machining process. III. the vari- net/​KEM.​579-​580.​86
able mass, cylindrical plasma model. J Appl Phys 73:7900–7909. 46. Xie BC, Wang YK, Wang ZL, Zhao WS (2011) Numerical simula-
https://​doi.​org/​10.​1063/1.​353942 tion of titanium alloy machining in electric discharge machining
30. Pandey PC, Jilani ST (1986) Plasma channel growth and the resolidi- process. Trans Nonferrous Met Soc China (English Ed) 21:s434–
fied layer in edm. Precis Eng 8:104–110. https://​doi.​org/​10.​1016/​ s439. https://​doi.​org/​10.​1016/​S1003-​6326(11)​61620-8
0141-​6359(86)​90093-0 47. Vishwakarma UK, Dvivedi A, Kumar P (2012) FEA modeling of
31. Zhang F, Gu L, Hu J, Zhao W (2016) A new thermal model consid- material removal rate in electrical discharge machining of Al6063/SiC
ering TIE of the expanding spark for anode erosion process of EDM composites. Int Conf Mech Ind Manuf Eng (ICMIME, 2012), Zurich,
in water. Int J Adv Manuf Technol 82:573–582. https://​doi.​org/​10.​ Switzerland 6:586–591. https://​doi.​org/​10.​5281/​zenodo.​10826​63
1007/​s00170-​015-​7389-3 48. Kansal HK, Singh S, Kumar P (2008) Numerical simulation of
32. Chu X, Zhu K, Wang C et al (2016) A study on plasma channel powder mixed electric discharge machining (PMEDM) using finite
expansion in micro-EDM. Mater Manuf Process 31:381–390. element method. Math Comput Model 47:1217–1237. https://​doi.​
https://​doi.​org/​10.​1080/​10426​914.​2015.​10594​45 org/​10.​1016/j.​mcm.​2007.​05.​016
33. Dhanik S, Joshi SS (2005) Modeling of a single resistance capaci- 49. Rajhi W, Kolsi L, Abbassi R et al (2022) Estimation of MRR and
tance pulse discharge in micro-electro discharge machining. J Manuf thermal stresses in EDM process: a comparative numerical study.
Sci Eng 127:759–767. https://​doi.​org/​10.​1115/1.​20345​12 Int J Adv Manuf Technol 121:7037–7055. https://​doi.​org/​10.​1007/​
34. Shabgard MR, Gholipoor A, Mohammadpourfard M (2018) Numer- s00170-​022-​09806-9
ical and experimental study of the effects of ultrasonic vibrations of 50. Liu JF, Guo YB (2016) Modeling of white layer formation in elec-
tool on machining characteristics of EDM process. Int J Adv Manuf tric discharge machining (EDM) by incorporating massive random
Technol 96:2657–2669. https://​doi.​org/​10.​1007/​s00170-​017-​1487-3 discharge characteristics. Procedia CIRP 42:697–702. https://​doi.​
35. Shabgard M, Farzaneh S, Gholipoor A (2017) Investigation of the org/​10.​1016/j.​procir.​2016.​02.​304
surface integrity characteristics in wire electrical discharge machin- 51. Liu JF, Guo YB (2016) Thermal modeling of EDM with progression
ing of Inconel 617. J Brazilian Soc Mech Sci Eng. https://​doi.​org/​ of massive random electrical dischargeS. Procedia Manuf 5:495–
10.​1007/​s40430-​016-​0556-0 507. https://​doi.​org/​10.​1016/j.​promfg.​2016.​08.​041
36. Gholipoor A, Shabgard MR, Mohammadpourfard M (2020) 52. Li S, Yin X, Jia Z et al (2020) Modeling of plasma temperature
A novel approach to plasma channel radius determination and distribution during micro-EDM for silicon single crystal. Int J
numerical modeling of electrical discharge machining process. Adv Manuf Technol 107:1731–1739. https://​doi.​org/​10.​1007/​
J Brazilian Soc Mech Sci Eng 42:1–10. https://​doi.​org/​10.​1007/​ s00170-​020-​05135-x
s40430-​020-​2244-3 53. Izquierdo B, Sánchez JA, Ortega N et al (2011) Insight into funda-
37. Singh A, Ghosh A (1999) A thermo-electric model of material mental aspects of the EDM process using multidischarge numerical
removal during electric discharge machining. Int J Mach Tools Manuf simulation. Int J Adv Manuf Technol 52:195–206. https://​doi.​org/​
39:669–682. https://​doi.​org/​10.​1016/​S0890-​6955(98)​00047-9 10.​1007/​s00170-​010-​2709-0

13

2444 The International Journal of Advanced Manufacturing Technology (2023) 126:2429–2444

54. Papazoglou EL, Markopoulos AP, Papaefthymiou S, Manolakos DE 56. Rajendran R, Vendan SP (2014) Study of workpiece thermal pro-
(2019) Electrical discharge machining modeling by coupling thermal file in electrical discharge machining (EDM) process. Int J Eng Res
analysis with deformed geometry feature. Int J Adv Manuf Technol Technol 3. https://​doi.​org/​10.​17577/​IJERT​V3IS1​1119
103:4481–4493. https://​doi.​org/​10.​1007/​s00170-​019-​03850-8
55. Tlili A, Ghanem F, Ben SN (2015) A contribution in EDM simula- Publisher’s note Springer Nature remains neutral with regard to
tion field. Int J Adv Manuf Technol 79:921–935. https://​doi.​org/​10.​ jurisdictional claims in published maps and institutional affiliations.
1007/​s00170-​015-​6880-1

13

You might also like