Precision Engineering 65 (2020) 156–163

Contents lists available at ScienceDirect

Precision Engineering
journal homepage: http://www.elsevier.com/locate/precision

Micro-EDM milling of zirconium carbide ceramics

G. D’Urso a, C. Giardini a, S. Lorenzi b, M. Quarto a, *, D. Sciti c, L. Silvestroni c
a
University of Bergamo, Department of Management, Information and Production Engineering, 24044, Dalmine, Italy
b
University of Bergamo, Department of Engineering and Applied Sciences, 24044, Dalmine, Italy
c
CNR-ISTEC, Institute of Science and Technology for Ceramics, 48018, Faenza, Italy

A R T I C L E I N F O A B S T R A C T

Keywords: Surface characteristics and machining performances of micro-pockets manufactured by μEDM as a function of
Micro-EDM different process conditions are studied in this paper. The micro-pockets were obtained using different combi­
Milling nations of process parameters on different ultra-high temperature ceramics (UHTCs) with the same base matrix
Carbides
(ZrC) and different volume fraction of the second phase (MoSi2). This work presents an analysis of the process
performances with different machining approaches, from pre-roughing down to fine-finishing. Microstructure
analysis of the as-sintered and machined materials was conducted to identify materials modification due to the
electrical discharges. A removing material mechanism during the μEDM process has been hypothesized, corre­
lating the ED-machined surface structure and the main process performances.

1. Introduction Amongst UHTCs, zirconium carbide, ZrC, finds industrial importance

because it is suitable for many applications such as coating nuclear
The industrial demand for advanced ceramics is continuously particle fuels [9]. It is one of the most refractory transition metal car­
increasing. Thanks to their constant ameliorating performances they bides and it is not only harder (e 25.5 GPa) and stiffer (elastic modulus e
represent a good choice for several applications in extreme environ­ 392 MPa) than the common advanced ceramics (Al2O3, ZrO2, etc.), but it
ments. Ceramics are employed in automotive, electrical, electronic is also more refractory (melting point e 3420 � C) and a better thermal (e
fields and for aerospace and propulsion applications. All these sectors 40 W/m� C) and electrical (e 106 S/m) conductor [4,10]. Despite these
require new materials that can perform in oxidizing or corrosive atmo­ excellent properties, materials based on ZrC have been hardly developed
spheres and sometimes over the course of a long working life. on an industrial scale due to the high cost of the raw materials, the
In particular, ultra-high temperature ceramics (UHTCs) are a class of complex aspects related to both processing, sintering, and machining.
promising materials used in extreme chemical and thermal environ­ Notwithstanding the above mentioned attracting properties, the use
ments associated with hypersonic flights, such as sharp leading edges of single-phase ZrC materials is not suitable for high-temperature
hot structures of the future generation of slender-shaped re-entry vehi­ structural applications owing to their poor oxidation resistance. Thus,
cles [1]. UHTCs have been studied as candidate materials for the many additives have been evaluated to eliminate the porosity in the
manufacturing of structural thermal protection systems and recently sintered pieces, through the formation of liquid phases, but especially to
have been considered as absorber materials in concentrating solar power improve their resistance to oxidation [11]. Amongst the additives able to
plants [2]. Their use in such severe conditions is possible thanks to their improve ZrC performances, SiC [12], TaSi2 [13] and MoSi2 [11] in
high melting point, good ablation resistance, excellent hardness and different volume percentage [14,15] have been tested. Assessed
strength at elevated temperatures [3]. This group of ceramic compounds methods for preparing dense UHT ceramics in bulk form with desirable
includes borides, carbides, and nitrides of transition metals (e.g. ZrB2, microstructure and properties have been recently reviewed [16].
HfB2, ZrC, HfC, TaC, HfN) with interstitial characteristics that lead to a Although sintering aids enable to reduce the processing temperatures
combination of metallic, covalent and ionic bonds. Carbides [4–6] as and pressures, the main limitations that prevent a widespread use in
such, display the physical properties of ceramics and the electronic industrial applications are related to the difficulties in processing and
properties of metals, with resulting performances going beyond the the high costs associated with their manufacture.
limits of metals [7,8]. One of the main requirements of ceramic materials machining is the

* Corresponding author. Via Pasubio 7/b, 24044, Dalmine, Bergamo, Italy.

E-mail address: [email protected] (M. Quarto).

https://doi.org/10.1016/j.precisioneng.2020.06.002
Received 5 August 2019; Received in revised form 10 March 2020; Accepted 1 June 2020
Available online 15 June 2020
0141-6359/© 2020 Elsevier Inc. All rights reserved.
G. D’Urso et al. Precision Engineering 65 (2020) 156–163

obtainment of fine shapes without micro-cracks, which is very difficult Table 1

to be addressed by traditional mechanical machining technologies, Nominal and final composition with density of ZrC-based conductive ceramic
owing to the intrinsic brittleness of this class of ceramics. Effective so­ materials[5].
lutions to shape these fragile materials are represented by the abrasive Label Nominal Density Final Relative Porosity
processes, such as ultrasonic machining and abrasive waterjet, and the composition [g/cm3] composition density evaluated by
thermal processes, such as laser ablation and electrical discharge [%] SEM [%]

machining. Among all these alternatives, in this study, μEDM was Z ZrC 4.81 ZrC 73.0 ~27
selected since it can prove to be very effective and advantageous when it ZM5 ZrC þ 5 5.52 ZrC þ 6 SiC 83.4 <1
MoSi2 þ 2 MoSi2
comes to machine complex 3D shapes [17–19]. μEDM is a
ZM10 ZrC þ 10 6.11 ZrC þ 15 SiC 92.7 <1
non-conventional technology characterized by contactless machining, in MoSi2 þ 5 MoSi2
fact, the material removal mechanism is performed by means of rapid
electrical discharges developed in the gap between the electrode and the
workpiece: the sparks cause melting and vaporizing of the material, in for increasing the final density. ZrC mixtures were prepared using
the constant presence of the dielectric. The continuous flushing of the commercial powders (H. C. Starck, Germany, Grade B, mean particle
dielectric in the working area allows to keep the electrical conditions size: 3.5 μm; particle size range 0.8–6 μm; specific surface area 3 m2/g;
under control and, at the same, time, it helps keeping the area clean from impurities (wt%) C 0.10, O 0.38, N < 0.01, Hf 1.55) and tetragonal
debris. Since these considerations, the μEDM is usually considered an MoSi2 (Aldrich, Germany, mean particle size 1 μm; particle size range
optimal solution for machining conductive materials (in particular 0.1–3 μm; specific surface area 1.60 m2/g; impurities: O 1) as sintering
metals) characterized by high hardness, high fragility and in general for aid, in amount of 5 vol% and 10 vol%. The powders were ball milled for
all those materials defined as difficult-to-cut due to their thermal, me­ 24 h in absolute ethanol with SiC milling media. Then, the slurries were
chanical and physical characteristics. Strengths of μEDM are related to dried in a rotating evaporator and the powders sieved through a 60-
its ability to machine difficult to cut materials avoiding the transmitting mesh screen. Hot pressing of all materials was conducted in a low vac­
of stresses and vibration to the workpiece since the contactless nature of uum using induction-heated graphite die with a constant uniaxial
the process [20,21]. Furthermore, it allows to obtain very complex pressure equal to 30 MPa. The maximum temperature was 1930 � C for
shape with high level of accuracy and, as a function of the selected all materials, held for 40 min, free cooling to room temperature fol­
process parameters, it is possible to obtain high level of surface finishing. lowed. The bulk density of the materials was measured by Archimedes’
μEDM is usually an expensive process, for this reason its application is method and it was used for the estimation of the percentage of relative
involved in fields where the products have a high value such as the density as the ratio between the measured density and theoretical one
precision mechanics, aerospace, medical, watchmaking and so on [22, (Table 1). The porosity of all materials was also ascertained by scanning
23]. Over the past years, the advances in this technology allow the electron microscope.
application of this process to the manufacture of conductive ceramic
materials [24,25]. The sparks generated during this process produce 2.2. Microstructure characterization
heat that melts and vaporizes the workpiece material that is removed
from the surface through the dielectric medium without contact be­ The microstructure of materials was analysed before and after the
tween the tool and the workpiece, thus reducing the probability of μEDM process. Metallographic sections of as produced ceramic speci­
damage the workpiece. The application of μEDM for machining this class mens and after machining were obtained. The specimens were ground
of materials could be very remarkable in the manufacturing of by means of emery paper up to 4000 grit and then they were polished
complex-shaped or hollow components. One of the most interesting with 1 μm diamond solution. The specimens were then observed by
applications could be the nose tip of the hypersonic atmospheric Scanning Electron Microscope associated with Energy Dispersive X-Ray
re-entering capsule, like SHARK which is featured by a micro-hole in the Spectroscopy (Sigma, Zeiss NTS Gmbh, Germany; and INCA Energy 300,
inner part where a thermocouple is placed to be protected by the ground Oxford Instruments, Abington, UK). Microstructure modifications were
impact [26]. This kind of components requires the use of advanced ce­ investigated after each machining step, from pre-roughing to fine-
ramics or advanced materials for critical environment applications in finishing, at the bottom/centre of the micro-pocket. The surface finish­
combination with features characterized by strict tolerances achievable ing was then evaluated in terms of the arithmetic average of roughness
only by μEDM machining. (Ra) by means of a Form Talysurf S21 PGI 420 (Taylor Hobson) system
A previous study [24] investigated the differences between with an interferometric sensor. The profile measurements were per­
ED-machined surfaces of stainless steel (AISI 316L) and Zirconium formed taking into account 5 profiles, such as indicated by the inter­
Carbide with MoSi2. That work showed completely different textures of national standard UNI EN ISO 4288:2000 along x-axis and an evaluation
the ED-machined surfaces. In particular, metal surfaces were charac­
terized by a smoother profile than the advanced ceramic. Thus, based on
Table 2
these results it was decided to deepen the analysis involving the process Process parameters [8].
performances and the material removal mechanism. In particular,
Machining Pre- Roughing Finishing Fine-
machining performances and the material characterization of
conditions Roughing Finishing
micro-pockets manufactured by μEDM as a function of different
Pulse Shape Long Long braked Short Very short
machining conditions are presented in the present work. The aim is the
Pulses pulses Pulses pulses
identification of possible different material removal mechanisms related Polarity -(neg) -(neg) -(neg) -(neg)
to the unique structure of the advanced ceramics selected for the Width [μs] 6.6 5 5 2
investigation with a focus on the presence of a secondary phase in the Frequency [kHz] 90 130 140 180
workpiece structure. Current [indexa] 80 50 100 100
Voltage [V] 150 130 100 90
Incremental depth 0.0075 0.005 0.001 0.0008
2. Materials and methods [mm]
Energy [indexa] 365 206 105 13
2.1. Materials preparation Gain 1000 1500 700 400
a
the value of the peak current is indicated as an index since this is the format
UHTCs plates were used as a workpiece, considering a single-phase used by Sarix machine. The machine has an autoregulating system and the
bulk material and two different volume fractions of the second phase instantaneous values are automatically set.

157
G. D’Urso et al. Precision Engineering 65 (2020) 156–163

2.4. Criteria for process performance evaluation

The μEDM machine generated a data report at the end of each

machining step. This report contains information about erosion time and
electrode wear recorded during the micro-pockets realization. Material
Removal Rate (MRR) and Tool Wear Ratio (TWR) were selected as
performance indicators for the micro-pockets machining.
MRR [mm3/s] was calculated as the rate of material removed from
the workpiece (MRW [mm3]), estimated as the volume of the designed
micro-pocket, with respect to the erosion time t [s] recorded by the
μEDM system (Eq. (1)).
MRR ¼ MRW=t (1)
TWR (Eq. (2)) was calculated as the ratio between the material
removed from the electrode (MRT [mm3] (Eq. (3)) and the material
removed from the workpiece so that this parameter is dimensionless.
The tool wear (TW), in terms of length, is measured through a touching
procedure executed in a reference position: the length of the electrode is
taken before and after the single milling operation. Specifically, during
Fig. 1. Geometry of the reference micro-pocket [7]. the machining set-up, the operator defined the starting length of the tool
and the control position. The control point was defined on a portion of
length equal to 0.75. the surface not involved in the machining process. This point was used as
a reference point for all the control touch, which were performed at the
2.3. μED-machining end of single layer removal process. After removing the single layer, the
tool was moved to the control point and through a control touch the
The experimental tests were carried out in analogy to what is re­ “new” length was measured and recorded in the report. In the automatic
ported in previous work [24], varying several process parameters, report were reported the cumulative tool wear [22]. Where Del is the
strictly related to the different machining conditions which identify electrode diameter.
different finishing levels of the machined surface, and consequently the
TWR ¼ MRT=MRW (2)
pulse type (in particular the width of a single pulse) used during the
process. Three runs were performed for each combination of workpiece
MRT ¼ πTWDel =4 (3)
material and machining conditions. The optimal process parameters,
presented in Table 2, were selected performing preliminary tests based
3. Results and discussion
on electrode characteristics (material and geometry), workpiece mate­
rial and machining conditions. Each machining condition is character­
3.1. Initial microstructure
ized by ranges in which the parameters can vary and, within these,
different combinations have been tested in order to limit the short cir­
Characterization of the as-sintered microstructure by Scanning
cuits and to optimize the material removal rate. The fundamental
Electron Microscope coupled with Energy Dispersive X-Ray Spectros­
principle to erode features by micro-EDM is layer by layer erosion, for
copy (SEM-EDS) was performed. The backscattered scanning electron
this reason in EDM milling, the incremental depth is one of the main
microscope images of the as-polished surfaces are shown in Fig. 2. The
process parameters as reported in Table 2. This value is related to the
final density of monolithic ZrC was about 73% of the theoretical one; in
different combination of workpiece and electrode materials, and the
fact, the microstructure was partially porous, but spatially uniform and
pulses type applied in the process. Once a specific set of parameters is
consisted of rounded equiaxed grains with mean grain dimensions
defined, the incremental depth is supposed to be constant for the entire
around 1.3 μm (Fig. 2a). About 27 vol% of closed porosity with pore
process.
dimensions in the range 0.3–1 μm was observed. The discrepancy be­
Tungsten carbide cylindrical electrodes having a diameter equal to
tween the final relative density and the residual porosity is related to the
0.3 mm were used. Hydrocarbon oil was used as the dielectric fluid. A
presence of a little amount of graphite, due to the free carbon present in
simple micro-pocket was chosen as a reference shape for the μEDM
the starting powder, 0.1 wt%.
milling tests with the dimensions in a range of 1–1.5 mm and a depth
The addition of MoSi2 notably decreased the final porosity in the
equal to 0.2 mm. The geometry of the microfeature realized in the ex­
sintered composite materials. An example of the polished surface is
periments is reported in Fig. 1. The experiments were performed by
shown in the scanning electron microscope images of Fig. 2b and c. The
means of a SARIX SX-200 μEDM milling machine.
bright phase corresponds to ZrC grains, which have a mean grain size of
1.7 μm and a rounded equiaxed shape; the dark agglomerates (in

Fig. 2. SEM backscattered images of Z (a), ZM5 (b) and ZM10 (c) specimens. The inset in (b) highlights residual MoSi2 phase between SiC grains formed upon
sintering[10].

158
G. D’Urso et al. Precision Engineering 65 (2020) 156–163

Fig. 3. Material removal rate (MRR) as a function of machining condition for Fig. 5. Machining time as a function of the machining condition and the
ZrC-based ceramics and AISI 316L steel[12]. workpiece material. The numbers on each cylinder indicate the percentage of
set-up time compared to the machining time[14].

the machining process.

The MRR plot shows similar trends for the machining performed on
the ceramics workpiece. It is well defined as the greater difference in
machining speed compared with AISI 316L. In general, pre-roughing
and roughing approaches are eight times faster than finishing and
fine-finishing approaches. This aspect can be justified both from the
process and material point of view. The high velocity of pre-roughing
and roughing approaches in comparison to the finishing can be
explained by the higher intensity of the discharges due to the longer
duration and the higher value of the peak of current that allows devel­
oping more thermo-electric energy to the machining area. In particular,
long discharges allow developing higher energy for a single electrical
discharge, thus causing a greater material removal rate. From the ma­
terial point of view, this difference is probably due to the unique
structure and composition of the workpiece materials. In particular, the
specimens are characterized by a porous structure and the presence of
different phases with different chemical compositions. Each phase has
its own chemical and thermal properties, which lead to different
behaviour of the material in terms of energy transfer, thermal conduc­
tivity and melting point. For the machining performed on AISI 316L, the
Fig. 4. Tool Wear Ratio (TWR) as a function of machining condition for ZrC- MRR variability as a function of machining conditions is negligible.
based ceramics and AISI 316L steel[13]. Furthermore, the results of MRR show that the machining speed is not
influenced by the relative density of the materials, but rather by the
amount around 10 vol% and 20 vol%, respectively) are SiC particles, presence of a low conducting phase, SiC (i.e. MMR is higher for ZM10
often embedded into residual MoSi2, which appears as a grey phase with which contains 15 vol% of SiC in its final microstructure).
irregular shape, inset of Fig. 2b. Due to the free carbon in the ZrC On the contrary, the TWR is characterized by higher stability for all
starting powder and the reducing atmosphere in the hot pressing machining conditions for ceramics materials (Fig. 4), and no appreciable
chamber, most of the MoSi2 phase was converted to SiC during the dependence on the materials composition is observed. For this perfor­
sintering process [27,28], resulting in different compositions as mance parameter, the machining performed on stainless steel plates are
compared to the nominal ones. Porosity in both cases was indeed below instead characterized by higher variability. Furthermore, the tool wear
1%, as confirmed by SEM inspection (Table 1). The final amount of ratio is smaller for ceramics than for AISI 316L due to the high proba­
MoSi2 detected by image analysis after sintering was indeed reduced to bility to find discontinuities in the ceramic structure, which reduces the
around 2–5 vol%, in agreement with previous investigations [12,27,28]. generation of abnormal discharges, such as arcing and shorts. In fact,
μEDM behaviour is affected not only by the machining parameters but
also by the material microstructures. In general, porosity increases the
3.2. μEDM performances
MRR and the pore size significantly influences its value [29,30].
Fig. 5 shows the machining time divided by erosion time and set up
Fig. 3 and Fig. 4 show the trend of μEDM process performances in
time. Set up time is the amount of machining time in which the dis­
terms of MRR and TWR as a function of workpiece materials and
charges do not occur (i.e. moving time, tool wear touching operations,
machining conditions. A comparison with the results obtained on AISI
etc.). It is important to underline that the experimental conditions are
316L stainless steel is also reported for comparison purposes. Its selec­
the same for all the performed tests. It is possible to observe that the set-
tion is related to the fact that it represents one of the most common
up time has more influence on the roughing approaches and this could
material machined by μEDM and allows defining the main differences in

159
G. D’Urso et al. Precision Engineering 65 (2020) 156–163

Fig. 6. SEM images of the specimens after the pre-roughing process. Effect of composition variation on the morphology of as produced surfaces. Left: images taken
with secondary electrons, right: images taken with backscattered electrons[16].

be related to the higher tool wear. In fact, thanks to the electrode wear this parameter, Fig. 6. In absence of MoSi2, uneven distribution of re-
compensation method applied automatically by the ED-machine, when solidified materials can be noticed. In the zones characterized by the
the electrode is too small to work, the spindle is moved to the control presence of a high amount of re-solidified materials, smooth edges of the
position and the electrode lengthens. High tool wear involves several pools and several cracks can be evidenced due to the high re-
elongations which affect the set-up time as reported in the histogram, solidification rate and to the residual internal tensile stresses of oxi­
explaining the higher incidence of set-up time for the machining char­ dised material with respect to the base. The dimensions and distribution
acterized by high wear. The histogram shows a lower value of incidence of the pools of re-solidified material are instead quite evenly distributed
of set-up time in the AISI 316L machining, despite the high tool wear. for ZM5 and ZM10, and they are smaller moving from 0 to 10 vol%
This could be related to the higher duration of the erosion process. MoSi2. The size and distribution of the different phases well reflect the
The material removal mechanism that occurred on AISI 316L is well material microstructure at different MoSi2/SiC contents. The back­
known and it is simply described by melting and evaporation phenom­ scattered images (Fig. 6, right) evidences the presence of dark zones
ena [24,25,31]. The situation is different for these particular ceramic which represent the material porosities, the grey phase is the re-
samples, which are characterized by a very high melting point. In this solidified material containing carbon, zirconium, silicon, molybdenum
case, more energy is needed to remove material, so the expectation was and oxygen, and the white zones are mainly constituted by tungsten
that the ceramics were characterized by a longer process but the porous which is the material transferred from the electrode to the workpiece
structure allows to improve the machining speed in comparison to more during the μEDM process. The amount of dark areas is quite high for
dense structure (AISI 316L). This aspect can be supported by the pure ZrC, which is characterized by the highest porosity, whilst lower
following considerations taken from the microstructural analysis of the amounts can be noticed at increasing MoSi2 content, due to a denser
machined surfaces. structure. It seems that the generation of sparks during the EDM process
is influenced by the presence of MoSi2 which causes a significant
3.3. Microstructure of machined ceramics modification in the microstructure by reducing the total porosity of the
composites and modifying the electroconductive homogeneity of the
The morphologies of the machined surfaces appear quite different as composites material. This allows supposing the generation of a prefer­
a function of the amount of secondary phases. In the specific, both the ential removal path for material removal. The mechanism seems to be
size and distribution of fused and re-solidified material are affected by mainly governed by the preferential melting of the ZrC grain

160
G. D’Urso et al. Precision Engineering 65 (2020) 156–163

Fig. 7. EDS spectra recorded on a squared area of the specimens before and after the pre-roughing process[17].

boundaries, the porosity pattern and the presence of microstructural

Table 3
discontinuities, i.e. the presence of second phases such as silicon carbide
Chemical composition measured by EDS analysis on a squared area of raw
or residual molybdenum disilicide. The EDS spectra of the machined
material and on machined surfaces[19].
surfaces compared to raw material are shown in Fig. 7 and qualitative
Material Z ZM5 ZM10 compositions are shown in Table 3. The enrichment in carbon can be
Weight Raw Machined Raw Machined Raw Machined noticed in both the ceramics with MoSi2, whilst no substantial difference
(%) can be noticed for pure ZrC. Also, the ratio between the elements of the
C 33.86 28.33 � 37.77 63.73 � 39.13 65.22 �
second phase – i.e. Mo, Si, and O - changes in function of the amount of
� 1.70 0.66 � 1.61 0.5 � 1.59 0.61
O 5.26 � 2.96 � 2.95 � 11.14 � 3.44 � 6.96 � MoSi2 and depletion in Mo can be noticed for both Z5 and Z10 specimens
0.67 0.25 0.61 0.37 0.61 0.39 after machining. The size and distribution of tungsten rich zones - i.e. the
Si – – 0.90 � 1.79 � 4.34 � 0.27 � light areas - and also the amount of tungsten estimated by EDS analysis
0.07 0.03 0.14 0.02 (Table 3) are related to the tool wear since the tungsten carbide elec­
Zr 60.88 66.95 � 55.16 20.98 � 50.93 24.11 �
� 1.61 0.65 � 1.48 0.27 � 1.38 0.39
trode was used.
Mo – – 3.22 � 0.84 � 2.15 � 1.41 �
0.47 0.1 0.45 0.13
3.4. Machining behaviour
W – 1.77 � 1.52 � – 2.02 �
0.16 0.08 0.08
As mentioned above, it seems that the distribution of sparks during
the μEDM process follows preferential paths and the presence of sec­
ondary phases, mainly SiC, with MoSi2 only in a minor amount, plays a
certain role. The material removal mechanism by μEDM seems to be

161
G. D’Urso et al. Precision Engineering 65 (2020) 156–163

Fig. 8. Polished cross-sections of the machined specimens with nominal 0% (a), 5 vol% (b) and 10 vol% MoSi2 (c) upon roughing/pre-roughing/finishing/fine-
finishing step. Above: Details of the micro-pocket profiles, below the EDS elemental mapping showing negligible oxygen and silicon variation across the section
[19]. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 9. Average profile roughness, Ra; of the machined specimens in relation to the porosity level[20].

162
G. D’Urso et al. Precision Engineering 65 (2020) 156–163

mainly governed by the preferential melting of the ZrC grains, the [3] Silvestroni L, Kleebe HJ, Fahrenholtz WG, Watts J. Super-strong materials for
temperatures exceeding 2000 � c. Sci Rep 2017;7.
porosity pattern and the presence of microstructural discontinuities, i.e.
[4] Pierson HO. Carbides of group VI. Handb Refract Carbides Nitrides 2008;100–117.
the secondary phases. A clear example is the ZM10 sample, in fact, it is [5] Justin JF, Jankowiak A. Ultra high temperature ceramics: densification, properties
characterized by porosity <1%, and the machining is performed with and thermal stability. AerospaceLab J 2011;3:AL3-08.
MRR higher than the MRR of pure zirconium, having a porosity of about [6] Opeka MM, Talmy IG, Zaykoski JA. Oxidation-based materials selection for 2000� C
þ hypersonic aerosurfaces: theoretical considerations and historical experience.
27%. This can be justified by the combination of the porosity with the J Mater Sci 2004;39:5887–904.
presence of secondary phases. [7] Upadhya K, Yang JM, Hoffman W. Advanced materials for ultrahigh temperature
Observations performed on the polished cross-section seem to structural applications above 2000� C. Am Ceram Soc Bull 1997;58:51–6.
[8] Scatteia L, Monteverde F, Alfano D, Cantoni S. Advances in ultra high temperature
confirm such a mechanism (Fig. 8). ZrC grains are well attached to each ceramics for hot structures. Trans JSASS Sp Tech Jpn 2009;7:73–8.
other, whilst rougher surfaces are obtained in the presence of high [9] Ogawa T, Ikawa K. Diffusion of metal fission products in ZrC1.0. J Nucl Mater
amount porosities in Z (Fig. 8a). The presence of second phases in ZM5 1982;105:331–4.
[10] Ryu HJ, Lee YW, Cha S Il, Hong SH. Sintering behaviour and microstructures of
and ZM10 (Fig. 8b and c) leaves a smoother surface compared to Z. In carbides and nitrides for the inert matrix fuel by spark plasma sintering. J Nucl
any case, the elemental map distribution of oxygen revealed negligible Mater 2006;352:341–8.
profile depth and no depletion of Si-based phases (Fig. 8 – bottom). [11] Silvestroni L, Sciti D. Sintering behavior, microstructure, and mechanical
properties: a comparison among pressureless sintered ultra-refractory carbides.
Rougher surfaces are obtained in the presence of a high amount of dis­ Ann Mater Sci Eng 2010;2010:1–11.
continuities [32] - mainly porosities - in the case of Z, which is char­ [12] Sciti D, Silvestroni L, Saccone G, Alfano D. Effect of different sintering aids on
acterized by the lowest relative density (about 73%). The roughness thermo-mechanical properties and oxidation of SiC fibers - reinforced
ZrB2composites. Mater Chem Phys 2013;137:834–42.
significantly reduces in presence of second phases, owing to the elimi­
[13] Silvestroni L, Sciti D. Transmission electron microscopy on Hf- and Ta-carbides
nation of porosity (Fig. 9). sintered with TaSi2. J Eur Ceram Soc 2011;31:3033–43.
[14] Sciti D, Guicciardi S, Nygren M. Spark plasma sintering and mechanical behaviour
4. Conclusions of ZrC-based composites. Scripta Mater 2008;59:638–41.
[15] Nú~nez-Gonz� alez B, Ortiz AL, Guiberteau F, Padture NP. Effect of MoSi2 content on
the lubricated sliding-wear resistance of ZrC-MoSi2 composites. J Eur Ceram Soc
ZrC-based materials, sintered pure and in presence of 5 and 10 vol% 2011;31:877–82.
of MoSi2, were machined by micro electro-discharge machining (μEDM) [16] Fahrenholtz WG, Wuchina EJ, Lee WE, Zhou Y, Ceramics U-HT. Materials for
extreme environment applications. Wiley2014 2014;111:57.
and different conditions of surface finishing steps, comprising pre- [17] Lauwers B, Kruth JP, Liu W, Eeraerts W, Schacht B, Bleys P. Investigation of
roughing, roughing, finishing and fine-finishing, were analysed. Ana­ material removal mechanisms in EDM of composite ceramic materials. J Mater
lyses on the ceramics were integrated by the evaluation of process per­ Process Technol 2004;149:347–52.
[18] D’Urso G, Maccarini G, Quarto M, Ravasio C, Caldara M. Micro-electro discharge
formances of μEDM process and a comparison with results obtained on machining drilling of stainless steel with copper electrode: the influence of process
one of the most common material machined by μEDM, i.e. stainless steel. parameters and electrode size. Adv Mech Eng 2016;8:1–16.
The initial microstructure of the ZrC-ceramics resulted very different [19] Gadow R, Landfried R, Kern F. Electrical discharge machining (EDM) of high-
performance ceramics. Proceedings of the III advanced ceramics and applications
depending on the amount of MoSi2 phase initially introduced. Pure ZrC Conference. 2016. p. 25–32.
had a quite high fraction of residual porosity, about 27%, whilst the [20] Kadirvel A, Hariharan P, Gowri S. A review on various research trends in micro-
other two materials were below 1%. On the other hand, most of MoSi2 EDM. Int J Mechatron Manuf Syst 2012;5:361–84.
[21] Raju L, Hiremath SS. A state-of-the-art review on micro electro-discharge
phase converted to SiC, owing to the reducing environment of the hot-
machining. Procedia Technol 2016;25:1281–8.
pressing chamber. [22] D’Urso G, Giardini C, Quarto M, Maccarini G. Cost index model for the process
The process performance evaluation underlined different behaviour performance optimization of micro-EDM drilling on tungsten carbide.
of ceramic materials moving from pre-roughing to fine-finishing; in fact, Micromachines 2017;8:251.
[23] D’Urso G, Quarto M, Ravasio C. A model to predict manufacturing cost for micro-
only for fine finishing approaches, the machining are affected by a very EDM drilling. Int J Adv Manuf Technol 2017;91:2843–53.
slow removal process. This is related to the lower quantity of energy [24] D’Urso G, Giardini C, Quarto M. Characterization of surfaces obtained by micro-
transferred to the surfaces from the short pulses. In general, the porous EDM milling on steel and ceramic components. Int J Adv Manuf Technol 2018;97:
2077–85.
structure and the presence of a secondary phase allow improving the [25] D’Urso G, Giardini C, Maccarini G, Quarto M, Ravasio C. Analysis of the surface
machining speed thanks to the selection of preferential path such as quality of steel and ceramic materials machined by micro-EDM. 2018. p. 431–2.
demonstrated by the microstructure of materials. Although notable [26] Saccone G, Gardi R, Alfano D, Ferrigno A, Del Vecchio, Laboratory A. On-ground
and in-flight investigation of ultra high temperature ceramic composite materials.
microstructure variations, only slight differences in machining perfor­ Aero Sci Technol 2016;58:490–7.
mances were measured. [27] Silvestroni L, Sciti D, Kling J, Lauterbach S, Kleebe HJ. Sintering mechanisms of
zirconium and hafnium carbides doped with MoSi 2. J Am Ceram Soc 2009;92:
1574–9.
Declaration of competing interest [28] Silvestroni L, Sciti D. Microstructure and properties of pressureless sintered ZrC-
based materials. J Mater Res 2008;23:1882–9.
The authors declare that they have no known competing financial [29] Zou R, Yu Z, Li W, Guo M, Li J. Influence of porous structure on the machining
performance of micro EDM. J Mater Process Technol 2016;232:43–51.
interests or personal relationships that could have appeared to influence
[30] Wang K, Zhang Q, Liu Q, Zhu G, Zhang J. Experimental study on micro electrical
the work reported in this paper. discharge machining of porous stainless steel. Int J Adv Manuf Technol 2017;90:
2589–95.
References [31] Basak I, Ghosh A. Mechanism of material removal in electrochemical discharge
machining: a theoretical model and experimental verification. J Mater Process
Technol 1997;71:350–9.
[1] Wuchina E, Opeka M, Causey S, Buesking K, Spain J, Cull A, Routbort J, Guitierrez- [32] Sani E, Mercatelli L, Meucci M, Silvestroni L, Balbo A, Sciti D. Process and
Mora F. Designing for ultrahigh-temperature applications: the mechanical and composition dependence of optical properties of zirconium, hafnium and tantalum
thermal properties of HfB2, HfCx, HfNx and АHf(N). J Mater Sci 2004;39:5939–49. borides for solar receiver applications. Sol Energy Mater Sol Cells 2016;155:
[2] Pierrat B, Balat-Pichelin M, Silvestroni L, Sciti D. High temperature oxidation of 368–77.
ZrC-20%MoSi2in air for future solar receivers. Sol Energy Mater Sol Cells 2011;95:
2228–37.

163