Surface & Coatings Technology 201 (2006) 2103 – 2108

www.elsevier.com/locate/surfcoat

Plasma spray deposition of ultra high temperature ceramics

Mario Tului a,⁎, Giuliano Marino b , Teodoro Valente c
a
Centro Sviluppo Materiali S.p.A. – Via di Castel Romano 100, 00128 Rome, Italy
b
CIRA S.C.p.A. – Via Maiorise, 81043 Capua, Italy
c
Rome University “la Sapienza”, ICMMPM Dept., CR-INSTM Eng. lab. on Surf. Treatm. (LITS) – Via Eudossiana 18, 00184 Rome, Italy
Available online 12 July 2006

Abstract

An innovative, proprietary methodology was developed to produce, by plasma spraying deposition, a ceramic composite containing SiC
particles dispersed in a ZrB2 matrix. With such a technique both coatings and free standing parts were fabricated. In spite of the well known
difficulty to obtain plasma sprayed coatings containing SiC, characterisation results evidenced that the used process did not affect the phase
composition. Thermal and mechanical behaviour of the obtained material have been extensively characterized, in order to assess its applicability to
specific space missions. In a first phase, tests were carried out on laboratory scale samples. Scope of the second phase of the activity, which is
currently running, is the validation of the developed material in simulated operative conditions, by demonstrative components testing in a PWT
(plasma wind tunnel) facility. A preliminary test campaign has been successfully completed, demonstrating that UHTC plasma sprayed coatings
are able to withstand high temperature oxidising conditions.
© 2006 Elsevier B.V. All rights reserved.

Keywords: Plasma spraying; Composite coatings; Free standing parts; Refractory ceramics

1. Introduction ZrB2, HfB2) with a dispersion of carbide particles (e.g., SiC, ZrC,
HfC). The most investigated composition is based on a ZrB2
In the next generation vehicles for orbital flights, the con- matrix and a SiC particles dispersion (1540 vol.%). At high
figuration presently adopted by e.g. the Space Shuttle, char- temperature, ZrB2 provides mechanical resistance, whilst SiC
acterised by blunt surfaces, will be substituted by sharp profiles, generates an oxidation protective scale on the material surface
which reduce the aerodynamic resistance and allow the vehicle to [5–8].
manoeuvre during the landing phase [1,2]. Such a favourable An innovative, proprietary methodology was developed to
configuration has not been adopted up to now due to the very produce, by plasma spraying deposition, a ceramic composite
high temperatures (higher than 2200 K) caused by aerodynamic containing SiC particles dispersed in a ZrB2 matrix [9]. With such
friction on sharp profiles; another aspect of paramount a technique both coatings and free standing parts were fabricated.
importance is that such high temperatures are reached in an In spite of the well known difficulty to obtain plasma sprayed
oxidising environment. Moreover, reusability and maintainabil- coatings containing SiC, characterisation results evidenced that
ity are critical issues. the used process did not affect the phase composition [10].
High temperature resistant ceramics, such as C/C and C/SiC, Thermal and mechanical behaviour of the obtained material
widely used in the realisation of space vehicles, cannot operate in have been extensively characterized, in order to assess its appli-
a so critical application. As a consequence, a completely cability to specific space missions. In a first phase, test were
innovative class of materials, denominated UHTC (ultra high carried out on laboratory scale samples. Results were already
temperature ceramic) is under development [3,4]. UHTC are published by the authors [10–13], and, for sake of clarity, they
composite ceramics constituted of a matrix of borides (TiB2, will be repeated in the present paper.
Scope of the second phase of the activity, which is currently
running, is the validation of the developed material in simulated
⁎ Corresponding author. Tel.: +39 06 5055 742. operative conditions, by demonstrative components testing in a
E-mail address: [email protected] (M. Tului). PWT (plasma wind tunnel) facility. A preliminary test campaign
0257-8972/$ - see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.surfcoat.2006.04.053
2104 M. Tului et al. / Surface & Coatings Technology 201 (2006) 2103–2108

Fig. 1. Tensile tests specimens: shape and size (dimensions in mm).

has been successfully completed, demonstrating that UHTC laboratory scale. Among them, 4 points bending tests and tensile
plasma sprayed coatings are able to withstand high temperature tests were carried out. Tests blocks of 3 × 4 × 50 mm3 in size were
oxidising conditions. used for 4-point bending tests, and, for tensile tests, on “dog
bone” samples with shape and size as reported in Fig. 1.
2. Experimental Characterisations were carried out according to the standard
procedures ASTM C1161-02c and ASTM C1273-95a, respec-
2.1. Tests on samples tively. Tests were carried out both on as sprayed samples and on
samples after exposure to 2173 K in air for 3 min.
Coating samples were deposited following a procedure Cycling tests were carried out on 4 × 1.5 × 22 mm3 blocks.
described in details elsewhere [10–15]. The coating material They were exposed to 2173 K in air for 2 min by flame heating
powder was prepared by blending ZrB2 and SiC fine powders (acetylene + oxygen) and then rapidly cooled in water (thermal
(75 and 25 wt.%, respectively) and agglomerating the blend by shock tests) or naturally cooled in air (thermal cycling tests).
spray drying. The powder was sprayed under IPS (inert gas Thermal diffusivity measurements were carried out by means
plasma spraying) conditions onto flat graphite substrates. of the laser flash technique. An unfocused laser pulse was ab-
Finally, the deposited layers were mechanically detached from sorbed into the surface of a cylindrically shaped test piece,
the substrates and machined by spark-erosion, followed, when causing its homogeneous heating. On the back of the test piece
necessary, by surface finishing.
A CAPS1 (controlled atmosphere plasma spray) equipment
[16], produced by Sulzer Metco (Switzerland), was used. It
consists of a 80 kW plasma torch installed within a 7 m3 vessel,
which allows depositions under air or inert atmosphere, in the
pressure range 10–4000 hPa.
XRD (X-ray diffraction) analyses were carried out on flat
samples to verify their phase composition. For comparison, the
starting material was also characterized by XRD. A Siemens
D500 (D) equipment was employed.
Some composite samples were cut, embedded in resin,
polished, and observed by SEM (scanning electron microsco-
py). The chemical composition of the various phases observed
was assessed by EDS (energy dispersion spectroscopy)
analyses. A Philips SEM 500 (NL) equipment was used.
The mechanical properties of the material obtained by
plasma spraying were evaluated by different tests on a

1
The use of the CAPS plant is shared by Rome University “La Sapienza”
and Centro Sviluppo Materiali S.p.A., based on a scientific collaboration
agreement. Fig. 2. Plasma wind tunnel test article.
 M. Tului et al. / Surface & Coatings Technology 201 (2006) 2103–2108 2105

Fig. 3. Schematic of “Scirocco” plasma wind tunnel.

the temperature increase was measured as a function of time. The was covered by a 5 mm thick layer of C/SiC, obtained by CVI
mathematical analysis of this temperature/time function allowed (Chemical Vapour Infiltration). On the C/SiC, a layer of UHTC
the determination of the thermal diffusivity, α. A Netzsch LFA was deposited by plasma spraying, following the procedure
427 (D) equipment was used. The measurements were carried described in the previous section.
out on a sample which was mounted in a furnace to allow the Tests were carried out at the “Scirocco” plasma wind tunnel
measurements of α at different temperatures in the range of 293– [17–19] plant, which represents one of the most powerful
1573 K. The measurements were carried out both during the hypersonic test facility in the world (70 MW) and produces a
heating and the cooling phases, and repeated during three very uniform and large test jet (up to 2 m diameter). In Fig. 3 a
heating/cooling cycles. schematic of “Scirocco” facility is depicted.
The process air is thermally energised into the segmented
2.2. Plasma wind tunnel tests constricted arc heater, reaching temperatures between 2000 and
10 000 K. This energy is then translated into kinetic energy by
The test article for the first experimental campaign in PWT
was a graphite cone, 125 mm of base diameter and 175 mm
high; the tip curvature radius was 5 mm (Fig. 2). The graphite Table 1
Parameters used for the PWT tests
Parameter Numerical values
Arc heater current (A) 2946
Arc heater voltage (V) 3619
Total mass flow rate (kg/s) 1.19
Arc Heater total enthalpy (MJ/kg) 4.19
Arc heater total pressure (bar) 3.58
Probe stagnation pressure (mbar) 9.47
Probe stagnation heat flux (kW/m2) 277

Table 2
Exposure times used during the PWT tests
Test Total exposure time Sweep time High temperature time
number (s) (s) (s)
1 30 5 20
2 45 5 35
Fig. 4. Test article (at the centre of the white circle) installed within the PWT
3 60 5 50
chamber.
2106 M. Tului et al. / Surface & Coatings Technology 201 (2006) 2103–2108

Fig. 5. Comparison between XRD spectra of (a) a mixture of 75 wt.% ZrB2 and 25 wt.% SiC powders and (b) of the material after spraying.

conveying the process air through a convergent–divergent Different nozzles with different length and exit diameter are
conical nozzle. As a result, an hypersonic test jet is generated available. After the air flow impact with the model in the test
with velocities ranging between 2000 to 6000 m/s and Mach chamber the jet is collected by a tube of 50 m length (the
numbers between 6 to 12 depending on the exit nozzle size. diffuser), goes through a heat exchanger, is pumped away by a
vacuum system and then expelled into the atmosphere.
Fig. 4 shows the test article installed within the PWT
chamber. Parameters employed in the PWT tests were reported
on Table 1. Three tests were carried out with the same
parameters reported in Table 1, but using increasing exposure
times, as shown in Table 2.

3. Results

3.1. Tests on samples

Fig. 5 shows the XRD diffraction patterns of the starting

mixture of ZrB2 and SiC powders (a) and the plasma sprayed
composite (b). By comparing the two spectra, it is evident that

Fig. 6. Cross section of a sample observed by SEM. Table 3

Results of mechanical tests
Average Standard Weibull Scale
deviation modulus parameter
UTS Tensile As 41 6.6 7.4 43.6
(MPa) sprayed
Heat 32 8.1 – –
treated
MOR 4 point As 81 6.1 9.6 79.4
(MPa) bending sprayed
Heat 60 10.4 – –
treated
E (GPa) Tensile As 47 7.8 – –
sprayed
Heat 76 12.2 – –
treated
4-point As 54 6.5 – –
bending sprayed
Heat 74 19.5 – –
Fig. 7. Cross section of a sample observed by SEM at high magnification:
treated
(A) ZrB2; (B) ZrB2 + SiC.
 M. Tului et al. / Surface & Coatings Technology 201 (2006) 2103–2108 2107

Fig. 8. Thermal diffusivity results.

Fig. 10. Test article after plasma wind tunnel tests.
deposition process did not affect the phase composition of the
material. In fact, no phases coming from the decomposition of High temperature exposure resulted in a decrease of UTS and
the initial constituents (e.g. ZrO2, SiO2) were present in the modulus of rupture, and in an increase of Young's modulus,
coating, at least in an amount high enough to be detected by possibly due to an increase of the surface defects by formation
XRD (∼1 vol.%). of an oxidation layer and to the densification of material by
Fig. 6 shows the cross-section of a coating sample observed induced sintering effects [10–14].
by SEM, utilising backscattered electrons. EDS analyses carried Results concerning the number of cycles to failure are
out on the areas of different grey shades allowed to identify that reproducible with an average of 9 and 24 cycles for water and air
the white areas were constituted of ZrB2, and the darker ones cooling, respectively. It was observed that the specimen surface
were a fine mixture of ZrB2 and SiC. tends to form a sort of exfoliation layers, due to oxidation and
As shown by Fig. 5, SiC did not decompose during spraying. phase composition modifications [10–14]. After each cooling
It can be explained by the formation of an eutectic phase with cycle, such layers detached from the surface, causing a reduction
ZrB2. In the literature [20] a SiC–ZrB2 pseudo-binary phase of the specimen resistant section. In the case of water cooling,
diagram is reported, actually showing an eutectic at about brittle fractures, typical of ceramic materials, were observed
2500 K. The examination of high magnification SEM pictures, after 9 cycle, when the specimen section was reduced at about
as the one reported in Fig. 7, seems to confirm this explanation. one half of the original one. In the case of air cooling, this
In fact, eutectic-like regions can be observed, where very fine phenomenon was extended until the last cycle to failure, with full
SiC particles are homogeneously dispersed in a ZrB2 matrix. material consumption. No brittle fractures were observed.
More microstructural investigations are in progress. Fig. 8 shows the results [13] of the thermal diffusivity
Results in terms of average ultimate tensile strength (UTS) measured by laser flash methodology. Data indicated by open
for as sprayed and exposed samples, average modulus of rup- diamonds connected by a solid line are the results of the first
ture (MOR) for as sprayed and exposed samples and Young's cycle of measurements, carried out on the sample in the as-
moduli (E) are reported in Table 3. Weibull and normal prob- produced state. As indicated by the arrows, the thermal dif-
ability based regression of experimental data were also per- fusivity values measured during the heating phase of the first
formed to evaluate scattering of experimental values. cycle were lower than the ones measured during the cooling

Fig. 9. Test article observed by a thermo-camera during PWT test.

2108 M. Tului et al. / Surface & Coatings Technology 201 (2006) 2103–2108

phase. Data belonging to the second and third cycles (solid References
symbols and dashed lines) did not show significant differences
when acquired during the heating or the cooling stage. [1] M.O. Wolfe, Proceedings 3rd Symposium on Naval Struct. Mech, Per-
The phenomenon described above could likely be explained gamon Press, New York, 1984.
[2] Report Boeing C.A. for NASA Evaluation of Civil Hypersonic Transport
with a partial sintering of the sample, during the first cycle, which Reliability, Langley Research Center, 1990.
caused a reduction of the porosity content and, as a consequence, [3] J. Bull, M.J. White, L. Kauffman. US patent 5,750,450 (1998).
an increase in the thermal diffusivity. During the following cycles [4] E. Clougherty, AMFL-TR-68-190: Part II, vol. VII, 1968.
no more sintering occurred, and the diffusivity value stabilised at a [5] B. Morel. U.S. patent 5,240,084 (1995).
value of about 2.8 × 10− 2 cm2/s. [6] E.V. Clougherty, E.T. Peters, D. Kalish, Mater. Processes 70s: 15th Nat.
SAMPE (Soc. Aerosp. Mater. Process Eng.) Symp. Exhib, 1969, p. 297.
[7] L. Kauffman, NSWC TR 86-242, 1986.
3.2. Plasma wind tunnel tests [8] J. Bull, NASA CP-3175: Part 1. Materials Park, OH, 2000, p. 777.
[9] M. Tului, T. Valente U.S. Pat. Appl. Publ., US 2002/0151427 (2002).
Fig. 9 shows the test article during the first test, observed by [10] C. Bartuli, T. Valente, M. Tului, Surf. Coat. Technol. 155 (2002) 260.
an infrared camera. Since it was a preliminary test, conditions [11] T. Valente, C. Bartuli, G. Visconti, M. Tului, in: C. Berndt (Ed.), Proc.
International Thermal Spray Conference: Thermal Spray 2000, “Thermal
were not very stressed: as a consequence, temperatures as high Spray. Surface Engineering via Applied Research”, ASM International,
as 1000 °C were recorded on the tip. Materials Park, OH, 2000, p. 837.
Fig. 10 shows the test article after the PWT tests. Visual [12] C. Bartuli, T. Valente, M. Tului, in: C.C. Berndt, K.A. Khor, E.F.
inspection did not reveal any detachment of the coating from the Lugscheider (Eds.), Proc. International Thermal Spray Conference:
substrate. A colour change, from pale grey to dark grey, was Thermal Spray 2001, New Surfaces for a New Millennium, 28–30 May
2001, Singapore, ASM Int. Publ., 2001, p. 259.
observed in the thermally most stressed part of the test article, [13] M. Tului, G. Marino, T. Valente, ESA SP SP-521 (2003) 161.
i.e., the tip. Destructive tests on the test article are currently in [14] T. Valente, G. Marino, M. Tului, International Thermal Spray Conference
progress. 2004 – Osaka (J) 10–12/05, 2004.
[15] M. Tului, F. Ruffini, F. Arezzo, S. Lasisz, Z. Znamirowski, L. Pawlowski,
Surf. Coat. Technol. 151–152 (2002) 483.
4. Conclusions
[16] F. Carassiti, U. Franzoni, M. Tului, T. Valente, 4th Euro Ceramics
Conference, Riccione, 2–6/10, 1995.
An innovative methodology to produce UHTC materials has [17] S. Caristia, F. De Filippis, A. Del Vecchio, C. Purpura, 4th European
been developed. Characterisation on laboratory scale samples, Symposium Aerothermodynamics for Space Applications, Capua, 2001.
as well preliminary tests under simulated operative conditions, [18] G. Russo, F. De Filippis, S. Borrelli, M. Marini, S. Caristia, in: AIAA
indicate that the obtained material should be adopted in thermal (Ed.), Advanced Hypersonic Test Facilities, 2002, p. 313.
[19] S. Caristia, F. De Filippis, A. Del Vecchio, E. Graps, Proceedings of 54th
structures for future space vehicles. International Astronautical Congress, Bremen, 2003.
Further test campaigns in plasma wind tunnel facilities are [20] D.A. Krivoshein, M.A. Maurakh, V.S. Dergunova, V.S. Petrov, Poroshko-
being planned to further assess the performance of the developed vaiya Metall. 8 (1980) 58 (in Russian).
material. The objective is to obtain sufficient data to plan a flight
test.