- Email: [email protected]
Joining of ceramic matrix composites to high temperature ceramics for thermal protection systems
G Model
ARTICLE IN PRESS
JECS-10300; No. of Pages 7
Journal of the European Ceramic Society xxx (2015) xxx–xxx
Contents lists available at www.sciencedirect.com
Journal of the European Ceramic Society journal homepage: www.elsevier.com/locate/jeurceramsoc
Joining of ceramic matrix composites to high temperature ceramics for thermal protection systems C. Jiménez a , K. Mergia b,∗ , M. Lagos a , P. Yialouris b , I. Agote a , V. Liedtke c , S. Messoloras b , Y. Panayiotatos b , E. Padovano d , C. Badini d , C. Wilhelmi e , J. Barcena a a
Tecnalia Research & Innovation, Industry and Transport Division, Mikeletegi Pasealekua, 2, Donostia-San Sebastian, Spain N.C.S.R “Demokritos”, Institute of Nuclear and Radiological Science and Technology, Energy and Safety, Aghia Paraskevi, 15310 Athens, Greece c Aerospace & Advanced Composites GmbH, Viktor-Kaplan-Straße 2, A-2700 Wiener Neustadt, Austria d Politecnico di Torino, Department of Applied Science and Technology, Corso Duca degli Abruzzi, 24, 10129 Torino, Italy e Airbus Defence and Space GmbH, Dept. TCC 2/CTOIWMS, 81663 Munich, Germany b
a r t i c l e
i n f o
Article history: Received 12 May 2015 Received in revised form 22 September 2015 Accepted 27 September 2015 Available online xxx Keywords: Ceramic matrix composite Silicon carbide Joining Thermal tests Aerospace
a b s t r a c t The current work reports a novel approach for the integration of external protective SiC multilayers with ceramic matrix composite (Cf /SiC) with the view of application in aerospace heat protection systems. The integration method is based on diffusion brazing bonding. As a joining agent the MAX-Phase Ti3 SiC2 , produced by self-propagating high temperature synthesis, has been employed. The pressure applied during the joining process and its effect on the microstructure of the integrated structure is discussed. Microstructural analysis of the resulting joints is conducted using scanning electron microscopy coupled with energy dispersive spectroscopy and X-ray diffraction measurements. Analysis of the joints showed that the bonds are uniform, dense, with few crack vertical to the interface which are not detrimental for the performance of the joints. Ground re-entry tests showed that the joints survive 5 re-entry cycles at 1391 and 1794 ◦ C without any detectable damage. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction One of the key technologies of today’s aerospace applications refers to the effective protection of components subject to large heat loads like the heat shields of re-entry vehicles or the inner walls of combustion chambers. This issue is of great importance for future missions where considerably increased energy densities are expected. Advanced ceramic matrix composite (CMC) materials based on SiC matrix reinforced with carbon fibers (Cf /SiC) are key materials for aerospace applications since they present superior mechanical properties and resistance against high temperatures and being at the same time lightweight and cost effective. For the application of these bulk ceramics in hostile environments appropriate coatings have to be developed that will prevent the oxidation of the carbon fibers. The current thermal protection systems are based on the use of ceramic matrix composite materials coated with SiC [1] or other carbide based materials. SiC exhibits excellent oxidation resistance
∗ Corresponding author. Fax: +30 210 6545496. E-mail address: [email protected] (K. Mergia).
at high temperatures, because the formed glassy silica films prevent oxygen diffusion very efficiently and thus serve as protection against further oxidation. However, the amorphous silica at temperatures above 1200 ◦ C crystallizes to cristobalite causing surface cracking and also reacts with water or Na/K vapor, resulting in severe degradation of the silica film [2]. Alternatively to SiC coatings ultra high temperature SiC based multilayer ceramics can be employed in combination with CMCs in order to provide integrated systems with enhanced performance under the most severe re-entry conditions. The advantage in using multilayered ceramics stems from the capability of adjusting their architecture and chemical composition e.g., by incorporating ZrB2 layers at the outer structure, and thus tuning the functionality of these protection layers to specific re-entry conditions i.e., the required service temperature and the oxidative environment these structures have to withstand. Therefore, advances in joining science and technology of CMC-ultra high temperature and oxidation resistant ceramics to SiC based multilayer ceramics are important in order to integrate these systems and realize the benefits of these advanced materials in aerospace applications. A number of techniques for joining monolithic ceramic materials or CMCs to themselves have been developed such as solid state
http://dx.doi.org/10.1016/j.jeurceramsoc.2015.09.038 0955-2219/© 2015 Elsevier Ltd. All rights reserved.
Please cite this article in press as: C. Jiménez, et al., Joining of ceramic matrix composites to high temperature ceramics for thermal protection systems, J Eur Ceram Soc (2015), http://dx.doi.org/10.1016/j.jeurceramsoc.2015.09.038
G Model JECS-10300; No. of Pages 7
ARTICLE IN PRESS C. Jiménez et al. / Journal of the European Ceramic Society xxx (2015) xxx–xxx
sintering of the ceramics, diffusion bonding using various active fillers [3,4], transient eutectic-phase routes [4], glass–ceramic joining [5], metallic braze-based joining [6,7], SiC reaction forming [8], MAX-phase bonding [4,9], spark plasma sintering [10], integrated assembly method [11] or mechanical fastening [12]. However, to our knowledge no joining procedure for CMCs and SiC based multilayers systems has been reported so far. The current work reports on a approach for the integration of external protective ceramic SiC based multilayers (SiC ML) with a thermostructural ceramic matrix composite (Cf /SiC). This structure is a part of a more complex system for a powerful thermal protection system under a reference mission (ESA’s Advanced Re-entry Vehicle) [13]. The integration method of the two materials is based on the diffusion brazing bonding process. As a joining agent a filler metal based on a MAX-Phase Ti3 SiC2 , produced by self propagating high temperature synthesis, has been employed. Ti3 SiC2 has good stability at elevated temperatures and it presents both metallic and ceramic properties and exhibits plastic behavior. It has a melting point higher than 3000 ◦ C and a coefficient of thermal expansion (CTE) 9.12 × 10−6 /K between 25 and 1200 ◦ C [14]. It has been used in the past for joining of SiC based ceramics for application in the temperature range 1200–1600 ◦ C [9]. The process parameters of the joining method and their effect on the microstructure of the integrated structure are discussed. The structure and microstructure of the assembly has been investigated by optical microscopy, scanning electron microscopy with energy dispersive analysis and X-ray diffraction measurements. Results from thermal tests under re-entry conditions are presented and discussed with respect to the envisaged applications.
2. Materials and methods The Cf /SiC (SICARBON) ceramic composites were supplied by Airbus Defence and Space GmbH [15]. They consist of carbon fibers embedded in a silicon carbide matrix. The production process of this material is based on the polymer infiltration pyrolysis process (PIP). The infiltration of the carbon fibers with a pre-ceramic polymer-based and powder-filled slurry system is performed by liquid polymer infiltration (LPI) via filament winding. From the supplied material samples of 40 × 40 mm2 were cut and used for all the experiments. The SiC multilayers were fabricated by the tape casting technique followed by pressureless sintering [16]. The processing method involved several steps: SiC slurry preparation; tape casting; stacking, debinding and pressureless sintering. The SiC multilayers used in the current study consist of eleven dense SiC layers. For the joining of the SiC to SICARBON, the ternary MAX compound Ti3 Si1.5 C2 was used as a filler metal. The material was produced by self-propagating high temperature synthesis (SHS) in a reaction furnace and in argon atmosphere using titanium, silicon and carbon powders with atomic ratios 3Ti:1.5Si:2C. After the SHS process a foam was obtained and this foam was milled and sieved under 63 m. The crystal structure of the filler was assessed by X-ray diffraction (XRD) measurements. The joints were fabricated employing the diffusion brazing technique in a high vacuum hot press furnace at 1600 ◦ C. The pressures used were 2.7, 7 and 25 MPa. For the joints fabricated at 2.7 and 7 MPa the pressure was applied at room temperature and kept constant during the thermal cycle. For the joint fabricated at 25 MPa at the beginning of the thermal cycle and up to 1400 ◦ C a lower pressure of 15 MPa was applied up and for the rest of the thermal cycle during heating and cooling the pressure of 25 MPa was applied. A heating rate of 15 ◦ C/min up to 1400 ◦ C, followed by a heating rate of 10 ◦ C/min up to 1600 ◦ C and a slow cooling rate of 5 ◦ C/min were applied.
Ti3SiC2 TiC Si2Ti
Intensity (a.u.)
2
20
30
40
50
60
70
80
2Θ (deg.) Fig. 1. XRD pattern of Ti3 Si1.5 C2 .
The XRD patterns were measured using a Bruker D8 diffractometer equipped with a Cu K␣ radiation, a parallel beam stemming from a Göbel mirror and a Våntec position sensitive detector with 9◦ angular acceptance. The microstructure of the joints was examined using FEI Quanta Inspect scanning electron microscopy (SEM) coupled with energy dispersive X-ray spectroscopy (EDS). For the EDS analysis, different accelerator voltages were used in the range between 12 and 20 kV and a standardless quantitative analysis was performed employing the ZAF correction approach. Tests of the SiC-ML/CMC under re-entry conditions, i.e. temperature profiles under vacuum, were carried out at a ground re-entry test rig [17] at a maximum temperature of 1794 and 1391 ◦ C for 5 cycles each. The tests were performed under vacuum and the thermal profiles used correspond to two control points of the ESA’s Advanced Re-entry vehicle (ARV) [18].
3. Results and discussion 3.1. Filler composition and structure The filler metal was produced by SHS. In the SHS process for the formation of Ti3 SiC2 , there is a competitive reaction between Ti and Si with C. The formation of TiC instead of Ti3 SiC2 is thermodynamically favorable and TiC appears always as a secondary phase in the formation of Ti3 SiC2 . Thus, excess amount of Si was used in order to reduce the TiC content in the final material i.e., the initial atomic content was Ti3 Si1.5 C2 . The XRD pattern of the filler compound produced by SHS is depicted in Fig. 1. From the XRD pattern we observe that in addition to the Ti3 Si1.5 C2 compound the minority phases TiC and TiSi2 are present. Ti3 Si1.5 C2 crystallizes in the hexagonal P63 /mmc space group, TiC which crystallizes in the fcc cubic Fm-3m space group, and TiSi2 in the orthorhombic face-centered Fddd space group. The formation of TiC as a secondary product during the combustion synthesis of the MAX phase Ti3 SiC2 has been studied in different works [19,20]. From Rietveld analysis of the XRD spectra of the filler the % mole content of the phases present was determined as Ti3 SiC2 : 58% (77.4 wt%), TiC: 37% (19.2 wt%) and TiSi2 : 5% (3.4 wt%). Converting the mole content of these three phases to atomic ratios we find 3Ti:0.95Si:2.1C. Compared to the initial atomic ratios 3Ti:1.5Si:2C of the mixed powders, this shows loss of Si during the combustion synthesis.
Please cite this article in press as: C. Jiménez, et al., Joining of ceramic matrix composites to high temperature ceramics for thermal protection systems, J Eur Ceram Soc (2015), http://dx.doi.org/10.1016/j.jeurceramsoc.2015.09.038
G Model JECS-10300; No. of Pages 7
ARTICLE IN PRESS C. Jiménez et al. / Journal of the European Ceramic Society xxx (2015) xxx–xxx
3
Fig. 2. SEM micrographs of the SiC ML-Cf /SiC joint fabricated at 25 MPa at magnification ×100 (a) and ×600 (b).
Fig. 3. SEM micrographs of the SiC ML-Cf /SiC joints fabricated at 7 MPa (a) and 2.7 MPa (b) pressure.
3.2. Microstructure and phase analysis of the joints The microstructure of the joints fabricated at the three different pressures is very similar. For example Fig. 2a depicts, at a magnification of ×100, the overview of the SiC ML-Cf /SiC joined structure fabricated at 25 MPa. The Cf /SiC consists of eleven SiC based interlayers, reinforced with carbon fibers, having an average thickness of 300 m. The fiber orientation is at 90◦ between the alternate interlayers. The top interlayer of the Cf /SiC has its carbon fibers vertical to the surface. The vertical to the SiC ML-Cf /SiC interface cracks that are observed in the Cf /SiC material are inherent in the material and are not produced during the joining process. Also few cracks appear in the filler vertical to the interface and in some cases they extend in the SiC multilayer up to its free surface. A more detailed micrograph of the same joint is presented in Fig. 2b at a magnification of ×600. It is observed that the filler wets efficiently both base materials and the joint is sound. The black dots at the lower part of Fig. 2b in the region of the Cf /SiC are the carbon fibers that reinforce the composite ceramic material. The microstructure of the joints fabricated at pressures of 7 and 2.7 MPa as it can be seen from Fig. 3 are quite similar to that of the 25 MPa. For both pressures the interface of the filler with the SiC is continuous and sound. However, the interface with Cf /SiC is, on average, less continuous compared with the joint fabricated at 25 MPa pressure (Fig. 3). The presence of cracks (Fig. 2a and b) is attributed in the release of stresses, accumulated during the cooling in the joining process. These arise from (a) the difference in the CTE between the base material and the filler and (b) lack of flatness of the SiC ML. In gen-
eral the vertical to the interface cracks are not expected to have a detrimental effect on the shear strength of the joint. It is also observed that as the applied pressure decreases the number density of the cracks also decreases. Fig. 4 depicts the SiC ML-filler and the filler-Cf /SiC interfaces for the joint fabricated at 25 MPa. Both interfaces are continuous and sound. The black areas in the SiC are pores or graphite inherent in the SiC material. In the filler region in all the joints two main areas are observed, (a) the light grey area F-I which is the main filler area and (b) dark islands F-II (Figs. 2 and 3). The main filler area F-I is predominant of the order of 60–90% of the joints cross sectional area. EDS measurements were carried out of the overall joint area and the areas F-I and F-II (Table 1) and the atomic percentage of the different atomic species (C, Si, Ti) were determined. The average atomic percentage in the table arises from measurements from different regions of the same type. Also in Table 1 the main possible phases for each region are given. The overall cross section, for the joint fabricated at 2.7 MPa, contains a mole percentage of the Ti3 SiC2 , TiC and TiSi2 phases of 73%, 19% and 8%, respectively. When the pressure increases to 7 and 25 MPa, the Ti3 SiC2 mole content increases to about 82%, whereas the TiSi2 phase almost disappears. This increase of the Ti3 SiC2 may be attributed to the reaction of TiC with TiSi2 to form the MAX phase during the joining process at high pressures according to the reaction 2TiC+TiSi2 → Ti3 SiC2 + Si
Please cite this article in press as: C. Jiménez, et al., Joining of ceramic matrix composites to high temperature ceramics for thermal protection systems, J Eur Ceram Soc (2015), http://dx.doi.org/10.1016/j.jeurceramsoc.2015.09.038
G Model
ARTICLE IN PRESS
JECS-10300; No. of Pages 7
C. Jiménez et al. / Journal of the European Ceramic Society xxx (2015) xxx–xxx
4
Fig. 4. SiC ML-filler (a) and filler-Cf /SiC (b) interfaces of the SiC ML-Cf /SiC joint fabricated at 25 MPa pressure. Table 1 EDS analysis of the filler area in the SiC ML-Cf /SiC joint fabricated at 25, 7 and 2.7 MPa pressure. The percentage of each phase is in mole%. Pressure (MPa) 25
7
2.7
C (at%)
Si (at%)
Ti (at%)
Main phases
Overall cross section
49.9 ± 1.0
11.7 ± 2.0
38.4 ± 2.8
F-I light grey
44.1 ± 1.4
10.7 ± 1.0
45.2 ± 0.5
F-II dark grey
52.5 ± 0.9
43.3 ± 1.1
4.2 ± 0.2
82% Ti3 SiC2 17.5% TiC 0.5% TiSi2 Ti3 SiC2 TiC Si–C TiC
Overall cross section
47.2 ± 0.5
10.7 ± 0.3
42.1 ± 0.5
F-I light grey
44.8 ± 0.4
8.0 ± 0.6
47.2 ± 0.3
F-II dark grey
48.6 ± 2.3
47.6 ± 2.6
3.8 ± 1.1
Overall cross section
53.0 ± 2.1
18.2 ± 0.7
28.8 ± 1.9
F-I light grey (close to SiC ML interface) F-I light grey (close to CMC interface) F-II dark grey
49.3 ± 0. 1
16.5 ± 5.2
34.3 ± 5.3
47.8 ± 0.5
4.8 ± 1.5
47.4 ± 1.7
Ti3 SiC2 TiC
51.7 ± 0.7
38.3 ± 1.0
10.0 ± 0.6
Si–C TiC
The above reaction is further corroborated by the small decrease in the TiC content when the pressure is larger than 2.7 MPa. Furthermore, the carbon atomic content in the overall cross section, when compared with the carbon content in the filler as determined by XRD measurements, is characterized by a considerable increase. This increase is attributed to carbon diffusion mainly from the SiC multilayer which is rich in carbon as it is discussed in Section 3.3. F-I area is rich in Ti3 SiC2 and TiC phases with similar atomic concentrations for the joints fabricated at 7 and 25 MPa. For the joint fabricated at 2.7 MPa the F-I area shows an increase of the Ti content accompanied by a parallel decrease of the Si content as we move from the SiC ML towards the Cf /SiC interface. This shows the existence of more TiC and Ti3 SiC2 in the half lower part of the joint close to the Cf /SiC interface. It is noted that irrespectively of the type of the majority phase, TiC or Ti3 SiC2 , in the joint area (F-I), all the joints are sound with continuous interfaces. The different atomic element concentration in the F-I area of the joint fabricated at 2.7 MPa with that of the joints fabricated at 7 and 25 MPa (Table 1) possibly arises from the difference in the applied pressure, as it is the only parameter changed in the joining cycle. Thus, the difference in the chemical evolution of the filler might be attributed to the
82.6% Ti3 SiC2 17.4% TiC 0.0% TiSi2 Ti3 SiC2 TiC Si–C TiC 73% Ti3 SiC2 19% TiC 8% TiSi2 Ti3 SiC2 TiC
effect of the pressure on the liquid phase of the TiSi2 as in the joining process the temperature of 1600 ◦ C is reached which is above the liquidus (1480 ◦ C) of TiSi2 . F-II area (dark grey islands) which covers a small part of the joint section is rich in Si and C and possibly SiC has been formed or has been diffused from SiC ML. The chemical composition of the F-II area is quite similar for all joints. The dispersion of the F-II phase in the filler region is quite inhomogeneous with a large scattering in the size of the Si–C rich islands. The joints fabricated at 7 and 25 MPa are characterized by a smaller size of the Si–C rich islands compared with the 2.7 MPa joint for which large islands are observed towards the SiC ML interface. These large islands indicate a possible diffusion of SiC from SiC ML as it has been pointed above. The part of the joint adjacent to the Cf /SiC is characterized by a finer dispersion of this phase.
3.3. Re-entry tests SiC ML-Cf /SiC joints fabricated with an applied pressure of 2.7 MPa were subject to re-entry tests up to 5 cycles under vacuum. The tests were performed using two different temperature
Please cite this article in press as: C. Jiménez, et al., Joining of ceramic matrix composites to high temperature ceramics for thermal protection systems, J Eur Ceram Soc (2015), http://dx.doi.org/10.1016/j.jeurceramsoc.2015.09.038
G Model
ARTICLE IN PRESS
JECS-10300; No. of Pages 7
C. Jiménez et al. / Journal of the European Ceramic Society xxx (2015) xxx–xxx 2000
400 200 0 00:00
02:00
o
Intermediate sample inspection
600
(b)
1400 1200 1000 800 600 400 200 0
04:00
06:00
Test Time (hh:mm)
08:00
Heater temperature CMC back temperature
1600
00:00
02:00
TC failure
800
Heater temperature CMC back temperature
Temperature ( C)
1000
1800
(a) Intermediate sample inspection
o
Temperature ( C)
1400 1200
5
04:00
06:00
08:00
10:00
Test Time (hh:mm)
Fig. 5. Temperature profiles used in the re-entry tests at 1391 ◦ C (a) and at 1794 ◦ C (b). (For interpretation of the references to color in text, the reader is referred to the web version of this article.)
Fig. 6. The CMC surface before (left) and after (right) five re-entry cycles at 1391 and 1794 ◦ C.
Fig. 7. The SiC ML surface before (left) and after (right) five re-entry cycles at 1391 and 1794 ◦ C.
Fig. 8. The surface of the SiC ML-Cf /SiC joint after the tests at 1391 ◦ C (a) and 1794 ◦ C (b).
Please cite this article in press as: C. Jiménez, et al., Joining of ceramic matrix composites to high temperature ceramics for thermal protection systems, J Eur Ceram Soc (2015), http://dx.doi.org/10.1016/j.jeurceramsoc.2015.09.038
G Model
ARTICLE IN PRESS
JECS-10300; No. of Pages 7
C. Jiménez et al. / Journal of the European Ceramic Society xxx (2015) xxx–xxx
6
Table 2 EDS analysis of the cross section of the SiC ML-Cf /SiC joints after the re-entry tests. The percentage of each phase is in mole%. C (at%)
Si (at%)
Ti (at%)
Main phases
Overall cross section
57.5 ± 1.4
18.9 ± 0.7
23.6 ± 1.2
F-I light grey
46.8 ± 0.5
7.0 ± 0.7
46.2 ± 0.7
F-II dark grey
47.7 ± 0.8
51.8 ± 0.9
0.5 ± 0.2
68.5% Ti3 SiC2 19.5% TiC 12.0% TiSi2 Ti3 SiC2 TiC Si–C
Overall cross section
52.8 ± 0.9
12.9 ± 1.0
34.3 ± 1.1
F-I light grey (close to SiC-ML interface) F-I light grey (close to CMC interface) F-II dark grey
42.2 ± 1.1
8.4 ± 1.3
46.4 ± 0.3
48.1 ± 0.2
1.6 ± 0.2
50.3 ± 0.4
TiC
45.4 ± 0.5
52.9 ± 0.9
1.6 ± 1.4
Si–C
SiC as fabricated o tested at 1391 C o tested 1794 C
SiC SiC SiC
30
SiC
40
SiC
SiC
SiC SiC C
20
50
60
SiC
SiC
SiC
SiC
profiles of the ARV, profile (a) with a peak temperature at 1391 ◦ C (Fig. 5a) and profile (b) with a peak temperature at 1794 ◦ C (Fig. 5b). During the test the backside temperature of the joint, i.e., the back side of the Cf /SiC is monitored using a K-type thermocouple. The red line in Fig. 5 depicts the thermocouple temperature attached at the back side of the Cf /SiC. The black line in Fig. 5 represents the temperature at the SiC top surface. This temperature is derived from the heat fluxes applied by the furnace heater during the tests and under the assumption that there are no considerable thermal losses. During the 2nd cycle of the test at 1794 ◦ C the temperature went beyond 1400 ◦ C, and the type K thermocouple failed and as a result the monitoring of the temperature at the back side of the CMC was interrupted. After the 1st cycle, the samples had been removed from the rig for optical inspection, and as no changes were found, 4 more cycles were carried out. The sample surface before and after the tests is presented in Figs. 6 and 7. The cracks appearing on the SiC ML surface pre-exist the re-entry tests. There is only a change in color that originates partially from an artefact of the scanner used for making the post test photographs and partially from graphite deposition at the surface as it will be explained below. The optical microscopy examination of the joints after the tests showed no detectable defects. Non-destructive testing using ultrasonic inspection was not possible as the cracks in the ceramic plate result in large “blackout” areas around the crack zones. The surface of the SiC ML that was subject to the thermal load was examined by SEM and XRD measurements. Fig. 8a and b presents the surface of the SiC ML after the test at 1391 and 1794 ◦ C, respectively. A high content of graphite is found on the surface of SiC which is not
82.1% Ti3 SiC2 17.4% TiC 0.5% TiSi2 Ti3 SiC2 TiC
C
1794
SiO2
1391
Intensity (arbitrary units)
Temperature (◦ C)
70
80
90
2theta (deg.) Fig. 9. XRD spectra of the SiC ML surface as fabricated (black line), tested at 1391 ◦ C (blue line) and at 1794 ◦ C (red line). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
related with the re-entry tests. The high graphite content is confirmed by XRD measurements on the SiC ML surface of the tested joints presented in Fig. 9 which show the presence of hexagonal graphite (SG: P63/mmc). A possible explanation for the presence of carbon is the thermal decomposition of SiC, where Si atoms diffuse
Fig. 10. SEM micrograph of the SiC ML-Cf /SiC cross section after the re-entry test at 1391 (a) and 1794 ◦ C (b).
Please cite this article in press as: C. Jiménez, et al., Joining of ceramic matrix composites to high temperature ceramics for thermal protection systems, J Eur Ceram Soc (2015), http://dx.doi.org/10.1016/j.jeurceramsoc.2015.09.038
G Model JECS-10300; No. of Pages 7
ARTICLE IN PRESS C. Jiménez et al. / Journal of the European Ceramic Society xxx (2015) xxx–xxx
outwards from the underlying SiC substrate and desorb subsequently, and the released carbon atoms diffuse and nucleate to form carbon rich areas. Annealing of SiC (up to 1400 ◦ C) under vacuum is a method currently used for the preparation of graphene-like layers [21]. The dark areas (S) in Fig. 8b are carbon and it comes from the graphite tiles that were used to place the structure to be joined in the hot press and remained on the surface of the SiC after the joining. The microstructure of the cross section of the joints after the thermal test at both temperatures shows the same general characteristics as in the as fabricated joints (Fig. 10). Table 2 presents the results from the EDS analysis of the cross sections of the joints after the re-entry tests. At the test at 1391 ◦ C the phase content of the overall cross section is similar to that of the as fabricated joint at 2.7 MPa (see Table 1). However, when the test temperature increases to 1794 ◦ C it has an effect analogous to that of pressure, i.e., the MAX phase content increases by around 20%, whereas the TiSi2 phase almost vanishes. Similarly, as before, this can be explained by the reaction of TiSi2 with TiC. This is to be expected because TiSi2 becomes liquid as the re-entry test temperature (1794 ◦ C) is above its liquidus (1480 ◦ C) and it has been reported [22] that Ti–Si melt reacts with TiC to produce Ti3 SiC2 . Furthermore, as it can be observed in Fig. 10, when the test temperature increases to 1794 ◦ C the Si–C rich islands are formed mainly in the middle part of the joint as in the joints fabricated at 25 MPa. This indicates that the test temperature has an effect similar to that of pressure, i.e., increased Si-C diffusion from the SiC ML. In the F-I (light grey) region of the joint tested at 1794 ◦ C there is a differentiation with respect to the Si and Ti content between the upper half part of the joint close to the SiC ML interface and the other half of the joint close to the CMC interface (Table 2). The upper part is characterized by a higher content of Si and lower content of Ti than those of the bottom part of the joint. The same effect, to a lesser degree, was observed in the as fabricated joint at 2.7 MPa. 4. Conclusions Successful joining of SiC multilayers to Cf /SiC ceramic composite materials using the ternary carbide Ti3 Si1.5 C2 and employing diffusion brazing bonding and pressures in the range 2.7–25 MPa was accomplished. As the fabrication pressure increases from 2.7 MPa to 7 MPa and above TiC reacts with TiSi2 to form the MAX phase Ti3 SiC2 . Cracks appear in the filler area and few of them can extend up to the free surface of the SiC multilayer. The cracks are attributed to the brittle behavior of the filler, the mismatch in the coefficient of thermal expansion between the joined parts and the filler and to the lack of flatness of the SiC specimens. These cracks are vertical to the SiC ML-Cf /SiC interface and are not detrimental to the shear strength of the joint. Thermal tests under two re-entry conditions, in vacuum, have not shown any detectable damage. Therefore these structures have the prospective to be utilized as re-usable thermal protection systems for re-entry applications. Acknowledgements This work has been performed within the framework of the European Project “SMARTEES (G.A. n. 262749) with the financial support by the European Community. The paper only reflects the view of the authors and the European Community is not liable for
7
any use of the information contained therein. References [1] I. Sakraker, C.O. Asma, Experimental investigation of passive/active oxidation behavior of SiC based ceramic thermal protection materials exposed to high enthalpy plasma, J. Eur. Ceram. Soc. 33 (2013) 351–359. [2] H. Kiyono, S. Shimada, K.J.D. MacKenzie, Kinetic and magic angle spinning-nuclear magnetic resonance studies of wet oxidation of beta-sialon powders, J. Electrochem. Soc. 148 (2001) B86–B91. [3] B.V. Cockeram, Flexural strength and shear strength of silicon carbide to silicon carbide joints fabricated by a molybdenum diffusion bonding technique, J. Am. Ceram. Soc. 88 (2005) 1892–1899. [4] Yutai Katoh, Lance L. Snead, Ting Chenga, Chunghao Shih, W. Daniel Lewis, Takaaki Koyanagi, Tatsuya Hinoki, Charles H. Henager Jr., Monica Ferraris, Radiation-tolerant joining technologies for silicon carbide ceramics and composites, J. Nucl. Mater. 448 (2014) 497–511. [5] Monica Ferraris, Milena Salvo, Valentina Casalegno, Andrea Ciampichetti, Federico Smeacetto, Massimo Zucchetti, Monica Ferraris, Joining of machined SiC/SiC composites for thermonuclear fusion reactors, J. Nucl. Mater. 375 (2008) 410–415. [6] M. Singh, R. Asthana, Joining of zirconium diboride-based ultra high-temperature ceramic composites using metallic glass interlayers, Mater. Sci. Eng. A 460–461 (2007) 153–162. [7] Mrityunjay Singh, Rajiv Asthana, Joining of ZrB2 -based ultra-high-temperature ceramic composites to Cu-clad-molybdenum for advanced aerospace applications, Int. J. Appl. Ceram. Technol. 6 (2009) 113–133. [8] B.H. Rabin, G.A. Moore, Reaction processing and properties of SiC-to-SiC joints, Mater. Res. Soc. Symp. Proc. 314 (1993) 197–203. [9] H. Dong, S. Li, Y. Teng, W. Ma, Joining of SiC ceramic-based materials with ternary carbide Ti3 SiC2 , Mater. Sci. Eng.: B 176 (1) (2011) 60–64. [10] Stefano Rizzoa, Salvatore Grassob, Milena Salvoa, Valentina Casalegnoa, Michael J. Reeceb, Monica Ferraris, Joining of C/SiC composites by spark plasma sintering technique, J. Nucl. Mater. 34 (2014) 903–913. [11] Alberto Ortona, Simone Pusterla, Sandro Gianella, An integrated assembly method of sandwich structured ceramic matrix composites, J. Eur. Ceram. Soc. 31 (2011) 1821–1826. [12] H. Böhrk, U. Beyermann, Secure tightening of a CMC fastener for the heat shield of re-entry vehicles, Compos. Struct. 92 (2010) 107–112. [13] Philippe Berthe, The advanced re-entry vehicle (ARV)—development step from ATV towards manned transportation systems, ESA/ESTEC, in: Proceedings of the 3rd International ARA days, 2–4 May, Arcachon, France, 2011. [14] M.W. Barsoum, T. El-Raghy, C.J. Rawn, W.D. Porter, H. Wang, E.A. Payzant, C.R. Hubbard, Thermal properties of Ti3 SiC2 , J. Phys. Chem. Solids 60 (1999) 429–439. [15] G. Motz, S. Schmidt, S. Beyer, The PIP-process: Precursor Properties and Applications, Ceramic Matrix Composites, in: W. Krenkel (Ed.), Wiley-VCH, Germany, 2008, pp. 165–186. [16] W.S. Yang, S. Biamino, E. Padovano, M. Pavese, S. Marchisio, G. D’Amico, S. Ceresa MioX, X. Chen, P. Fino, C. Badini, Thermophysical properties of SiC multilayer prepared by tape casting and pressureless sintering, Compos. Struct. 96 (2013) 469–475. [17] J. Barcena, M. Lagos, I. Agote, C. Jimenez, X. Hernandez, C. Badini, E. Padovano, S. Gianella, D. Gaia, V. Liedtke, K. Mergia, S. Messoloras, P. Yialouris, Y. Panayiotatos, A. Ortona, C. D’Angelo, C. Wilhelmi, SMARTEES FP7 space project—towards a new tps reusable concept for atmospheric reentry from low earth orbit, in: 7th European Workshop on Thermal Protection Systems and Hot Structures, 8–10 April 2013, ESA-ESTEC, Noordwijk, The Netherlands, 2013. [18] T. Salmon, X. Vo, M. Bottacini, TPS Architecture on ARV Proceedings of the 7th European Symposium on Aerothermodynamics, in: L. Ouwehand, Noordwijk, Netherlands, European Space Agency, 2011, id.120. [19] S.G. Vadchenko, A.E. Sytschev, D. Yu Kovalev, A.S. Shchukin, S.V. Konovalikhin, Self-propagating high-temperature synthesis in the Ti–Si–C system in features of product patterning, Nanotechnol. Russ. 10 (1) (2015) 67–74. [20] S.V. Konovalikhin, D. Yu Kovalev, A.E. Sytschev, S.G. Vadchenko, A.S. Shchukin, Formation of nanolaminate structures in the Ti–Si–C system: a crystallochemical study, Int. J. Self-Propag. High-Temp. Synth. 23 (2014) 216–220. [21] Qingyan Wang, Wenhao Zhang, Lili Wang, Ke He, Xucun Ma, Qikun Xue, Large-scale uniform bilayer graphene prepared by vacuum graphitization of 6H-SiC(0001) substrates, J. Phys.: Condens. Matter 25 (2013) 095002. [22] M. Abu, J. Mohamed, Z. Ahmad, Effect of excess silicon on the formation of Ti3 SiC2 using free Ti/Si/C powders synthesized via arc melting, ISRN Ceram. 2012 (2012) 10, Article ID 341285.
Please cite this article in press as: C. Jiménez, et al., Joining of ceramic matrix composites to high temperature ceramics for thermal protection systems, J Eur Ceram Soc (2015), http://dx.doi.org/10.1016/j.jeurceramsoc.2015.09.038
ARTICLE IN PRESS
JECS-10300; No. of Pages 7
Journal of the European Ceramic Society xxx (2015) xxx–xxx
Contents lists available at www.sciencedirect.com
Journal of the European Ceramic Society journal homepage: www.elsevier.com/locate/jeurceramsoc
Joining of ceramic matrix composites to high temperature ceramics for thermal protection systems C. Jiménez a , K. Mergia b,∗ , M. Lagos a , P. Yialouris b , I. Agote a , V. Liedtke c , S. Messoloras b , Y. Panayiotatos b , E. Padovano d , C. Badini d , C. Wilhelmi e , J. Barcena a a
Tecnalia Research & Innovation, Industry and Transport Division, Mikeletegi Pasealekua, 2, Donostia-San Sebastian, Spain N.C.S.R “Demokritos”, Institute of Nuclear and Radiological Science and Technology, Energy and Safety, Aghia Paraskevi, 15310 Athens, Greece c Aerospace & Advanced Composites GmbH, Viktor-Kaplan-Straße 2, A-2700 Wiener Neustadt, Austria d Politecnico di Torino, Department of Applied Science and Technology, Corso Duca degli Abruzzi, 24, 10129 Torino, Italy e Airbus Defence and Space GmbH, Dept. TCC 2/CTOIWMS, 81663 Munich, Germany b
a r t i c l e
i n f o
Article history: Received 12 May 2015 Received in revised form 22 September 2015 Accepted 27 September 2015 Available online xxx Keywords: Ceramic matrix composite Silicon carbide Joining Thermal tests Aerospace
a b s t r a c t The current work reports a novel approach for the integration of external protective SiC multilayers with ceramic matrix composite (Cf /SiC) with the view of application in aerospace heat protection systems. The integration method is based on diffusion brazing bonding. As a joining agent the MAX-Phase Ti3 SiC2 , produced by self-propagating high temperature synthesis, has been employed. The pressure applied during the joining process and its effect on the microstructure of the integrated structure is discussed. Microstructural analysis of the resulting joints is conducted using scanning electron microscopy coupled with energy dispersive spectroscopy and X-ray diffraction measurements. Analysis of the joints showed that the bonds are uniform, dense, with few crack vertical to the interface which are not detrimental for the performance of the joints. Ground re-entry tests showed that the joints survive 5 re-entry cycles at 1391 and 1794 ◦ C without any detectable damage. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction One of the key technologies of today’s aerospace applications refers to the effective protection of components subject to large heat loads like the heat shields of re-entry vehicles or the inner walls of combustion chambers. This issue is of great importance for future missions where considerably increased energy densities are expected. Advanced ceramic matrix composite (CMC) materials based on SiC matrix reinforced with carbon fibers (Cf /SiC) are key materials for aerospace applications since they present superior mechanical properties and resistance against high temperatures and being at the same time lightweight and cost effective. For the application of these bulk ceramics in hostile environments appropriate coatings have to be developed that will prevent the oxidation of the carbon fibers. The current thermal protection systems are based on the use of ceramic matrix composite materials coated with SiC [1] or other carbide based materials. SiC exhibits excellent oxidation resistance
∗ Corresponding author. Fax: +30 210 6545496. E-mail address: [email protected] (K. Mergia).
at high temperatures, because the formed glassy silica films prevent oxygen diffusion very efficiently and thus serve as protection against further oxidation. However, the amorphous silica at temperatures above 1200 ◦ C crystallizes to cristobalite causing surface cracking and also reacts with water or Na/K vapor, resulting in severe degradation of the silica film [2]. Alternatively to SiC coatings ultra high temperature SiC based multilayer ceramics can be employed in combination with CMCs in order to provide integrated systems with enhanced performance under the most severe re-entry conditions. The advantage in using multilayered ceramics stems from the capability of adjusting their architecture and chemical composition e.g., by incorporating ZrB2 layers at the outer structure, and thus tuning the functionality of these protection layers to specific re-entry conditions i.e., the required service temperature and the oxidative environment these structures have to withstand. Therefore, advances in joining science and technology of CMC-ultra high temperature and oxidation resistant ceramics to SiC based multilayer ceramics are important in order to integrate these systems and realize the benefits of these advanced materials in aerospace applications. A number of techniques for joining monolithic ceramic materials or CMCs to themselves have been developed such as solid state
http://dx.doi.org/10.1016/j.jeurceramsoc.2015.09.038 0955-2219/© 2015 Elsevier Ltd. All rights reserved.
Please cite this article in press as: C. Jiménez, et al., Joining of ceramic matrix composites to high temperature ceramics for thermal protection systems, J Eur Ceram Soc (2015), http://dx.doi.org/10.1016/j.jeurceramsoc.2015.09.038
G Model JECS-10300; No. of Pages 7
ARTICLE IN PRESS C. Jiménez et al. / Journal of the European Ceramic Society xxx (2015) xxx–xxx
sintering of the ceramics, diffusion bonding using various active fillers [3,4], transient eutectic-phase routes [4], glass–ceramic joining [5], metallic braze-based joining [6,7], SiC reaction forming [8], MAX-phase bonding [4,9], spark plasma sintering [10], integrated assembly method [11] or mechanical fastening [12]. However, to our knowledge no joining procedure for CMCs and SiC based multilayers systems has been reported so far. The current work reports on a approach for the integration of external protective ceramic SiC based multilayers (SiC ML) with a thermostructural ceramic matrix composite (Cf /SiC). This structure is a part of a more complex system for a powerful thermal protection system under a reference mission (ESA’s Advanced Re-entry Vehicle) [13]. The integration method of the two materials is based on the diffusion brazing bonding process. As a joining agent a filler metal based on a MAX-Phase Ti3 SiC2 , produced by self propagating high temperature synthesis, has been employed. Ti3 SiC2 has good stability at elevated temperatures and it presents both metallic and ceramic properties and exhibits plastic behavior. It has a melting point higher than 3000 ◦ C and a coefficient of thermal expansion (CTE) 9.12 × 10−6 /K between 25 and 1200 ◦ C [14]. It has been used in the past for joining of SiC based ceramics for application in the temperature range 1200–1600 ◦ C [9]. The process parameters of the joining method and their effect on the microstructure of the integrated structure are discussed. The structure and microstructure of the assembly has been investigated by optical microscopy, scanning electron microscopy with energy dispersive analysis and X-ray diffraction measurements. Results from thermal tests under re-entry conditions are presented and discussed with respect to the envisaged applications.
2. Materials and methods The Cf /SiC (SICARBON) ceramic composites were supplied by Airbus Defence and Space GmbH [15]. They consist of carbon fibers embedded in a silicon carbide matrix. The production process of this material is based on the polymer infiltration pyrolysis process (PIP). The infiltration of the carbon fibers with a pre-ceramic polymer-based and powder-filled slurry system is performed by liquid polymer infiltration (LPI) via filament winding. From the supplied material samples of 40 × 40 mm2 were cut and used for all the experiments. The SiC multilayers were fabricated by the tape casting technique followed by pressureless sintering [16]. The processing method involved several steps: SiC slurry preparation; tape casting; stacking, debinding and pressureless sintering. The SiC multilayers used in the current study consist of eleven dense SiC layers. For the joining of the SiC to SICARBON, the ternary MAX compound Ti3 Si1.5 C2 was used as a filler metal. The material was produced by self-propagating high temperature synthesis (SHS) in a reaction furnace and in argon atmosphere using titanium, silicon and carbon powders with atomic ratios 3Ti:1.5Si:2C. After the SHS process a foam was obtained and this foam was milled and sieved under 63 m. The crystal structure of the filler was assessed by X-ray diffraction (XRD) measurements. The joints were fabricated employing the diffusion brazing technique in a high vacuum hot press furnace at 1600 ◦ C. The pressures used were 2.7, 7 and 25 MPa. For the joints fabricated at 2.7 and 7 MPa the pressure was applied at room temperature and kept constant during the thermal cycle. For the joint fabricated at 25 MPa at the beginning of the thermal cycle and up to 1400 ◦ C a lower pressure of 15 MPa was applied up and for the rest of the thermal cycle during heating and cooling the pressure of 25 MPa was applied. A heating rate of 15 ◦ C/min up to 1400 ◦ C, followed by a heating rate of 10 ◦ C/min up to 1600 ◦ C and a slow cooling rate of 5 ◦ C/min were applied.
Ti3SiC2 TiC Si2Ti
Intensity (a.u.)
2
20
30
40
50
60
70
80
2Θ (deg.) Fig. 1. XRD pattern of Ti3 Si1.5 C2 .
The XRD patterns were measured using a Bruker D8 diffractometer equipped with a Cu K␣ radiation, a parallel beam stemming from a Göbel mirror and a Våntec position sensitive detector with 9◦ angular acceptance. The microstructure of the joints was examined using FEI Quanta Inspect scanning electron microscopy (SEM) coupled with energy dispersive X-ray spectroscopy (EDS). For the EDS analysis, different accelerator voltages were used in the range between 12 and 20 kV and a standardless quantitative analysis was performed employing the ZAF correction approach. Tests of the SiC-ML/CMC under re-entry conditions, i.e. temperature profiles under vacuum, were carried out at a ground re-entry test rig [17] at a maximum temperature of 1794 and 1391 ◦ C for 5 cycles each. The tests were performed under vacuum and the thermal profiles used correspond to two control points of the ESA’s Advanced Re-entry vehicle (ARV) [18].
3. Results and discussion 3.1. Filler composition and structure The filler metal was produced by SHS. In the SHS process for the formation of Ti3 SiC2 , there is a competitive reaction between Ti and Si with C. The formation of TiC instead of Ti3 SiC2 is thermodynamically favorable and TiC appears always as a secondary phase in the formation of Ti3 SiC2 . Thus, excess amount of Si was used in order to reduce the TiC content in the final material i.e., the initial atomic content was Ti3 Si1.5 C2 . The XRD pattern of the filler compound produced by SHS is depicted in Fig. 1. From the XRD pattern we observe that in addition to the Ti3 Si1.5 C2 compound the minority phases TiC and TiSi2 are present. Ti3 Si1.5 C2 crystallizes in the hexagonal P63 /mmc space group, TiC which crystallizes in the fcc cubic Fm-3m space group, and TiSi2 in the orthorhombic face-centered Fddd space group. The formation of TiC as a secondary product during the combustion synthesis of the MAX phase Ti3 SiC2 has been studied in different works [19,20]. From Rietveld analysis of the XRD spectra of the filler the % mole content of the phases present was determined as Ti3 SiC2 : 58% (77.4 wt%), TiC: 37% (19.2 wt%) and TiSi2 : 5% (3.4 wt%). Converting the mole content of these three phases to atomic ratios we find 3Ti:0.95Si:2.1C. Compared to the initial atomic ratios 3Ti:1.5Si:2C of the mixed powders, this shows loss of Si during the combustion synthesis.
Please cite this article in press as: C. Jiménez, et al., Joining of ceramic matrix composites to high temperature ceramics for thermal protection systems, J Eur Ceram Soc (2015), http://dx.doi.org/10.1016/j.jeurceramsoc.2015.09.038
G Model JECS-10300; No. of Pages 7
ARTICLE IN PRESS C. Jiménez et al. / Journal of the European Ceramic Society xxx (2015) xxx–xxx
3
Fig. 2. SEM micrographs of the SiC ML-Cf /SiC joint fabricated at 25 MPa at magnification ×100 (a) and ×600 (b).
Fig. 3. SEM micrographs of the SiC ML-Cf /SiC joints fabricated at 7 MPa (a) and 2.7 MPa (b) pressure.
3.2. Microstructure and phase analysis of the joints The microstructure of the joints fabricated at the three different pressures is very similar. For example Fig. 2a depicts, at a magnification of ×100, the overview of the SiC ML-Cf /SiC joined structure fabricated at 25 MPa. The Cf /SiC consists of eleven SiC based interlayers, reinforced with carbon fibers, having an average thickness of 300 m. The fiber orientation is at 90◦ between the alternate interlayers. The top interlayer of the Cf /SiC has its carbon fibers vertical to the surface. The vertical to the SiC ML-Cf /SiC interface cracks that are observed in the Cf /SiC material are inherent in the material and are not produced during the joining process. Also few cracks appear in the filler vertical to the interface and in some cases they extend in the SiC multilayer up to its free surface. A more detailed micrograph of the same joint is presented in Fig. 2b at a magnification of ×600. It is observed that the filler wets efficiently both base materials and the joint is sound. The black dots at the lower part of Fig. 2b in the region of the Cf /SiC are the carbon fibers that reinforce the composite ceramic material. The microstructure of the joints fabricated at pressures of 7 and 2.7 MPa as it can be seen from Fig. 3 are quite similar to that of the 25 MPa. For both pressures the interface of the filler with the SiC is continuous and sound. However, the interface with Cf /SiC is, on average, less continuous compared with the joint fabricated at 25 MPa pressure (Fig. 3). The presence of cracks (Fig. 2a and b) is attributed in the release of stresses, accumulated during the cooling in the joining process. These arise from (a) the difference in the CTE between the base material and the filler and (b) lack of flatness of the SiC ML. In gen-
eral the vertical to the interface cracks are not expected to have a detrimental effect on the shear strength of the joint. It is also observed that as the applied pressure decreases the number density of the cracks also decreases. Fig. 4 depicts the SiC ML-filler and the filler-Cf /SiC interfaces for the joint fabricated at 25 MPa. Both interfaces are continuous and sound. The black areas in the SiC are pores or graphite inherent in the SiC material. In the filler region in all the joints two main areas are observed, (a) the light grey area F-I which is the main filler area and (b) dark islands F-II (Figs. 2 and 3). The main filler area F-I is predominant of the order of 60–90% of the joints cross sectional area. EDS measurements were carried out of the overall joint area and the areas F-I and F-II (Table 1) and the atomic percentage of the different atomic species (C, Si, Ti) were determined. The average atomic percentage in the table arises from measurements from different regions of the same type. Also in Table 1 the main possible phases for each region are given. The overall cross section, for the joint fabricated at 2.7 MPa, contains a mole percentage of the Ti3 SiC2 , TiC and TiSi2 phases of 73%, 19% and 8%, respectively. When the pressure increases to 7 and 25 MPa, the Ti3 SiC2 mole content increases to about 82%, whereas the TiSi2 phase almost disappears. This increase of the Ti3 SiC2 may be attributed to the reaction of TiC with TiSi2 to form the MAX phase during the joining process at high pressures according to the reaction 2TiC+TiSi2 → Ti3 SiC2 + Si
Please cite this article in press as: C. Jiménez, et al., Joining of ceramic matrix composites to high temperature ceramics for thermal protection systems, J Eur Ceram Soc (2015), http://dx.doi.org/10.1016/j.jeurceramsoc.2015.09.038
G Model
ARTICLE IN PRESS
JECS-10300; No. of Pages 7
C. Jiménez et al. / Journal of the European Ceramic Society xxx (2015) xxx–xxx
4
Fig. 4. SiC ML-filler (a) and filler-Cf /SiC (b) interfaces of the SiC ML-Cf /SiC joint fabricated at 25 MPa pressure. Table 1 EDS analysis of the filler area in the SiC ML-Cf /SiC joint fabricated at 25, 7 and 2.7 MPa pressure. The percentage of each phase is in mole%. Pressure (MPa) 25
7
2.7
C (at%)
Si (at%)
Ti (at%)
Main phases
Overall cross section
49.9 ± 1.0
11.7 ± 2.0
38.4 ± 2.8
F-I light grey
44.1 ± 1.4
10.7 ± 1.0
45.2 ± 0.5
F-II dark grey
52.5 ± 0.9
43.3 ± 1.1
4.2 ± 0.2
82% Ti3 SiC2 17.5% TiC 0.5% TiSi2 Ti3 SiC2 TiC Si–C TiC
Overall cross section
47.2 ± 0.5
10.7 ± 0.3
42.1 ± 0.5
F-I light grey
44.8 ± 0.4
8.0 ± 0.6
47.2 ± 0.3
F-II dark grey
48.6 ± 2.3
47.6 ± 2.6
3.8 ± 1.1
Overall cross section
53.0 ± 2.1
18.2 ± 0.7
28.8 ± 1.9
F-I light grey (close to SiC ML interface) F-I light grey (close to CMC interface) F-II dark grey
49.3 ± 0. 1
16.5 ± 5.2
34.3 ± 5.3
47.8 ± 0.5
4.8 ± 1.5
47.4 ± 1.7
Ti3 SiC2 TiC
51.7 ± 0.7
38.3 ± 1.0
10.0 ± 0.6
Si–C TiC
The above reaction is further corroborated by the small decrease in the TiC content when the pressure is larger than 2.7 MPa. Furthermore, the carbon atomic content in the overall cross section, when compared with the carbon content in the filler as determined by XRD measurements, is characterized by a considerable increase. This increase is attributed to carbon diffusion mainly from the SiC multilayer which is rich in carbon as it is discussed in Section 3.3. F-I area is rich in Ti3 SiC2 and TiC phases with similar atomic concentrations for the joints fabricated at 7 and 25 MPa. For the joint fabricated at 2.7 MPa the F-I area shows an increase of the Ti content accompanied by a parallel decrease of the Si content as we move from the SiC ML towards the Cf /SiC interface. This shows the existence of more TiC and Ti3 SiC2 in the half lower part of the joint close to the Cf /SiC interface. It is noted that irrespectively of the type of the majority phase, TiC or Ti3 SiC2 , in the joint area (F-I), all the joints are sound with continuous interfaces. The different atomic element concentration in the F-I area of the joint fabricated at 2.7 MPa with that of the joints fabricated at 7 and 25 MPa (Table 1) possibly arises from the difference in the applied pressure, as it is the only parameter changed in the joining cycle. Thus, the difference in the chemical evolution of the filler might be attributed to the
82.6% Ti3 SiC2 17.4% TiC 0.0% TiSi2 Ti3 SiC2 TiC Si–C TiC 73% Ti3 SiC2 19% TiC 8% TiSi2 Ti3 SiC2 TiC
effect of the pressure on the liquid phase of the TiSi2 as in the joining process the temperature of 1600 ◦ C is reached which is above the liquidus (1480 ◦ C) of TiSi2 . F-II area (dark grey islands) which covers a small part of the joint section is rich in Si and C and possibly SiC has been formed or has been diffused from SiC ML. The chemical composition of the F-II area is quite similar for all joints. The dispersion of the F-II phase in the filler region is quite inhomogeneous with a large scattering in the size of the Si–C rich islands. The joints fabricated at 7 and 25 MPa are characterized by a smaller size of the Si–C rich islands compared with the 2.7 MPa joint for which large islands are observed towards the SiC ML interface. These large islands indicate a possible diffusion of SiC from SiC ML as it has been pointed above. The part of the joint adjacent to the Cf /SiC is characterized by a finer dispersion of this phase.
3.3. Re-entry tests SiC ML-Cf /SiC joints fabricated with an applied pressure of 2.7 MPa were subject to re-entry tests up to 5 cycles under vacuum. The tests were performed using two different temperature
Please cite this article in press as: C. Jiménez, et al., Joining of ceramic matrix composites to high temperature ceramics for thermal protection systems, J Eur Ceram Soc (2015), http://dx.doi.org/10.1016/j.jeurceramsoc.2015.09.038
G Model
ARTICLE IN PRESS
JECS-10300; No. of Pages 7
C. Jiménez et al. / Journal of the European Ceramic Society xxx (2015) xxx–xxx 2000
400 200 0 00:00
02:00
o
Intermediate sample inspection
600
(b)
1400 1200 1000 800 600 400 200 0
04:00
06:00
Test Time (hh:mm)
08:00
Heater temperature CMC back temperature
1600
00:00
02:00
TC failure
800
Heater temperature CMC back temperature
Temperature ( C)
1000
1800
(a) Intermediate sample inspection
o
Temperature ( C)
1400 1200
5
04:00
06:00
08:00
10:00
Test Time (hh:mm)
Fig. 5. Temperature profiles used in the re-entry tests at 1391 ◦ C (a) and at 1794 ◦ C (b). (For interpretation of the references to color in text, the reader is referred to the web version of this article.)
Fig. 6. The CMC surface before (left) and after (right) five re-entry cycles at 1391 and 1794 ◦ C.
Fig. 7. The SiC ML surface before (left) and after (right) five re-entry cycles at 1391 and 1794 ◦ C.
Fig. 8. The surface of the SiC ML-Cf /SiC joint after the tests at 1391 ◦ C (a) and 1794 ◦ C (b).
Please cite this article in press as: C. Jiménez, et al., Joining of ceramic matrix composites to high temperature ceramics for thermal protection systems, J Eur Ceram Soc (2015), http://dx.doi.org/10.1016/j.jeurceramsoc.2015.09.038
G Model
ARTICLE IN PRESS
JECS-10300; No. of Pages 7
C. Jiménez et al. / Journal of the European Ceramic Society xxx (2015) xxx–xxx
6
Table 2 EDS analysis of the cross section of the SiC ML-Cf /SiC joints after the re-entry tests. The percentage of each phase is in mole%. C (at%)
Si (at%)
Ti (at%)
Main phases
Overall cross section
57.5 ± 1.4
18.9 ± 0.7
23.6 ± 1.2
F-I light grey
46.8 ± 0.5
7.0 ± 0.7
46.2 ± 0.7
F-II dark grey
47.7 ± 0.8
51.8 ± 0.9
0.5 ± 0.2
68.5% Ti3 SiC2 19.5% TiC 12.0% TiSi2 Ti3 SiC2 TiC Si–C
Overall cross section
52.8 ± 0.9
12.9 ± 1.0
34.3 ± 1.1
F-I light grey (close to SiC-ML interface) F-I light grey (close to CMC interface) F-II dark grey
42.2 ± 1.1
8.4 ± 1.3
46.4 ± 0.3
48.1 ± 0.2
1.6 ± 0.2
50.3 ± 0.4
TiC
45.4 ± 0.5
52.9 ± 0.9
1.6 ± 1.4
Si–C
SiC as fabricated o tested at 1391 C o tested 1794 C
SiC SiC SiC
30
SiC
40
SiC
SiC
SiC SiC C
20
50
60
SiC
SiC
SiC
SiC
profiles of the ARV, profile (a) with a peak temperature at 1391 ◦ C (Fig. 5a) and profile (b) with a peak temperature at 1794 ◦ C (Fig. 5b). During the test the backside temperature of the joint, i.e., the back side of the Cf /SiC is monitored using a K-type thermocouple. The red line in Fig. 5 depicts the thermocouple temperature attached at the back side of the Cf /SiC. The black line in Fig. 5 represents the temperature at the SiC top surface. This temperature is derived from the heat fluxes applied by the furnace heater during the tests and under the assumption that there are no considerable thermal losses. During the 2nd cycle of the test at 1794 ◦ C the temperature went beyond 1400 ◦ C, and the type K thermocouple failed and as a result the monitoring of the temperature at the back side of the CMC was interrupted. After the 1st cycle, the samples had been removed from the rig for optical inspection, and as no changes were found, 4 more cycles were carried out. The sample surface before and after the tests is presented in Figs. 6 and 7. The cracks appearing on the SiC ML surface pre-exist the re-entry tests. There is only a change in color that originates partially from an artefact of the scanner used for making the post test photographs and partially from graphite deposition at the surface as it will be explained below. The optical microscopy examination of the joints after the tests showed no detectable defects. Non-destructive testing using ultrasonic inspection was not possible as the cracks in the ceramic plate result in large “blackout” areas around the crack zones. The surface of the SiC ML that was subject to the thermal load was examined by SEM and XRD measurements. Fig. 8a and b presents the surface of the SiC ML after the test at 1391 and 1794 ◦ C, respectively. A high content of graphite is found on the surface of SiC which is not
82.1% Ti3 SiC2 17.4% TiC 0.5% TiSi2 Ti3 SiC2 TiC
C
1794
SiO2
1391
Intensity (arbitrary units)
Temperature (◦ C)
70
80
90
2theta (deg.) Fig. 9. XRD spectra of the SiC ML surface as fabricated (black line), tested at 1391 ◦ C (blue line) and at 1794 ◦ C (red line). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
related with the re-entry tests. The high graphite content is confirmed by XRD measurements on the SiC ML surface of the tested joints presented in Fig. 9 which show the presence of hexagonal graphite (SG: P63/mmc). A possible explanation for the presence of carbon is the thermal decomposition of SiC, where Si atoms diffuse
Fig. 10. SEM micrograph of the SiC ML-Cf /SiC cross section after the re-entry test at 1391 (a) and 1794 ◦ C (b).
Please cite this article in press as: C. Jiménez, et al., Joining of ceramic matrix composites to high temperature ceramics for thermal protection systems, J Eur Ceram Soc (2015), http://dx.doi.org/10.1016/j.jeurceramsoc.2015.09.038
G Model JECS-10300; No. of Pages 7
ARTICLE IN PRESS C. Jiménez et al. / Journal of the European Ceramic Society xxx (2015) xxx–xxx
outwards from the underlying SiC substrate and desorb subsequently, and the released carbon atoms diffuse and nucleate to form carbon rich areas. Annealing of SiC (up to 1400 ◦ C) under vacuum is a method currently used for the preparation of graphene-like layers [21]. The dark areas (S) in Fig. 8b are carbon and it comes from the graphite tiles that were used to place the structure to be joined in the hot press and remained on the surface of the SiC after the joining. The microstructure of the cross section of the joints after the thermal test at both temperatures shows the same general characteristics as in the as fabricated joints (Fig. 10). Table 2 presents the results from the EDS analysis of the cross sections of the joints after the re-entry tests. At the test at 1391 ◦ C the phase content of the overall cross section is similar to that of the as fabricated joint at 2.7 MPa (see Table 1). However, when the test temperature increases to 1794 ◦ C it has an effect analogous to that of pressure, i.e., the MAX phase content increases by around 20%, whereas the TiSi2 phase almost vanishes. Similarly, as before, this can be explained by the reaction of TiSi2 with TiC. This is to be expected because TiSi2 becomes liquid as the re-entry test temperature (1794 ◦ C) is above its liquidus (1480 ◦ C) and it has been reported [22] that Ti–Si melt reacts with TiC to produce Ti3 SiC2 . Furthermore, as it can be observed in Fig. 10, when the test temperature increases to 1794 ◦ C the Si–C rich islands are formed mainly in the middle part of the joint as in the joints fabricated at 25 MPa. This indicates that the test temperature has an effect similar to that of pressure, i.e., increased Si-C diffusion from the SiC ML. In the F-I (light grey) region of the joint tested at 1794 ◦ C there is a differentiation with respect to the Si and Ti content between the upper half part of the joint close to the SiC ML interface and the other half of the joint close to the CMC interface (Table 2). The upper part is characterized by a higher content of Si and lower content of Ti than those of the bottom part of the joint. The same effect, to a lesser degree, was observed in the as fabricated joint at 2.7 MPa. 4. Conclusions Successful joining of SiC multilayers to Cf /SiC ceramic composite materials using the ternary carbide Ti3 Si1.5 C2 and employing diffusion brazing bonding and pressures in the range 2.7–25 MPa was accomplished. As the fabrication pressure increases from 2.7 MPa to 7 MPa and above TiC reacts with TiSi2 to form the MAX phase Ti3 SiC2 . Cracks appear in the filler area and few of them can extend up to the free surface of the SiC multilayer. The cracks are attributed to the brittle behavior of the filler, the mismatch in the coefficient of thermal expansion between the joined parts and the filler and to the lack of flatness of the SiC specimens. These cracks are vertical to the SiC ML-Cf /SiC interface and are not detrimental to the shear strength of the joint. Thermal tests under two re-entry conditions, in vacuum, have not shown any detectable damage. Therefore these structures have the prospective to be utilized as re-usable thermal protection systems for re-entry applications. Acknowledgements This work has been performed within the framework of the European Project “SMARTEES (G.A. n. 262749) with the financial support by the European Community. The paper only reflects the view of the authors and the European Community is not liable for
7
any use of the information contained therein. References [1] I. Sakraker, C.O. Asma, Experimental investigation of passive/active oxidation behavior of SiC based ceramic thermal protection materials exposed to high enthalpy plasma, J. Eur. Ceram. Soc. 33 (2013) 351–359. [2] H. Kiyono, S. Shimada, K.J.D. MacKenzie, Kinetic and magic angle spinning-nuclear magnetic resonance studies of wet oxidation of beta-sialon powders, J. Electrochem. Soc. 148 (2001) B86–B91. [3] B.V. Cockeram, Flexural strength and shear strength of silicon carbide to silicon carbide joints fabricated by a molybdenum diffusion bonding technique, J. Am. Ceram. Soc. 88 (2005) 1892–1899. [4] Yutai Katoh, Lance L. Snead, Ting Chenga, Chunghao Shih, W. Daniel Lewis, Takaaki Koyanagi, Tatsuya Hinoki, Charles H. Henager Jr., Monica Ferraris, Radiation-tolerant joining technologies for silicon carbide ceramics and composites, J. Nucl. Mater. 448 (2014) 497–511. [5] Monica Ferraris, Milena Salvo, Valentina Casalegno, Andrea Ciampichetti, Federico Smeacetto, Massimo Zucchetti, Monica Ferraris, Joining of machined SiC/SiC composites for thermonuclear fusion reactors, J. Nucl. Mater. 375 (2008) 410–415. [6] M. Singh, R. Asthana, Joining of zirconium diboride-based ultra high-temperature ceramic composites using metallic glass interlayers, Mater. Sci. Eng. A 460–461 (2007) 153–162. [7] Mrityunjay Singh, Rajiv Asthana, Joining of ZrB2 -based ultra-high-temperature ceramic composites to Cu-clad-molybdenum for advanced aerospace applications, Int. J. Appl. Ceram. Technol. 6 (2009) 113–133. [8] B.H. Rabin, G.A. Moore, Reaction processing and properties of SiC-to-SiC joints, Mater. Res. Soc. Symp. Proc. 314 (1993) 197–203. [9] H. Dong, S. Li, Y. Teng, W. Ma, Joining of SiC ceramic-based materials with ternary carbide Ti3 SiC2 , Mater. Sci. Eng.: B 176 (1) (2011) 60–64. [10] Stefano Rizzoa, Salvatore Grassob, Milena Salvoa, Valentina Casalegnoa, Michael J. Reeceb, Monica Ferraris, Joining of C/SiC composites by spark plasma sintering technique, J. Nucl. Mater. 34 (2014) 903–913. [11] Alberto Ortona, Simone Pusterla, Sandro Gianella, An integrated assembly method of sandwich structured ceramic matrix composites, J. Eur. Ceram. Soc. 31 (2011) 1821–1826. [12] H. Böhrk, U. Beyermann, Secure tightening of a CMC fastener for the heat shield of re-entry vehicles, Compos. Struct. 92 (2010) 107–112. [13] Philippe Berthe, The advanced re-entry vehicle (ARV)—development step from ATV towards manned transportation systems, ESA/ESTEC, in: Proceedings of the 3rd International ARA days, 2–4 May, Arcachon, France, 2011. [14] M.W. Barsoum, T. El-Raghy, C.J. Rawn, W.D. Porter, H. Wang, E.A. Payzant, C.R. Hubbard, Thermal properties of Ti3 SiC2 , J. Phys. Chem. Solids 60 (1999) 429–439. [15] G. Motz, S. Schmidt, S. Beyer, The PIP-process: Precursor Properties and Applications, Ceramic Matrix Composites, in: W. Krenkel (Ed.), Wiley-VCH, Germany, 2008, pp. 165–186. [16] W.S. Yang, S. Biamino, E. Padovano, M. Pavese, S. Marchisio, G. D’Amico, S. Ceresa MioX, X. Chen, P. Fino, C. Badini, Thermophysical properties of SiC multilayer prepared by tape casting and pressureless sintering, Compos. Struct. 96 (2013) 469–475. [17] J. Barcena, M. Lagos, I. Agote, C. Jimenez, X. Hernandez, C. Badini, E. Padovano, S. Gianella, D. Gaia, V. Liedtke, K. Mergia, S. Messoloras, P. Yialouris, Y. Panayiotatos, A. Ortona, C. D’Angelo, C. Wilhelmi, SMARTEES FP7 space project—towards a new tps reusable concept for atmospheric reentry from low earth orbit, in: 7th European Workshop on Thermal Protection Systems and Hot Structures, 8–10 April 2013, ESA-ESTEC, Noordwijk, The Netherlands, 2013. [18] T. Salmon, X. Vo, M. Bottacini, TPS Architecture on ARV Proceedings of the 7th European Symposium on Aerothermodynamics, in: L. Ouwehand, Noordwijk, Netherlands, European Space Agency, 2011, id.120. [19] S.G. Vadchenko, A.E. Sytschev, D. Yu Kovalev, A.S. Shchukin, S.V. Konovalikhin, Self-propagating high-temperature synthesis in the Ti–Si–C system in features of product patterning, Nanotechnol. Russ. 10 (1) (2015) 67–74. [20] S.V. Konovalikhin, D. Yu Kovalev, A.E. Sytschev, S.G. Vadchenko, A.S. Shchukin, Formation of nanolaminate structures in the Ti–Si–C system: a crystallochemical study, Int. J. Self-Propag. High-Temp. Synth. 23 (2014) 216–220. [21] Qingyan Wang, Wenhao Zhang, Lili Wang, Ke He, Xucun Ma, Qikun Xue, Large-scale uniform bilayer graphene prepared by vacuum graphitization of 6H-SiC(0001) substrates, J. Phys.: Condens. Matter 25 (2013) 095002. [22] M. Abu, J. Mohamed, Z. Ahmad, Effect of excess silicon on the formation of Ti3 SiC2 using free Ti/Si/C powders synthesized via arc melting, ISRN Ceram. 2012 (2012) 10, Article ID 341285.
Please cite this article in press as: C. Jiménez, et al., Joining of ceramic matrix composites to high temperature ceramics for thermal protection systems, J Eur Ceram Soc (2015), http://dx.doi.org/10.1016/j.jeurceramsoc.2015.09.038