1 Oxidation Microstructure Studies of Reinforced Carbon/Carbon Nathan S. Jacobson* NASA Glenn Research Center Cleveland, OH 44135 Donald M. Curry NASA Johnson Space Center Houston, TX 77058 Keywords: Carbon/carbon composites, Oxidation, Coating, Reactivity, Graphitic carbon Abstract Laboratory oxidation studies of reinforced carbon/carbon (RCC) are discussed with particular emphasis on the resulting microstructures. This study involves laboratory furnace (500-1500oC) and arc-jet exposures (1538oC) on various forms of RCC. RCC without oxidation protection oxidized at 800 and 1100oC exhibits pointed and reduced diameter fibers, due to preferential attack along the fiber edges. RCC with a SiC conversion coating exhibits limited attack of the carbon substrate at 500, 700 and 1500oC. However samples oxidized at 900, 1100, and 1300oC show small oxidation cavities at the SiC/carbon interface below through-thickness cracks in the SiC coating. These cavities have rough edges with denuded fibers and can be easily distinguished from cavities created in processing. Arc-jet tests at 1538oC show limited oxidation attack when the SiC coating and glass sealants are intact. When the SiC/sealant protection system is damaged, attack is extensive and proceeds through matrix cracks, creating denuded fibers on the edges of the cracks. Even at 1538oC, where diffusion control dominates, attack is non-uniform with fiber edges oxidizing in preference to the bulk fiber and matrix. * Corresponding Author: [email protected] , Phone 216-433-5498, FAX 216-433-5544 2 1. Introduction The oxidation of carbon has been studied extensively. Perhaps the most important conclusion from these studies is that the mechanism and microstructure of oxidation is highly dependent on the type of carbon [1]. Most studies of carbon oxidation focus on either the reaction mechanism/kinetics or the microstructural characteristics of attack, although the oxidation mechanism and the resultant microstructure are closely connected. This report is on oxidation microstructures observed in the reinforced carbon/carbon (RCC) used for the thermal protection of the wing leading edge and nose cap of the Space Shuttle Orbiter. Consider first unprotected carbon/carbon. There is extensive literature on the oxidation of these materials [2-7]. Oxidation begins to be significant above about 500oC. Previous studies indicate oxidation is controlled by the chemical reaction of carbon and oxygen below about 700-800oC and both chemical reaction and gas-phase diffusion above about 700-800oC [2, 5]. Surface reaction rates are quite dependent on the type of carbon, so selective attack of the higher energy surfaces is expected. Attack is typically along the fiber axis at the fiber/matrix interface [3]. Crystalline or graphitic carbon is less susceptible to oxidation [6, 8]. Fibers derived from polyacrylonitrile (PAN) polymers [8] are more crystalline in the interior than the edges and hence the interior is more oxidation resistant than the edges. Oxidation creates a pointed morphology on such fibers [5, 6]. The rayon-derived fibers in RCC are expected to behave similarly. The conclusions from oxidation studies of carbon fiber/SiC matrix composites [8, 9] are also useful in understanding oxidation of RCC. Halbig and Cawley [9] show that reaction control at 750oC leads to more uniform attack deep into the substrate; whereas diffusion control at 1250oC leads to attack around the edges of the specimen. The structure of RCC has been described in detail elsewhere [10, 11]. RCC is protected from oxidation via a conversion coating of SiC. Due to the different coefficients of 3 thermal expansion between carbon and SiC, through-thickness cracks develop in the SiC coating. Several sealants are applied to fill these cracks. At the highest temperatures, the SiC conversion coating should be in compression and the cracks and fissures close. Ideally the sealants are fluid and fill the cracks and fissures as well. However, in non- ideal conditions and at lower temperatures, cracks and fissures may be open and provide pathways for oxidation. The authors have done a series of experiments with intentionally drilled holes through the SiC, leading to oxidation-created cavities below the SiC coating [12] at 600, 1000, and 1400oC. At 600oC attack was minimal; however at 1000 and 1400oC oxidation cavities formed at a rate described by a diffusion model. This paper reports on a series of controlled oxidation experiments on RCC with emphasis on the microstructure of oxidation damage in the carbon/carbon substrate. Oxidation damage in unprotected RCC, protected RCC with coating cracks, and protected RCC with coating damage is discussed. The objectives of this paper are two-fold (1) Relate the post-oxidation microstructure to mechanism (2) Identify distinguishing features of oxidation-induced porosity as compared to porosity from fabrication. 2.1. Experimental: Starting Materials A schematic of RCC is shown in Fig. 1. The substrate consists of a two-dimensional lay- up of rayon-derived carbon fabric. Fibers perpendicular to the direction of the particular cut are referred to as ‘longitudinal’ and the fibers parallel to the direction of the cut are referred to as ‘transverse’. Porosity in the carbon/carbon substrate is filled with repeated application of a liquid carbon precursor [10, 11]. As noted, oxidation protection is afforded by the formation of a SiC conversion coating. The cracks in the SiC are sealed with a vacuum infiltration of tetra-ethyl silicate (TEOS), which decomposes to SiO on 2 heating. Two applications of an outer coating of a sodium silicate-based glass (Type A Sealant) are applied for additional protection. 4 Figure 1. Schematic of RCC. 5 Three variants of RCC are studied here, as listed in Table 1. Each one taken after a different processing step in manufacturing. This selection of samples permits a study of the influence of each added component on oxidation. Table 1. RCC materials studied. Designation Material RCC-P3 C/C substrate with three liquid carbon precursor infiltrations RCC-P3/SiC C/C substrate with three liquid carbon precursor infiltrations, SiC conversion coating RCC-P3/SiC/TEOS/Sealant C/C substrate with three liquid carbon precursor infiltrations, SiC conversion coating, two sodium silicate glass sealant applications Figs. 2(a)-(c) illustrate the appearance of the as-fabricated material. Fig. 2(a) is an electron micrograph of the as-cut RCC P-3, showing both transverse and longitudinal fiber bundles. Fig. 2(b) is an optical micrograph of the RCC P-3/SiC material, illustrating typical voids and cracks through the fiber bundles produced in processing. Fig. 2(c) is a higher magnification photo illustrating a void from processing. Voids form from incomplete compaction and shrinkage of the matrix material. 6 Figure 2(a). RCC-P3 before oxidation, showing both transverse and longitudinal fibers. 7 Figure 2(b) Cross-sectional view of as fabricated RCC P-3/SiC. 8 Figure 2(c).Cross-sectional view of a pore from fabrication in RCC P-3/SiC + TEOS. 2.2. Experimental: Oxidation Exposures and Microstructural Examination Specimens were oxidized in either a reduced pressure laboratory tube furnace at the NASA Glenn Research Center or the arc-jet facility at the NASA Johnson Space Center (JSC). The reduced pressure laboratory tube furnace has been described elsewhere [12] and is shown schematically in Fig. 3. Using the vacuum pump and needle valves shown, it was possible to obtain excellent pressure control to 667 ± 13 Pa (5 ± 0.1 torr), which is representative of the pressure encountered in the Orbiter’s re-entry to the earth’s atmosphere. Gas flowing at about 100 cc/min STP was used. Typically controlled tests were done by heating in Ar (the MoSi heating elements reached temperature in ~30 2 minutes), switching to air for a specified hold time, and cooling in Ar. 9 Figure 3. Schematic of reduced pressure laboratory furnace. The arc-jet facility at JSC has been described in detail elsewhere [13, 14]. This large facility is for simulation of re-entry conditions. Air is electrically heated and expanded through a channel or conical nozzle. This creates a high flux of dissociated molecules on the standard 7.1 cm (2.8 inch) diameter RCC discs. Cross sections of all specimens were examined. The uncoated RCC-P3 specimens were characterized without additional preparation using a Field Emission Scanning Electron Microscopy (FE-SEM, Model S4700, Hitachi). The coated specimens were sectioned, mounted with high pressure epoxy infiltration technique, polished and examined with optical microscopy (Reichert MeF3A, Optronics). 10 3.1. Results and Discussion: Laboratory Oxidation of RCC-P3 Small cubes of RCC-P3 (~0.5 cm sides) were cut with a diamond saw and oxidized in the reduced pressure tube furnace as described in Table 2. Table 2. Oxidation of RCC-P3. Temperature (oC) Conditions Percent weight change 500 5 torr air/0.5 hr -2.2 800 5 torr air/0.5 hr -27.0 1100 5 torr air/0.5 hr -68.6 Figs. 4-6 show views of the carbon/carbon after each treatment. After the 500oC treatment, the sample showed limited oxidation, as indicated by the small weight change. The surface after oxidation (Fig. 4) is similar to the surface before treatment (Fig. 2(a)).