Isothermal oxidation tests were performed on the coated carbon-carbon composite sample at 1100, 1200, 1300 and 1400C. Generally the weight change verses time curves showed two types of behavior. One type is slow weight change, usually weight gain, over a long period of time. The other behavior is a dramatic weight loss when the coating fails. All coated samples showed enhanced life over uncoated CC composite. The oxidation kinetics exhibited by the samples do not conform to a parabolic rate law. Instead many processes are corrurring, causing the irregular oxidation curves seen in some of the figures. The results of these tests will now be discussed in terms of the individual coatings.
The HP24 sample loses weight rapidly from the start of the oxidation test at all temperatures except 1200C. The high weight loss means that the SiC conversion layer does not form a continuous protective SiO2 scale. Instead it is necessary to postulate that a defect in the coating bypasses the protection provided by SiO2. These holes/cracks provide a rapid diffusion path for oxidants that gasify the CC composite. The rate of gasification is not as great as for uncoated CC composite for two reasons. First the oxidants react with a smaller surface area, i.e., the entire sample is not exposed. Second the transport of the oxidant to the CC composite is limited by diffusion in the crack. The rate of diffusion is determined by the crack size.
Thermal stresses are the cause of the crack/hole defect. The rapid loss of weight at the beginning of testing supports this observation. The thermal stresses either open preexisting cracks or create new cracks. At 1100C the thermal stresses are greater than the 35 k.s.i.
Further complications arise in the B4C stringers that outline the converted fibers in the coating. These particles might be continuously linked from the substrate to the surface and could be preferentially oxidized into B2O3 and CO(g). This would lead to bubbling in the coating and on the surface of the composite.
Any boron oxide that forms in contact with the silica may alter the protective properties of the scale. The boron oxide can do this by changing the defect structure of the glass or crystalline silica. A change in the structure can result in rapid diffusion of oxygen through the scale. Boron oxide can also lower the viscosity of the silica scales.
The appearance of free Si extending from the surface to the CC composite substrate is another deleterious morphology and may be a greater problem than the B4C particle stringers. If the unprotective silica reached the substrate rapid weight losses would ensue from CO(g) production from high carbon activity phases. This type of coating failure was not observed in these tests, but could at longer exposures.
The fabrication of the HP24 coating is very important in determining the oxidation behavior that can be expected in service. The observations of coating failures and successes at these temperatures show that irregularities exist during fabrication, causing a scatter in oxidation behavior. The irregularities involve three microstructural variations: cracks, high carbon activity phases in the coating, and deleterious phases in the coating.