Small-scale high temperature corrosion cell during sample extraction operations
A series of materials selection trials carried out by TWI on behalf of an industrial member company has allowed the company to optimise its component fabrication procedures prior to, and during, the manufacturing process. The high temperature study, which has relevance to all industries where high temperature gases are present, identified materials with increased durability and resistance in a specific aggressive environment. As a further result of the work, the member company made several modifications to the design of its reactor interior to boost component lifetime.
TWI was contacted by an industrial member company regarding an issue with high temperature (above 1000°C) processing of oxide powders in a specialised flowing gas mixture. While the furnace design was robust, several components within the furnace were experiencing accelerated degradation and failure within a few days of exposure. Preliminary investigations determined that the material used to construct these components was not compatible with the extremely aggressive environment present. The company commissioned TWI to carry out materials performance trials, to select a material with a longer lifetime and better resistance to the environment.
Given that the composition of the gas changed depending on temperature, location, amount of powder and time, the project team decided to use simulated gas mixtures mimicking the most severe conditions observed within the furnace. Several different high temperature alloys were exposed to these environments for at least two weeks each. To more closely copy the microenvironment present at the component surface, some samples were also covered with the oxide powder, which partially reduced as the test progressed.
TWI constructed a new small-scale high temperature corrosion cell for the project. This cell is capable of safely testing specimens under flowing gas mixtures containing HCl, CO, CO2, H2, CH4, moisture and oxygen, with test durations of up to eight weeks and temperatures up to 1120°C. Up to five samples may be tested simultaneously with no cross-contamination.
Adherence of oxide to coupon surface and sub-surface carburisation of Haynes® Alloy 214
In consultation with the member company, the project team selected a test temperature of 1100°C and two specific reducing/carburising gas mixtures (10%CO-1%CO2-Argon or 90%CO-2.5%H2-Argon) to provide a range of test environments from mildly carburising to heavily carburising. Based on TWI’s extensive industrial experience and a review of the available literature, it then chose nine candidate nickel-based, iron-based and cobalt-based high temperature alloys with the required strength. Beta-aluminide diffusion coatings were applied to the exterior of a subset of these coupons to assess its efficacy as a protective coating. Different surface finishes (as-received, ground, pre-oxidised or polished) were also applied to assess the effect of surface pre-treatment.
Each coupon was submerged in oxide powder (or tested without the presence of oxide powder) and exposed to the flowing gas mixtures over several weeks at 1100°C. After extraction, the oxide powder was brushed away, leaving alloy coupons that showed vastly different performance. Some were visibly corroded, others apparently untouched by the environment. Yet, when sectioned and examined in detail by light microscopy and scanning electron microscopy, every single alloy coupon displayed some sign of change. When no oxide powder was present, the project team observed mass gain from carbon uptake (up to 5mg/cm2). When oxide powder was present, mass change measurements were not meaningful, as oxide powder adhered strongly to the surface in some areas of the sample surface but material loss was clearly visible in other areas.
The corrosion was a mixture of carburisation and oxidation, with several chromium-rich and aluminium-rich oxides forming at the surface. A discolouration and change in the surface texture of each coupon was visible to the naked eye, with the texture change indicating removal of some material from the surface. In some cases, the team observed general attack and voids which had formed below the surface. In others, the attack was primarily intergranular with associated carbide precipitation along the grain boundaries. There was no direct correlation between elemental composition and resistance to the environment, though generally chromium depletion along grain boundaries would be mitigated by the presence of more stable carbide formers such as niobium, as well as higher overall chromium contents. Additional molybdenum and lower iron contents may also be beneficial.
Haynes® Alloy 230 and Alloy 625 both showed good performance in these trials, with no change observed to the substrate beyond a depth of 65µm, and minimal material losses. Alloy 600 also performed reasonably well, though the attack was more general in this case. At the opposite end of the scale, Kanthal APM experienced massive attack in this environment, with pits up to 0.8mm deep, material loss and multiple sub-surface voids. Other alloys tested (601, 602CA, 693, Haynes® 214, Hastelloy® X) showed intermediate performance. In addition, the oxide powder bonded to the surface of the coupons during testing, leading to formation of mixed oxide phases and detachment of metallic particles when the oxide was removed.
The surface finish did not make a significant difference to the depth or mechanism of corrosion, even when the coupon was pre-oxidised. However, there was a significant difference in adherence of the oxide powder to the coupon depending on the finish, with the polished surface retaining the least oxide powder after testing.
The aluminide diffusion coating showed good protective characteristics in this environment, with less extensive sub-surface carburisation, and reduced oxidation. However, any flaws, cracks or imperfections in the coating led to catastrophic carburisation and a focus on a single region. On an uncoated specimen of Alloy 601, carburisation was observed up to 110µm below the substrate surface. In a region of coated Alloy 601 with a small imperfection in the coating, attack at the surface was extensive and carburisation reached depths of 2600µm, radiating outwards from the coating flaw. This is shown below.
Extensive and heavy carburisation of Alloy 601 due to a small flaw in the diffusion coating
Based on its materials assessment, TWI recommended that Haynes® Alloy 230 and Alloy 625 be studied further for fabrication of components in this challenging environment. It also recommended polishing or machining components to reduce adherence of the oxide powder to the surface. This would lead to an increased lifetime as less material could be removed by adhesion-removal or by erosion/wear.
The applied beta-aluminide coating was also generally protective against the environment but the coating quality required further refinement, as any local flaws could lead to catastrophic and rapid failure. TWI did not recommend its use at the current time, but recommended further refinement of the coating application process.
For more information about our services for fabrication in high temperature environments, please see the Materials and Corrosion Management section of our website, or contact us.