CORROSION-RESISTANT ALLOY AND APPLICATIONS

Abstract

The present disclosure is directed to new ferritic alloys comprising iron, chromium, and silicon. The novel alloys have surprisingly enhanced corrosion resistant properties, with excellent workability, strength, and ductility. The disclosed alloys, and products made therefrom are useful in making low cost, efficient, and clean biomass stove combustors, as well as other devices for use with biomass fuel.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims benefit of priority pursuant to U.S. Provisional Patent Application No. 62/308,551 entitled “Corrosion-Resistant Alloy and Applications,” filed on Mar. 15, 2017, which is hereby incorporated by reference in its entirety.

GOVERNMENT LICENSE RIGHTS

This invention was made with Government support under grant number DE-AC05-00OR22725 awarded by the Department of Energy. The Government has certain rights in the invention.

FIELD

The disclosed compositions, methods, and systems are directed to novel corrosion-resistant metal alloys, making and testing such alloys, and use of such alloys in biomass cookstoves and biomass pyrolysis and combustion environments for energy production.

BACKGROUND

Nearly 40% of the world cooks on open fires or inefficient biomass-fueled cookstoves. The resulting smoke is a health hazard, contributing to an estimated 4 million premature deaths per year, as well as a major source of carbon black emissions. One solution is the introduction of improved, clean-burning biomass cookstoves. One of the most challenging components to achieve clean-burning is the combustor, which must operate at high temperatures (often ≥600° C.) in the presence of highly corrosive species released from biomass fuel combustion, yet be sufficiently low cost to permit widespread adoption.

Approximately 3 billion people worldwide cook their food by burning solid biomass fuel over open fires or basic, inefficient biomass-burning cookstoves, releasing toxic emissions that cause an estimated ˜˜4 million premature deaths a year. Such emissions also account for approximately 20% of the world's black carbon, with significant regional and global climate change implications Inefficient cooking methods also result in the frequent need for time-consuming gathering of biomass fuel, particularly by women and children, which takes time away from activities such as education or income generation, and in some circumstances represents a safety risk. Low-cost, highly efficient improved cookstoves with increased efficiencies/reduced emissions are needed as one approach to help solve this complex problem.

Significant recent progress has been made for improved cookstove design, combustion modeling, and emissions reductions. The success of clean biomass cookstoves from an engineering standpoint is, however; also dependent on the materials of construction. Among the more challenging components is the combustor, which operates at high temperatures (often ≥600° C.) in the presence of highly corrosive species (water vapor, chlorine, sulfur, salts, ash, and deposits) derived from the combustion of biomass fuel. High temperature, corrosive conditions pose a significant challenge in the developing world, since materials must be low-cost, durable, and, in most cases light weight (low mass), in order to permit widespread adoption in the developing world. Metallic alloys offer the best combination of ease of manufacturing, lightweight, robustness/mechanical integrity, and design flexibility, but are also far more susceptible to corrosion than other materials, for example ceramic combustion chambers. In particular, mixed oxidant and ash/salt deposit-induced, biomass-related high-temperature corrosion can result in corrosion rates (by about an order of magnitude or higher; depending on alloy composition, temperature, and environment), as compared to air oxidation.

State-of-the-art metal combustors for clean biomass cookstoves are type 310 stainless steel (Fe-25Cr-20Ni wt. % base) and the FeCrAl family of alloys, which are generally in the range of Fe-(10-20)Cr-(2-8)Al wt. % base. However, the relatively high cost of 310 stainless steel and the FeCrAl alloys is near the commercialization viability borderline for clean biomass cookstoves in many developing areas of the world. These metals generally form oxide base layers to aid in corrosion protection. For example, Type 310 and related stainless steels form chromium-oxide base layers to achieve corrosion protection at high temperatures, whereas FeCrAl alloys form aluminum-oxide base layers for protection. As disclosed herein, a third option for protecting metal alloys from high temperature corrosion is the addition of silicon to the formulation. The silicon helps to form silicon-oxide base layers and/or mixed iron-silicon oxides base layers to protect the alloy from corrosion.

Corrosion of alloys under biomass combustion conditions has been an active area of research in recent years for industrial scale power generation applications (e.g. power plants, boilers, etc.) that use biomass as a renewable alternative to fossil fuels. Laboratory corrosion test exposure standards and protocols are available, particularly with regards to introduction of ash/salt deposits and their interactions with flue gas (combustion exhaust products containing SOx, Cl, etc. species). Although there is some qualitative overlap with biomass cookstoves (e.g. with regards to the corrosive species encountered, particularly salts, sulfur, and water vapor), the temperatures, pressures, combustion environments, biomass fuel sources, corrosive species concentrations and deposit tendencies, alloy mechanical property requirements, and component lifetimes encountered in these power generation applications can differ considerably from those of low-cost (˜$10-50 range), individual cookstoves that burn local biomass fuels. Moreover, the typical biomass cookstove has a targeted lifetime on the order of about 3000-5000 hot hours.

Existing metals either lack suitable corrosion resistance, or are high cost specialty formulations. What is needed is a low-cost and cost-effective alternative to existing alloy metals for use in corrosive environments. In addition, there is little or no research specifically devoted to corrosion issues in improved biomass cookstoves.

Disclosed herein is the development of accelerated corrosion test screening protocols employing highly corrosive salt and water vapor species, specifically designed to evaluate alloys for clean biomass cookstove combustors. Also disclosed are corrosion analysis for a range of commercial alloys and newly developed corrosion resistant alloys for use in cookstove. A new Fe (iron)-Cr (chromium)-Si (silicon) base alloy is disclosed that provides for surprisingly improved corrosion resistance at lower cost than state-of the art FeCrAl-based alloys and stainless steel alloys. The disclosed methods and alloys are useful in the production and use of low-cost structural alloys, for example with biomass cookstove metal combustors. The disclosed methods and compositions may also find use in other biomass combustion environments such as pyrolysis equipment for bio-oil production, and biomass boilers and gasification systems for energy production.

SUMMARY

Disclosed herein are metallic alloys with enhanced corrosion resistance. In many embodiments, the disclosed alloys may be ferritic-based matrix structural alloys. In most embodiments, the composition of the disclosed alloys are in a range of about 13-17 wt % chromium (Cr), about 2-3.5 wt % silicon (Si), about 0.2-1 wt % manganese (Mn), about 0.3-0.7 wt % titanium (Ti), about 0.1-0.6 wt % carbon (C), with the remainder iron (Fe). The alloy may be substantially free of nickel, or may include from about 0.1 to 1.0% nickel. In one embodiment the disclosed alloy is about 2.4-3.0 wt % Si, about 14.5-15.5 wt % Cr, and about 0.2-0.6 wt % C. In another embodiment, the disclosed alloy is about 2.4-3.0 wt % Si, about 14.5-15.5 wt % Cr, and about 0.30-0.6 wt % C. In yet another embodiment, the disclosed alloy is about 2.4-3.0 wt % Si, about 14.5-15.5 wt % Cr, and about and 0.4-0.6 wt % C.

Also disclosed is a ferritic-alloy, further comprising boron (B) or nickel (Ni). In many embodiments, the alloy comprises boron in the range of about 0.01 to 0.5 wt % and/or nickel in the range of about 0.1 to 1.0 wt %. In many embodiments, boron may help to optimize the behavior of the Si-rich and/or Fe—Si rich oxide for corrosion resistance. In one embodiment, the alloy comprises boron from about 0.05 to 0.15 wt % B. In many embodiments, the addition of 0.1 to 1 wt. % Ni may help further improve alloy ductility.

The disclosed alloys may be useful in devices that may be exposed to corrosive environments. In one embodiment, the alloy is used in constructing the combustor for a biomass fuel. In one embodiment, the combustor is in a biomass-fueled cookstove. In another embodiment, the combustor is in a biomass-fueled energy or heat generator.

Also disclosed are methods of working the disclosed alloys, for example to create sheets and foils of the alloy. In many embodiments, the disclosed alloy may be hot rolled at a temperature above about 700° C. and below about 1250° C. In some embodiments, the hot rolling may be at about 1100° C. In many embodiments, the disclosed alloys may be cold rolled, for example at room temperature. In many embodiments, the disclosed alloy is first hot rolled and then cold rolled. In some embodiments, cold rolling may be preceded or followed by annealing at a temperature from about 600° C. to about 1250° C., for example about 850° C., 950° C., or 1000° C. In many embodiments, an annealing step may help to increase the ductility of the disclosed alloy and may further enhance its corrosion resistance.

The disclosed alloys may resist corrosion when exposed to a corrosive environment. In many embodiments, the disclosed alloys may resist corrosion when biomass fuel having greater than about 800 μg of halogen per gram of fuel is burned. In most embodiments, the disclosed alloys show less than about 10% mass change over about 500 hours in a corrosive environment. In many embodiments, the disclosed alloys may lose less than about 10% of the thickness of the alloy when exposed to a corrosive environment for about 500 hours.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 shows computational equilibrium thermodynamic predictions for Fe-15Cr-0.5Ti as a function of Si and C content (FIG. 1 a, Fe-15Cr-5Si-0.5Ti-0.08C wt. %; FIG. 1 b, Fe-15Cr-3Si-0.5Ti-0.08C wt. %; and FIG. 1c, Fe-15Cr-3Si-0.5Ti-0.5C wt. %) using JMatPro_v6 and the Stainless Steel (Fe) database (Sente Software Ltd. Surrey Technology Centre 40 Occam Road GU2 7YG United Kingdom).

FIG. 2 Top, left, photograph of lab furnace testing alumina tray with samples, right, in-situ cookstove testing. Bottom, temperature profiles for the well-controlled portion of the burn cycle with 3 wood sticks for the two cookstove test beds.

FIG. 3 shows graph of corrosion data for the commercial FeCrAlY alloy from in-situ cookstove testing using as-received clear Pine wood and salted clear Pine wood.

FIGS. 4a-d Specific mass change (4a, 4c) and metal loss (4b, 4d) corrosion data from 600° C. lab furnace testing of commercial (4a, 4b) and developmental alloys (4c, 4d). The metal loss data is plotted as the average ±1 standard deviation of the 3 locations of greatest attack. The environment was air +10% H₂O, with salt added at the start of testing and re-applied after every 100 h cycle.

FIGS. 5a-d shows plots of specific mass change (5a, 5c) and metal loss (5b, 5d) corrosion data from 800° C. lab furnace testing of commercial (5a, 5b) and developmental alloys (5c, 5d). The metal loss data is plotted as the average ±1 standard deviation of the 3 locations of greatest attack. The environment was air +10% H₂O, with salt added at the start of testing and re-applied after every 100 h cycle.

FIGS. 6a-d shows graphs of specific mass change (6a, 6c) and metal loss (6b, 6d) corrosion data from in-situ cookstove testing of commercial (6a, 6b) and developmental alloys (6c, 6d). The metal loss data is plotted as the average ±1 standard deviation of the 3 locations of greatest attack. FeCrAlY A and FeCrAlY B are duplicate samples run in each of the two cookstove test beds utilized. Salted clear Pine wood was burned to induce accelerated corrosion conditions.

FIGS. 7a-c depicts cross-section backscattered electron mode SEM images of 310S stainless steel after 500 h exposure in (7a) 600° C. lab furnace testing, (7b) 800° C. lab furnace testing, and (7c) in-situ cookstove testing. A cross-section of a 310S combustor from a field-operated cookstove is shown in (7d) for comparative purposes.

FIGS. 8a-c depicts cross-section backscattered electron mode EPMA images and corresponding elemental maps for (8a) a 310S field operated cookstove combustor, (8b) 310S after 1000 h of in-situ cookstove testing, and (8c) 310S after 500 h of 800° C. lab furnace testing.

FIGS. 9a-d shows cross-section backscattered electron mode SEM images for pure Ni after (9a) 500 h, 600° C. lab furnace testing, and (9b) after 500 h, 800° C. lab furnace testing. A cross-section backscattered electron mode EPMA image and elemental maps are shown in (9c, 9d) for pure Ni after 500 h of in-situ cookstove testing.

FIGS. 10a-f shows cross-section backscattered electron mode SEM images after 1000 h of in-situ cookstove testing for (10a) 316L, (10b) 446, (10c) FeCrAlY, (10d) 25CrFeCrAl, (10e) FeCrAlSi, and (10f) FeCrSi.

FIGS. 11a-c shows cross-section light microscopy images after 1000 h of lab furnace testing at 800° C. for (11a) 310S, (11b) FeCrAIY, and (11c) FeCrSi.

FIGS. 12a-c shows cross-section backscattered electron mode EPMA images and elemental maps for (12a) FeCrAlY after 1000 h of in-situ cookstove testing, (12b) FeCrSi after 1000 h of in-situ cookstove testing, and (12c) FeCrSi after 500 h of 800° C. lab furnace testing.

FIG. 13 is a graph showing predicted phase equilibrium, re-calculated for alloy having Fe-15Cr-2.4Si-0.56C-0.5Mn-0.5Ti (generated by JMatPro v.9).

FIG. 14 shows micrographs showing microstructure of disclosed Fe-15Cr-2.4Si-0.5Mn-0.5Ti-0.57C wt. % alloy with and without post rolling annealing.

FIG. 15 is a graph of Vickers Hardness vs. temperature for the disclosed Fe-15Cr-2.4Si-0.5Mn-0.5Ti-0.57C wt. % alloy.

FIG. 16 shows micrographs of the edge surfaces of two annealed samples of the Fe-15Cr-2.4Si-0.5Mn-0.5Ti-0.57C wt. % alloy after 90% cold rolling.

FIG. 17 show ductility studies on various disclosed alloys.

DETAILED DESCRIPTION

Disclosed herein are alloys useful in a broad range applications where corrosion resistance is useful, and methods of making, using, and working the disclosed alloys. The disclosed alloys may be useful in applications where corrosion may be present, for example applications related to energy conversion and combustion, in addition to applications involving alloy, chemical, and petrochemical process environments. In some embodiments, the environment may include aggressive oxidizing species such as O, S, C, H₂O, salts, heavy metals, etc. and temperatures may be in excess of about 400-500° C.

Resistance to high-temperature (greater than about 600° C.) corrosion is achieved by the formation of continuous, protective, slow-growing oxide surface layers (scales), which are typically based on Al₂O₃, Cr₂O₃, and/or SiO₂, and (occasionally) NiO. Disclosed herein is a series of studies involving novel ferritic alloys comprising high concentrations of carbon and moderate concentrations of silicon. The disclosed alloys possess superior corrosion resistance, without suffering from typical drawbacks seen with higher and lower concentrations of silicon. In addition, the disclosed alloys are more readily workable, for example by cold rolling, than other ferritic alloys. The disclosed alloys were compared against existing formulations for resistance to corrosion in environments with, for example, high halogen content.

The disclosed novel alloys have surprisingly good manufacturability. Typically, high levels of Si result in poor manufacturability, poor mechanical properties, and weldability issues. These problems are due, in part, to promoting brittle alpha prime and sigma Cr-rich phases. Under some high-temperature oxidation conditions, high levels of Si increase the tendency for oxide scale spallation due to the large coefficient of thermal expansion mismatch between SiO₂ and Fe—Cr base alloys. For this reason, commercial ferritic stainless steels limit the amount of silicon to ˜≤0.5 to 1 wt. %. Austenitic stainless steels can generally tolerate higher levels of Si due to their high Ni content, which helps resist alpha prime and sigma formation, typically up to ˜2 wt. % Si range, although some grades as high as 3 to 3.5 wt. % Si are available. Reports in salt-containing, high-temperature corrosion conditions also indicate a benefit of Si at 3 wt. % and higher in austenitic alloys, but not at 1.6 wt. %. Such high Si austenitics are certainly of interest for improved biomass cookstove combustors, although their relatively high cost due to their Ni content, and increased manufacturability challenges from the high Si content, are drawbacks for improved biomass cookstoves.

The disclosed ferritic FeCrSi alloys were designed in consideration of the detrimental effects of Si on manufacturability, mechanical properties, and weldability. Computational thermodynamic calculations were used to identify specific alloy compositions. These calculations indicated that, compared to a ferritic Fe-15Cr-5Si base alloy, reduced Si combined with increased C levels may suppress the predicted formation of brittle sigma and alpha prime phases (FIG. 1). The disclosed FeCrSi alloy had an intended nominal composition of Fe-15Cr-3Si-0.5Mn-0.5Ti-0.06C, but was found to be Fe-15Cr-2.4Si-0.5Mn-0.5Ti-0.57C on analysis (Table 1). Titanium was added to help limit formation of M23C6 (M=Cr) type precipitates, which can tie up Cr in a manner detrimental to corrosion resistance (sensitization). However, the high C levels in the FeCrSi alloy greatly exceeded the levels that can be mitigated by the Ti level added, as indicated in the computational thermodynamics which predicted significant M23C6 formation for this alloy (FIG. 1).

Surprisingly, and in contrast with expected characteristics of such compositions, Applicants discovered that the disclosed FeCrSi alloy is readily manufacturable by both hot and cold rolling despite the high Si and C content. Without wishing to limited by theory, the amenability to hot rolling was likely aided by austenite stabilization in the 1100° C. hot-rolling condition employed (FIG. 1), resulting from the high levels of C additions. Further, the FeCrSi alloy also exhibited excellent high-temperature corrosion resistance under biomass-cookstove relevant conditions. This enhanced corrosion resistance was achieved despite the high levels of C and associated M23C6 formation, and at half the Si level of the literature reported Fe-15Cr-5Si alloy. Additions of 1 wt. % Si to as-cast Fe-14Cr-0.4C have been reported to permit ready dissolution of Cr-carbides to support chromia scale formation under high-temperature oxidation conditions, and a similar effect likely occurred in the FeCrSi alloy of the present work.

These findings indicate that FeCrSi alloy compositions in the range of about 13-17 wt % Cr, about 2-3.5 wt % Si, about 0.2-1 wt % Mn, about 0.3-0.7 wt % Ti, about 0.1-0.6 wt % C, and the remainder Fe show potential for use in biomass cookstove combustor components and are of strong interest for further study and scale-up development. Higher C levels may help to improve high-temperature strength, which, with ferritic alloys, is at the borderline for use in improved biomass cookstoves (depending on combustor/cookstove design and operation temperature). Future may evaluate both weldability (riveting is also an option in cookstove applications) and hot/cold rolling manufacturability, as well as longer-term laboratory corrosion and field testing. This Fe—Cr—Si—Mn—Ti—C base alloy composition range may also find use in biomass related power generation applications as alloys and/or coatings. However, it is important to note that the cookstove combustor application has far less stringent requirements for corrosion resistance, mechanical properties, and weldability than do alloys for biomass-fired power generation applications involving industrial scale components with operational lifetimes of hundreds of thousands of hours.

Annealing Temperature

Workability may be enhanced by processing the disclosed alloys at an annealing temperature. In many embodiments, the annealing temperature may be after a hot rolling and/or after a cold rolling. In most embodiments, the annealing temperature is between about 600° C. and 1200° C., for example greater than about 600° C., 650° C., 700° C., 750° C., 800° C., 850° C., 900° C., 950° C., 1000° C., 1050° C., 1100° C., or 1150° C., and less than about 1200° C., 1150° C., 1100° C., 1050° C., 1000° C., 950° C., 900° C., 850° C., 800° C., 750° C., 700° C., or 650° C. In one embodiment, the annealing temperature is selected from about 850° C., 900° C., 950° C., and 1000° C.

The disclosed alloys may be subjected to an annealing temperature for various lengths of time. In some embodiments the annealing time may be between about 1 min and 120 min. In many embodiments, the annealing time is greater than about 1 min, 5 min, 10 min, 20 min, 30 min, 60 min, 70 min, 80 min, 90 min, 100 min, or 110 min, and less than about 120 min, 110 min, 100 min, 90 min, 80 min, 70 min, 60 min, 50 min, 40 min, 30 min, 20 min, 10 min, 5 min, 4 min, 3 min, 2 min, or 1 min. The preferred annealing time may be sufficient to allow the average Vickers Hardness of the disclosed alloy to be less than about 500 HV, 450 HV, 400 HV, 350 HV, 300 HV, 250 HV, or 200 HV, and greater than about 150 HV, 200 HV, 250 HV, 300 HV, 350 HV, 400 HV, or 450 HV.

Corrosive Environment

The disclosed alloys may be resistant to a corrosive environment. In many embodiments the disclosed corrosive environment may be an environment wherein the average temperature is greater than about 400° C., 450° C., 500° C., 550° C., 600° C., 650° C., 700° C., 750° C., 800° C., 850° C., or 900° C. and less than about 950° C., 900° C., 850° C., 800° C., 750° C., 700° C., 650° C., 600° C., 550° C., 500° C., or 450° C., for greater than 50 h, 100 h, 200h, 300 h, 400 h, 500 h, 600 h, 700 h, 800 h, 900 h, 1000 h, 2000 h, or 3000 h, and less than about 3500 h, 3000 h, 2500 h, 2000 h, 1500 h, 1000 h, 900 h, 800 h, 700 h, 600 h, 500 h, 400 h, 300 h, 200 h, or 100 h. In many embodiments, the corrosive environment is one wherein the combusted fuel contains one or more halogen or halogen salt, such as sodium chloride, NaCl, chloride, Cl, fluoride, Fl, or iodide, I. In many embodiments, the average halogen content of one gram of fuel in a corrosive environment is greater than about 800 μg, 850 μg, 900 μg, 950 μg, 1000 μg, 1050 μg, 1100 μg, or 1200 μg, and less than about 1250 μg, 1200 μg, 1100 μg, 1000 μg, 900 μg, or 850 μg.

Corrosion Resistance

The disclosed alloys may resist losing mass or thickness in the corrosive environment. In many embodiments, the disclosed alloys may lose less than about 50% of their mass per cm², over greater than 50 h, 100 h, 200 h, 300 h, 400 h, 500 h, 600 h, 700 h, 800 h, 900 h, 1000 h, 2000 h, or 3000 h, and less than about 3500 h, 3000 h, 2500 h, 2000 h, 1500 h, 1000 h, 900 h, 800 h, 700 h, 600 h, 500 h, 400 h, 300 h, 200 h, or 100 h. In many embodiments, the average mass loss or average loss of thickness is less than about 50%, 45%, 40%, 35%, 30%, 25%, 20%, 15%, 10%, or 5% and more than about 2%, 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, or 45%.

Processing the Disclosed Alloy

The disclosed alloys may be processed in a variety of ways well known in the art. As described above, the disclosed alloys may be hot or cold rolled to produce a product with a reduced thickness. In some embodiments, the thickness of the disclosed alloys may be reduced by more than 50% by rolling, for example by more than 55%, 60%, 65%, 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99%, and less than about 99.5%, 99%, 98%, 97%, 96%, 95%, 94%, 93%, 92%, 91%, 90%, 85%, 80%, 75%, 70%, 65%, or 60%.

The disclosed alloys may be processed to various thicknesses, for example less than about 3 mm. In many embodiments, the thickness of the product produced from the alloy is greater than about 0.1 mm and less than about 5 mm, for example greater than about 0.10 mm, 0.15 mm, 0.20 mm, 0.25 mm, 0.30 mm, 0.35 mm, 0.40 mm, 0.45 mm, 50 mm, 0.60 mm, 0.70 mm, 0.80 mm, 0.90 mm, 1.0 mm, 1.5 mm, 2.0 mm, or 2.5 mm and less than about 3.0 mm, 2.0 mm, 1.5 mm, 1.0 mm, 0.9 mm, 0.8 mm, 0.7 mm, 0.6 mm, 0.5 mm, 0.4 mm, 0.3 mm, 0.2 mm, or 1.5 mm. In a preferred embodiment, the alloy is processed into a sheet or foil with an average thickness of about 0.3 to 0.6 mm.

Biomass may refer to any organic fuel that may be burned, for example, without limitation, coal, charcoal, wood, dung, leaves, sticks, oil.

EXAMPLES

I. Alloy Compositions and Corrosion

Alloys for study are shown in Table 1 (compositions are given in weight percent, wt. %). The alloys were selected to provide information on a range of protective oxide-scale forming types and a range of base alloy compositions that may be useful in cookstove combustors. The tested alloys include ferritic alloys, based on body-centered-cubic (BCC) Fe, and austenitic alloys, based on face-centered cubic (FCC) Fe, which offer better high-temperature strength than ferritics but are also more costly due to their high Ni contents (Ni stabilizes FCC Fe). Commercial alloys were also evaluated. These alloys included ferritic, alumina-forming FeCrAlY (Fe-20Cr-5Al base) and austenitic, chromia-forming type 310S stainless steel (Fe-25Cr-20Ni base), which are considered state-of-the-art cookstove combustor alloys for their balance of relatively low cost and good high-temperature corrosion resistance. Lower-cost commercial austenitic type 316L (Fe-17Cr-10Ni base), austenitic type 201 (Fe-18Cr-7Mn-5Ni base) and ferritic type 446 (Fe-25Cr base) stainless steels were also evaluated. Materials such as mild or galvanized steels were not evaluated as they are not suitable for the 600° C. temperatures encountered in the combustion chamber of improved biomass cookstove designs, and rapidly corrode at these temperatures.

TABLE 1
Table 1. Analyzed compositions in wt. % of developmental and commercial alloys by
inductively coupled plasma and combustion techniques.
(Impurities ≤ ~0.05 wt. % not reported).
	Fe	Ni	Cr	Al	Si	Mo	Mn	Ti	C	other
FeCrAlSi	75.99		14.56	5	2.81		0.47	0.49	0.066
FeCrSi	80.85		15.15		2.44		0.47	0.47	0.57
Fe25CrAl	68.99		25.1	4.84	0.01			0.47	0.07
FeCrAlY	73.04	0.12	20.42	5.65	0.24		0.18		0.016	0.05Y,
										0.04Hf
										0.05Zr
310S	53.23	18.83	25.17		0.59	0.43	0.91		0.03	0.3Cu
										0.17Co
										0.09V
										0.06Nb
446	74.1	0.17	24.09		0.26		0.75		0.058	0.14V
										0.11Nb
316L	67.86	9.98	17.08		0.43	2.1	1.56		0.026	0.46Cu
										0.07V
201	70.73	4.42	16.21		0.5	0.26	6.79		0.085	0.71Cu
										0.07Co
										0.068N
AFA-25Ni	51.83	25.04	13.84	3.56	0.13	0.18	1.99		0.114	2.51Nb
										0.51Cu
										0.16W
AFA-20Ni	57.61	19.96	13.87	3.06	0.12	1.99	1.99		0.152	0.6Nb
										0.52Cu
AFA-12Ni	62.39	12.02	13.91	2.52	0.11	0.11	4.97		0.201	3.06Cu
										0.59Nb
Pure Ni	0.09	99.44			0.04		0.23		0.068	0.06Co

Developmental model alloys based on ferritic Fe-15Cr with 3Si, Fe-15Cr with 5Al and 3Si, and Fe-25Cr-5Al were studied to further assess the relative effects of Al, Cr, and Si additions. Three developmental alumina-forming austenitic (AFA) alloys were also studied, employing levels of Ni additions at 25Ni, 20Ni, and 12Ni wt. % (Ni levels are a key cost driver for stainless steel alloys). Unalloyed Ni (referred to as “pure” Ni for simplicity) was studied as a surrogate for a Ni-cladding as well as for insights into the corrosion mechanism. Pre-oxidized FeCrAlY (treated at 1100° C. for 2 h in air, which yielded specific mass gains of 0.5-0.6 mg/cm²) with a ˜1-3 μm alumina surface layer pre-formed prior to testing was also studied, both for mechanistic insights and as a potential route to improved corrosion resistance.

Manufacturing of the disclosed novel alloys was also considered. For example, high levels of Si can result in poor alloy manufacturability, poor mechanical properties, and weldability issues, due in part to promoting brittle alpha prime and sigma Cr-rich phases. Under some high-temperature oxidation conditions, high levels of Si can also increase the tendency for oxide scale spallation due to the large coefficient of thermal expansion mismatch between SiO₂ and Fe—Cr base alloys. Because of this, commercial ferritic stainless steels typically are limited to ˜≤0.5 to 1 wt. % Si. Austenitic stainless steels can generally tolerate higher levels of Si due to their high Ni content which helps resist alpha prime and sigma formation, typically up to ˜2 wt. % Si range, although some grades as high as 3 to 3.5 wt. % Si are available. Reports in salt-containing, high-temperature corrosion conditions also indicate a benefit of Si at 3 wt. % and higher in austenitic alloys, but not at 1.6 wt. %. Such high Si austenitics are certainly of interest for improved biomass cookstove combustors, although their relatively high cost due to their Ni content, and increased manufacturability challenges from the high Si content, are drawbacks for improved biomass cookstoves.

The disclosed ferritic FeCrSi alloys were designed in consideration of the detrimental effects of Si on manufacturability, mechanical properties, and weldability. Computational thermodynamic calculations indicate that, compared to a ferritic Fe-15Cr-5Si base alloy, reduced Si and increased C levels can suppress the predicted formation of brittle sigma and alpha prime phases (FIG. 1, bottom). The disclosed FeCrSi alloy had an intended nominal composition of Fe-15Cr-3Si-0.5Mn-0.5Ti-0.06C, but was found to be Fe-15Cr-2.4Si-0.5Mn-0.5Ti-0.57C on analysis (Table 1). Titanium was added to help limit formation of M23C6 (M=Cr) type precipitates, which can tie up Cr in a manner detrimental to corrosion resistance (sensitization). However, the high C levels in the FeCrSi alloy greatly exceeded the levels that can be mitigated by the Ti level added, as indicated in the computational thermodynamics which predicted significant M23C6 formation for this alloy (FIG. 1).

Surprisingly, the disclosed FeCrSi alloy was readily manufacturable by both hot and cold rolling despite the high Si and C content. The amenability to hot rolling was likely aided by austenite stabilization in the 1100° C. hot-rolling condition employed (FIG. 1), resulting from the high levels of C additions. Further, the FeCrSi alloy also exhibited excellent high-temperature corrosion resistance under biomass-cookstove relevant conditions. This enhanced corrosion resistance was achieved despite the high levels of C and associated M23C6 formation, and at half the Si level of the literature reported Fe-15Cr-5Si alloy. Additions of 1 wt. % Si to as-cast Fe-14Cr-0.4C have been reported to permit ready dissolution of Cr-carbides to support chromia scale formation under high-temperature oxidation conditions, and a similar effect likely occurred in the FeCrSi alloy of the present work.

Alloy Corrosion Test Sample Preparation

Commercial 316 L, 310S, and FeCrAlY alloys were procured in sheet form, 446 in round bar form, and the pure Ni and 201 in plate form (see Table 1). Three developmental AFA alloys (see Table 1) were obtained in solutionized plate form from hot-rolled and machined ˜15 kg vacuum induction melted ingots. FeCrAlSi, FeCrSi, and Fe25CrAl alloys (see Table 1) were vacuum arc cast in small, laboratory scale 2.5 cm×2.5 cm×10 cm ingots, solutionized at 1200° C. for 1 h in Ar-4H₂ gas, and then hot rolled at 1100° C. to 50% reduction using 5-10% reduction per rolling pass. Test samples measuring 20 mm×10 mm×0.75 to 1.2 mm for lab furnace evaluation and 25 mm×12.5 mm×0.75 to 1.2 mm with a 4 mm diameter hole for in-situ cookstove evaluation were electro-discharged machine (EDM) cut. For the thin sheet 316L, 310S, and FeCrAlY samples, the as-processed surface finish was retained for corrosion testing. For all of the other alloy product forms, the EDM-cut surface finish of the test sample faces was polished to 600 grit finish by standard metallographic techniques in water using SiC grinding papers.

Corrosion Testing

Corrosion evaluation under cookstove-relevant conditions was studied by two methods: 1) lab furnace testing and 2) in-situ exposure in an operating cookstove. The lab furnace testing was conducted in air with 10 volume percent (vol. %) H₂O to simulate water vapor release from burning biomass, and direct deposition of salt onto the test samples to simulate the burning of highly corrosive biomass feedstocks. In particular, relatively high levels of salt species are encountered in many types of biomass and can lead to significantly accelerated alloy corrosion rates. The in-situ cookstove testing was conducted using wood fuel that was pre-soaked in a saltwater solution to yield accelerated, highly corrosive conditions.

Test sample corrosion was evaluated by two approaches: specific mass change (balance accuracy of ˜±0.04 mg or 0.01 mg/cm²), which can be measured nondestructively with the sample returned to testing for additional exposure time, and metal loss measurements, which required removal of a given sample from testing for destructive, cross-section analysis to determine the thickness of intact metal remaining after a given exposure period. During corrosion, the test sample mass increases from incorporation of oxygen into the alloy from the environment and the subsequent formation of oxides on and/or within the sample (also holds true for nitrogen, sulfur, etc. species). Volatilization and/or flaking off of the oxide, etc. results in mass losses. In this manner the kinetics of corrosion can be followed by mass change measurements. Combined with cross-section metal loss measurements, a more complete picture of corrosion kinetics can be obtained.

Corrosion Data

Corrosion data from 600° C. lab furnace testing, 800° C. lab furnace testing, and in-situ cookstove testing are shown in FIG. 4, FIG. 5; FIG. 6. All alloy samples in both lab furnace and in-situ cookstove testing showed loose, non-adherent oxide scales typical of Fe-oxide formation/relatively rapid corrosion developing within the first 100 h of exposure under all test conditions evaluated. In the 600° C. lab furnace testing (FIG. 4), specific mass changes were on the order of ±5 mg/cm² over the course of 1000 h of total exposure (100 h cycles) for all commercial and developmental alloys, with the exception of type 201 stainless steel which exhibited extensive corrosion and specific mass loss in excess of about 50 mg/cm². Cross-section metal loss measurements (FIG. 4b, d) proved far more useful in distinguishing the relative degree of alloy corrosion resistance than the specific mass change measurements. This was likely due to a combination of mass gain from oxide scale formation and mass loss from oxide scale spallation (or volatilization), which can result in low net mass changes despite appreciable corrosion. The cross-sectioning analysis also permits assessment of the extent of internal corrosion vs. external scaling.

Lab furnace Corrosion Testing

Electrical-resistance heated lab furnace corrosion exposures were conducted in 100 h cycles at 600 and 800° C. in air with 10 vol. % water vapor. Sample mass changes were measured after every 100 h furnace cycle (air cooling of the test samples). Three replicates of each alloy were exposed, with samples sequentially removed from testing after 100 h, 500 h, and 1000 h for cross-section analysis. (The 1000 h of testing represents ˜2.7 hot hours per day of cookstove use for a year; although it is important to note that the corrosive conditions employed in the present work were likely accelerated by the use of salt species). The water vapor was introduced by atomization of distilled water into the flowing gas stream above its condensation temperature. The tests were conducted in a ˜8.1 cm diameter alumina tube furnace using flow rates of ˜925 or 760 cm³/min (air) and ˜4-5 cm³/h (water) at 600 and 800° C., respectively. To introduce salt to the test samples, a 3.5 wt. % salt solution was made using distilled water and Instant Ocean® Sea Salt (Blacksburg, Va. 24060-6671 USA), a product which simulates the salt species (Na, K, Mg, etc. chlorides) present in natural seawater.

The test samples were placed across alumina rods in a flat alumina tray (see FIG. 1a). The top surfaces of the test samples were sprayed with salt water and allowed to air dry prior to initial exposure in the furnace, and salt water was re-applied after every 100 h cycle. (Mass change measurements for corrosion assessment were taken after every cycle prior to adding salt). The salt solution was sprayed only on the top-oriented test sample face (exposed surface), and the same top face was maintained for salting across all test cycles. (Orientation of the samples was maintained to salt the same top face across all test cycles.) Salt additions after drying were typically in the range of ˜0.5 to 1.5 mg/cm² of salt per top-exposed sample face.

In-Situ Cookstove Testing

The in-situ cookstove evaluation was performed using two Envirofit International (Fort Collins, Colo. USA) B1200 ceramic lined, rocket elbow type stoves fed with salted lab wood. Test fixtures were created to hang samples at a fixed depth inside the combustion chamber of each stove such that they would be exposed directly to the flame region (FIG. 1b). Each fixture held up to fifteen samples at a time. In order to ensure even exposure, fixtures were rotated every hour of burn time. Test samples were removed after every 50 h of burn time, mass change registered, and returned to testing. As with the lab furnace evaluation, three replicates of each alloy were exposed, with samples sequentially removed from testing after 100, 500, and 1000 h for cross-section analysis.

The “lab wood” burned in the in-situ test was ˜20 mm×by 20 mm cross-section by 305 mm length clear Pine trim. In order to produce accelerated corrosion conditions, a procedure was developed to controllably introduce salt (halogen) content into the lab wood to a level similar to that found in a biomass fuel known to cause highly corrosive conditions. Two types of Haitian charcoal known to produce highly corrosive combustor conditions were analyzed for halogen content: Mangrove lump charcoal and Chabon Ticadaie briquette charcoal. The fuels produced 760 μg and 1390 μg of halogen per gram of charcoal, respectively, yielding an average value of 1075 μg halogen per gram of charcoal. (It should be noted that high halogen is found in a wide range of biomass around the world and is not limited only to Haitian charcoal). The average 1075 μg/g was used as a target value for the halogen content of the lab clear Pine wood. The target halogen content was achieved by soaking the Pine wood in a tank of synthetic sea water solution for several days. Water was continuously circulated by a pump to promote even mixing of the salt. After soaking, the wood was air-dried for at least 2 days in ambient air and then dried in an oven at 104° C. for an additional 2 days before burning. Instant Ocean® Sea Salt was again used to create the salting solution. An initial target value for salt concentration in the water was determined by analyzing the halogen content of processed wood that was soaked in 1.9 liter jars with different salt concentrations. Based on this initial salt concentration target value, several small batches of salted Pine wood were produced and analyzed for halogen content. Because the halogen content of the initial salted fuel was lower than desired (Table 2), the salt quantities were increased in subsequent batches in order to more closely match the halogen content of the Haitian charcoals. A total of ten batches of wood were processed and burned in the in-situ cookstove testing. An average value of ˜1070 μg of halogen per gram of wood was achieved (Table 2).

TABLE 2
Halogen content of lab-salted wood.
				Halogen
Batch #	Water (gal)	Salt (g)	Wood (kg)	Content (μg/g)
1	36	400	36	281
2	34	400	40	650
3	33	350	42	776
4	150	3675	113	700
5	165	5250	192	821
6	190	5250	199	1780
7	180	5250	117	548
8	175	5250	218	781
9	197	5250	175	1550
10	197	5250	239	1275
Weighted Avg	169	4702	177	1072
All Batches
Weighted Avg*	184	5250	198	1160
Batches 5-10
Std. Dev.	13	0	44	485
Batches 5-10

Soaking process was being developed during batches 1-5, and thus halogen content differed from batch to batch. The final salting process was used for batches 5-10, which accounted for 83% of the total wood used in the in-situ cookstove alloy corrosion testing.

Three salted clear Pine wood sticks were fed at a time into the cookstove, and sticks were kept so they were always touching to reduce variability in burn rate due to stick spacing. After every 2-3 sets of sticks burned, char and ash that had accumulated in the chamber was removed. Each day of testing, cookstoves were burned continuously for an average of ˜6 h. The average fuel consumption rate was 570 g/h. To determine the range of temperatures that the alloy test samples would experience, a thermocouple was placed inside the chimney of each stove at the same height as the coupon fixture. Typical combustion chamber temperature profiles for the cookstoves, where test coupons were placed, are shown in FIG. 2. The average gas temperature range during steady state in-situ testing was 663° C.±85° C. If transient operation events such as startup, shutdown, and char/ash removal are included in the chamber temperature calculations, the average exposure temperature decreases to an average of 609° C.±188° C. As the majority of the testing was performed in the steady state configuration, under controlled conditions, the average exposure temperature and variability was likely closer to 663° C.±85° C. than the 609° C.±188° C. over the course of the 1000 h of testing.

FIG. 3 shows in-situ cookstove corrosion specific mass change data for the state-of-the-art FeCrAlY alloy (Table 1) when the stove was burned with as-received clear Pine wood vs. salted clear Pine wood. Parabolic-like (rate decreasing with time) corrosion kinetics with relatively low specific mass gains were observed for the FeCrAlY when as-received clear Pine wood was burned, consistent with protective oxide scale formation. In contrast, much more rapid mass gains and a period of mass loss resulting from oxide scale spallation were observed when the salted clear Pine wood was burned. The salting procedure adopted therefore clearly induced a more corrosive test environment to serve as an accelerated test method for evaluation of candidate combustor materials.

Cross-Section Analysis of Corroded Samples

Exposed test samples were cross-sectioned by low-speed diamond saw and prepared by standard metallographic techniques (oil-based rather than aqueous polishing media was used to avoid dissolution of possible chloride and related corrosion products during sample preparation). The cross-sections were analyzed by light microscopy, scanning electron microscopy (SEM) with energy dispersive x-ray spectroscopy (EDS), and electron probe microanalysis (EPMA) using both EDS and wavelength dispersive spectroscopy (WDS). Initial sample thickness was measured with a micrometer. Cross-section thickness measurements of corroded samples were made using an optical measurescope (light microscope with digitized micrometer stage measurement attachment). The test sample cross-sections along the 10-12.5 mm sample width were divided into 3 regions, not including 1 mm from the sample end corners. In each of the 3 sample regions, the area of greatest corrosion attack was selected for measurement of intact metal remaining in cross-section. The boundary of the intact metal remaining was defined as the point at which the underlying metal was free of oxide scale and internal attack. The average of the three locations of greatest attack was used to plot metal loss vs. time in FIG. 4, FIG. 5; FIG. 6 (reported in the plots as ±1 standard deviation).

Results

Consistent with the specific mass change measurement of −50 mg/cm², the type 201 stainless steel exhibited extensive cross-sectional metal loss, ˜−240 μm after 1000 h of exposure in the 600° C. lab furnace test. However, despite relatively low specific mass changes (−1 to 5 mg/cm²), significant cross-section metal loss was also observed for the three AFA alloys, with ˜−170 μm, −210 μm, and −330 μm metal loss for AFA-25Ni, AFA-20Ni, and AFA-12Ni, respectively. The commercial stainless steels exhibited lower corrosion rates, with metal losses of about −60 μm, −70 μm, and −90 μm after 1000 h for type 446, type 310S, and type 316L, respectively.

The lowest metal losses for the Fe-base alloys in the 600° C. lab furnace test were registered for the commercial FeCrAlY, which had metal losses of about −30 μm after 500 h, and −20 μm after 1000 h. The 500 and 1000 h data are from different FeCrAlY test samples and are in the range expected for sample-to-sample variation and initial test sample thickness measurement scatter, estimated to be about −10 to −20 μm. Pre-oxidation of the FeCrAlY resulted in an even lower metal loss measurement, −10 μm after 1000 h. The developmental FeCrSi, FeCrSiAl, and Fe25CrAl alloys also exhibited low metal losses, in the range of −25 to −35 μm (FIG. 4d). As a whole, these metal loss measurement values for the Fe—Cr—Al and Fe—Cr—Si alloys indicate similar, relatively good levels of corrosion resistance under these accelerated corrosion conditions. Consistent with its known good corrosion resistance to high-temperature salt species attack under some conditions, the pure Ni showed little visible evidence of corrosion, with specific mass changes of only 0.3 mg/cm² (Ni tested only to 500 h total exposure at 600° C.) and metal loss of −11 μm, which is within the expected error for the cross-section metal loss corrosion assessment method.

Much faster corrosion rates were observed in the 800° C. lab furnace testing (FIG. 5), where evaluation of most alloys stopped after 500 h of exposure due to excessive corrosion (mass losses in excess of −40 to −50 mg/cm² and cross-section metal losses in excess of several hundred μm). Of the alloys tested to 1000 h, (FeCrAlY, pre-oxidized FeCrAlY, 310S, FeCrAlSi, FeCrSi, Fe25CrAl, and pure Ni), only the FeCrSi and pure Ni samples exhibited good corrosion resistance, with 1000 h metal losses of about −160 μm and about −135 μm, respectively. The FeCrAlY and 310S alloy samples were consumed through-thickness in some cross-section locations (starting sample thickness of ˜1 mm and ˜0.75 mm, respectively). The pre-oxidized FeCrAlY and the Fe25CrAl samples exhibited metal losses in excess of about −500 μm at 1000 h, and the FeCrAlSi sample about −370 μm.

FIG. 6 shows corrosion data for the in-situ cookstove testing. A comparison with FIG. 4; FIG. 5 shows that the corrosion rates were essentially intermediate between the 600° C. and 800° C. lab furnace testing, consistent with the about 663° C. average temperature of the in-situ cookstove testing (FIG. 2). Relative alloy corrosion resistance trends were generally similar to the lab furnace testing, indicating the utility of the lab furnace protocol as a screening tool to down -select candidate alloys for the more costly and time intensive in-situ cookstove testing. Type 201 stainless steel, type 316L stainless steel, and the 12 and 20Ni AFA alloys all exhibited relatively poor corrosion resistance in the in-situ cookstove testing, with metal losses in excess of −200 μm after only 500 h of exposure, consistent with the lab furnace trends.

The lowest corrosion rates in the in-situ cookstove testing were exhibited by the FeCrAlY, pre-oxidized FeCrAlY, FeCrAlSi, and FeCrSi alloys, with metals losses in the −100 to −150 μm range after 1000 h (FIG. 6). The types 310S and 446 stainless steels exhibited moderately worse corrosion resistance, with metal loss values of −190 μm and −230 μm after 1000 h. The two major exceptions between lab furnace and in-situ cookstove testing were the pure Ni and the 25Ni AFA alloy samples. Despite exhibiting the best corrosion resistance in the lab furnace testing, the pure Ni suffered from −300 μm metal loss after only 500 h in the in-situ cookstove testing. Conversely, although the 25Ni AFA alloy performed poorly in the lab furnace testing, at 500 h of in-situ cookstove testing, the 25Ni AFA sample showed relatively moderate metal loss of only −62 μm (this sample was not run to longer exposure times).

Cross-Section Microstructures after Corrosion

FIG. 7 shows SEM backscattered electron mode cross-section images of type 310S stainless steel after 500 h of exposure in 600° C. lab furnace, 800° C. lab furnace, and in-situ cookstove testing. Also shown for comparison is a cross-section of a 310S cookstove combustor component (FIG. 7d) from field operation of a cookstove (information on time of operation was not available; the 310S combustor analyzed was not from the same 310S alloy batch used in the other experiments). The corrosion features were quite similar across the lab furnace, in-situ cookstove, and field-used 310S, with loosely adherent external oxide scales tens of microns thick overlying a zone of internal attack.

Elemental mapping by EPMA for 310S after field use, 1000 h in-situ cookstove testing, and 500 h lab furnace testing at 800° C. are shown in FIG. 8. The oxide scales were Fe-rich, with Cr and Si enrichment in the inner oxide scale regions. The internal attack regions were Cr- and Si-rich oxides, with the surrounding metal rich in Ni and depleted in Cr and Si. For the field-operated and the in-situ cookstove test 310S, S (released by the burning biomass) was also detected in the internal attack zone. Despite the near-absence of S observed in the lab furnace exposure 310S sample, internal attack of Cr and Si was still observed in 310S, which suggests that salt species alone were sufficient to make the alloy susceptible to extensive internal attack. Overall, this comparison indicates that the lab furnace, in-situ cookstove, and the field-operated cookstove all exhibited similar corrosion behavior for type 310S stainless steel, and provides support for the utility of the developed lab furnace and in-situ cookstove test protocols as alloy screening approaches.

Similar Fe-rich oxide scale and internal attack features for lab furnace and in-situ cookstove exposures also were observed for all stainless steel, AFA, and Fe—Cr—X (X═Al, Al+Si) alloys studied, with the notable exception of the pure Ni. (The FeCrSi samples exhibited similar oxide scale structure, but no internal attack in both lab furnace and in-situ cookstove exposures). FIG. 9 shows cross-section images for pure Ni after 500 h 600° C. lab furnace, 800° C. lab furnace, and in-situ cookstove testing. The excellent corrosion resistance in the lab furnace testing resulted from formation of an external Ni-rich oxide scale, with minor Ni metal grain boundary internal attack evident in the 800° C. exposure and essentially none in the 600° C. exposure. However, extensive internal attack along the Ni metal grain boundaries was observed in the in-situ cookstove testing, which resulted in high metal loss values (FIG. 6). Elemental mapping indicated that the internal attack at the alloy grain boundaries was Ni and O rich. Sulfur was detected near the metal/external scale interface, but not in the internal grain boundary attack zone. Local grain boundary attack areas containing fine Si, C, Fe rich particles were occasionally detected (Si, C, Fe, Mn, Cr, Co, Ca, Mg present as Ni metal impurities, Table 1), but it was not clear if they were definitively associated with the corrosion attack.

FIG. 10 shows a compilation of SEM backscattered electron mode cross-section images for several key alloys after 1000 h of in-situ cookstove testing: commercial stainless steels types 316L and 446, commercial FeCrAIY, and the developmental model alloys Fe25CrAl, FeCrAlSi, and FeCrSi. The 316L cross-section (FIG. 10a) was similar to the 310S (FIG. 8b), with an external Fe-rich oxide scale overlying an internal attack zone preferentially along alloy grain boundaries. The depths of the internal attack zones for the 316L and 310S samples were similar, in the range of about −80 μm, suggesting that the better corrosion resistance of 310S (FIG. 6, about −190 μm metal loss for 310S after 1000 h of in-situ cookstove testing vs about −370 μm for 316L) is due to slower oxide scaling from the higher Cr and Ni levels in type 310S (Table 1). The ferritic 446, FeCrAlY, Fe25CrAl, and FeCrAlSi samples all showed both external scale formation, and internal attack, again primarily along alloy grain boundaries. The extent of internal attack was moderately less than that observed with the austenitic type 310S and 316L stainless steels, typically in the range of about 20-80 μm. In contrast, the FeCrSi sample exhibited external Fe-rich oxide scaling but not internal attack.

The resistance of the FeCrSi alloy to internal attack was the source of its superior corrosion resistance in the 800° C. lab furnace testing (FIG. 5). Low magnification, light microscopy cross-sections of the test sample ends for 310S, FeCrAlY, and FeCrSi after 1000 h of lab furnace testing are shown in FIG. 11. Both the type 310S and FeCrAlY alloys suffered from a transition to extensive internal attack, with the entire sample thickness consumed in some locations. In contrast, no internal attack was observed for the FeCrSi alloy.

FIG. 12 shows backscattered electron mode cross-sections and corresponding EPMA elemental mapping for FeCrAlY after 1000 h of in-situ cookstove testing, FeCrSi after 1000 h of in-situ cookstove testing, and FeCrSi after 500 h of 800° C. lab furnace testing. In the in-situ cookstove testing, the scales formed on both FeCrAlY and FeCrSi were Fe-rich, with fine local porosity associated with Na ingress (Cl also detected in these regions). Chromium and trace Si (Si not shown in maps) were also detected in the external Fe-rich scale formed on FeCrAlY. At the alloy-scale interface in the FeCrAlY (FIG. 12a), locally quasi-continuous regions of Al-rich oxide were observed, overlying an internal attack zone rich in Al—O—S extending inward along alloy grain boundaries (some Si enrichment was also detected in this region). Fine, internal oxidation (and possibly internal nitridation, N not specifically analyzed for) particles associated with Al were also observed between the internally attacked alloy grain boundary regions. The elemental mapping indicated that the alloy in this internal attack region was depleted in Cr and Al relative to the bulk alloy.

For the FeCrSi alloy in both the in-situ cookstove exposure and the 800° C. lab furnace testing, Si was detected in the outer, Fe-rich oxide scales (FIG. 12b, c). At the alloy/scale interface, both Cr and Si enrichment were also observed. Particularly in the in-situ cookstove test exposure, S enrichment was also detected at the alloy/scale interface. No internal attack was evident in the FeCrSi alloy, although the elemental maps indicated Cr depletion of the alloy beneath the oxide scale, particularly in the 800° C. lab furnace exposure (FIG. 12c).

Discussion

A qualitative ranking of alloys in order from best to worst corrosion resistance based primarily on the cross-section metal loss measurements in conjunction with cross-section microstructure features is provided in Table 3 for each of the 3 test conditions: in-situ cookstove, 600° C. lab furnace, 800° C. lab furnace. The alumina-forming FeCrAl-class alloys exhibited better corrosion resistance than the chromia-forming austenitic 200 and 300 series and ferritic 400 series stainless steels examined under biomass cookstove relevant conditions, although a discrete, thin, and protective alumina-based scale was not formed. Rather, the scales formed by the FeCrAl alloys were still thick and Fe-rich, with enrichment of Al-oxides near the alloy-scale interface, and internal attack of Al beneath the scale (FIG. 10; FIG. 12).

TABLE 3
Qualitative ranking of relative corrosion resistance of alloys in the 3 conditions studied.
Most corrosion resistant alloys listed at top, least at bottom. Multiple alloys listed in the same
table cell indicate similar levels of corrosion resistance. Note that all of the raw specific
mass change and metal loss corrosion data are available in the supplementary material.
In-Situ Cookstove	600° C. Lab Furnace	800° C. Lab Furnace
FeCrAlY, Pre-oxidized FeCrAlY,	Pure Ni, Pre-oxidized	Pure Ni, FeCrSi
FeCrAlSi, FeCrSi	FeCrAlY
AFA-25Ni, 310S, 446, 25Cr—	FeCrAlY, FeCrAlSi, FeCrSi,	FeCrAlSi
FeCrAl	25Cr—FeCrAl
316L, 201, AFA-20Ni	446, 310S, 316L	25Cr—FeCrAl, FeCrAlY, Pre-
		oxidized FeCrAlY,
AFA-12Ni, Pure Ni	AFA-25Ni, AFA-20Ni	316L, 446, 310S
	201	201, AFA-25Ni, AFA-20Ni,
		AFA-12Ni
	AFA-12Ni

Pre-oxidation of the commercial FeCrAlY alloy to initially form an exclusive alumina surface layer moderately enhanced the high-temperature corrosion resistance of the FeCrAlY, although the benefit was lost (pre-formed alumina scale degraded) with longer exposure times/higher exposure temperatures. These findings are consistent with literature reports that alumina-forming alloys frequently perform better than chromia-forming alloys under biomass—relevant, high-temperature corrosion conditions, even when appreciable corrosion with thick Fe-rich outer oxide scales are encountered. The alumina-forming austenitic (AFA) stainless steels exhibited significantly worse corrosion resistance than did the ferritic FeCrAl-class alloys. The AFA alloys have lower levels of Al and Cr than FeCrAl alloys (Table 1) in order to form an austenitic-based microstructure to achieve good high-temperature creep strength. A consequence of the lower Al and Cr levels is that AFA falls nearer to the composition borderline for protective alumina scale formation, which makes them more susceptible than FeCrAl to corrosive attack under highly aggressive conditions.

With the exception of the pure Ni, the relative rankings of the alloys were similar for the lab furnace and in-situ cookstove test methods. This finding suggests that the lab furnace test protocol developed is a useful initial screening method to evaluate candidate alloys for improved biomass cookstove combustors. The in-situ cookstove method is still the preferred alloy evaluation method, with the use of controlled salted wood yielding desired accelerated corrosion conditions (FIG. 3), but is also more costly and time consuming to run due to the need to constantly monitor the cookstove and feed it with new wood fuel every about 20 min. The in-situ cookstove testing also yielded similar S ingress characteristics to those observed in field-operated cookstove combustors (FIG. 8), which may play a significant role in the corrosion mechanism for some alloys, particularly if low-melting Ni—S phases are formed, although salt species-related attack dominated the alloys studied in the accelerated corrosion conditions of the present work. The extensive metal grain boundary attack (FIG. 9) observed for the pure Ni in the in-situ cookstove exposure (but not the lab furnace testing) did not appear to be the result of S-related attack based on initial EPMA mapping. However, further characterization work is needed to better understand the nature and mechanism of this internal grain boundary attack, and to identify which species from the biomass (S, C, or other) caused the attack.

The greatest level of corrosion resistance in the present work was exhibited by the developmental FeCrSi alloy (Table 1), particularly in the 800° C. lab furnace testing (FIG. 5, FIG. 11; FIG. 12). It has been previously shown that sufficient additions of Si to ferritic Fe—Cr base alloys can beneficially improve high-temperature corrosion resistance in biomass combustion, waste incinerator, and related environments involving high-temperature exposure to salt species. Model alloys such as Fe-15Cr-5Si wt. % and coatings based on Fe, (20-50) Cr, (3-10) Si wt. % have been reported to exhibit significantly enhanced high-temperature corrosion resistance in the presence of salt species, beyond that achieved with Al additions. The beneficial effects of Si have been postulated to be related to both increased Cr diffusivity in the alloy to favor formation of a more dense and protective Cr₂O₃ scale and formation of an inner layer rich in SiO₂ at the alloy-scale interface (SiO₂ does not form at the alloy-scale interface in FeCrAlSi alloys because of the greater thermodynamic stability with oxygen of Al/Al₂O₃ vs Si/SiO₂). Enrichment of Si near the alloy-scale interface was observed for the FeCrSi alloy in the present work (FIG. 12), and both local SiO₂ formation and increased Cr diffusivity effects are consistent with the observed resistance to internal attack by the FeCrSi alloy.

II. Alloy Compositions and Manufacturability

The ability to hot or cold roll the disclosed alloys was tested. For these experiments, a second predicted phase equilibrium was generated, and is depicted at FIG. 13. Using these prediction, experiments were performed where the disclosed alloy was hot-rolled at 1100° C. (the as-received material), and then additionally annealed (tempered), post rolling, at 950 either 850° C. for 1 h, followed by water quenching. Thereafter, the rolled alloys were checked at the microstructure level, and for hardness.

FIG. 14 shows microstructure analysis of the Fe-15Cr-2.4Si-0.5Mn-0.5Ti-0.57C wt. % alloy and the effect of annealing after hot rolling. Panels of FIG. 14, left to right show the microstructures at increasing magnification: 50×, 200×, and 1000×. The top row of panels show microstructure of alloy hot rolled at 1100° C., while the middle and bottom rows depict samples annealed at 950° C. and 850° C., respectively, for one hour post hot rolling. These micrographs demonstrate that the microstructure becomes more consistent post-annealing.

FIG. 15 is a graph of Vickers vs. temperature (hot rolling and annealing) for the Fe-15Cr-2.4Si-0.5Mn-0.5Ti-0.57C wt. % alloy. As shown in this graph, the as hot-rolled (HR) consisted of not only pearlite with ˜283 HV but also a large volume of retained austenite with ˜480 HV. These are a source of deformation resistance at room temperature (RT) for cold rolling. Annealing (or tempering) at 950° C. or 850° C. resulted in eliminating the retained austenite species, with a concomitant drop in Vickers Hardness, which greatly increased amenability to cold rolling.

FIG. 16 shows micrographs of the edge surfaces of two annealed samples of the Fe-15Cr-2.4Si-0.5Mn-0.5Ti-0.57C wt. % alloy after 90% cold rolling. The top micrograph shows the alloy annealed at 950° C. for 1 h followed by water quenching, and cold rolling to reduce the sample thickness by about 90%. The bottom micrograph shows the second annealed alloy, treated at 850° C. for 1 h followed by water quenching, and cold rolling. Only minor edge cracking was observed after the 850° C. annealing and 90% cold rolling, with better edge quality and even less local cracking for the 950° C. annealed and cold rolled alloy.

These experiments demonstrate that the disclosed alloys are amenable to cold rolling, especially when treated with a high-temperature annealing step after hot-rolling. Sheet and foil product 0.75 mm thick and 0.25 mm thick were readily produced by cold rolling.

Additional testing was performed to analyze ductility of disclosed alloys. Specifically, three alloys were studied after 5 min of annealing at 800° C. and air cooling (AC). The three alloys were labeled Cook2 (Fe-15Cr-2.4Si-0.5Mn-0.5Ti-0.57C wt. %), Cook4 (Fe-15Cr-2Si-0.5Mn-0.5Ti-0.1C wt. %), and Cook7 (Fe-15Cr-3Si-0.5Mn-0.5Ti-0.3C wt. %). Two samples of each alloy were tested for ductility: 1—an as-cold rolled sample, and an annealed sample. Results are presented below in Table 4, and graphed in FIG. 17 (top row). Testing (presented in FIG. 17, bottom graph) indicated that hardness drops and is stable after 5 min of annealing, with no change to hardness after 30 minutes of annealing.

TABLE 4
		YS (MPa)	UTS (MPa)	Uniform strain, %	Failure strain, %
Cook2	As rolled	1225	1248	0.7	5
	800 C./5 min/AC	548	764	17	30
Cook4	As rolled	1067	1068	0.2	8
	800 C./5 min/AC	451	611	16	30
Cook7	As rolled	1174	1186	0.6	5
	800 C./5 min/AC	521	714	16	28
*Tested by using SS-J2 specimens

These experiments helped produce a method for processing the disclosed that is useful in creating a variety of products from the disclosed alloys, for example sheets and foils. In many embodiments, a first step may include a homogenization and hot-deformation step (e.g. forging or rolling) at or above 1100° C., for example from about 600 to about 1200° C. This step may aid in reducing or eliminating any segregation issues that may result from solidification. This step may also help to break coarse grain structure in the as-cast materials. In many embodiments, this step may be followed by a cooling step, where the temperature of the alloy is cooled, (e.g. by air-cooling) to about room temperature (i.e. about 20° C.). This cooling may help avoid or reduce issues, such as crack formation during cooling and/or during the following thermo-mechanical treatment. In many embodiments, where the alloy may comprise more than about 0.2 wt. % C, additional annealing within a temperature range of about 700° C. to about 1200° C., for example from about 850-950° C., may be useful. This step may be followed by water quenching, before proceeding to additional thermo and/or mechanical treatments. Annealing the disclosed alloys within this temperature range may help to (a) soften the martensitically transformed grains formed during cooling after homogenization and hot-deformation, and (b) avoid or lessen sensitization through the decomposition of excess amounts of carbides (e.g. M23C6).

In most embodiments, the maximum hardness of the alloys after annealing may be less than about 500 HV, 450 HV, 400 HV, 350 HV, 3000 HV, 250 HV, or 20 HV (or equivalent hardness). In a preferred embodiment the maximum hardness will be about 250 HV. In many embodiments, the disclosed alloys may then be subjected to additional hot-rolling (for example in a temperature range of about 750-1200° C., and preferably between about 850-950° C.). This may aid in reducing the thickness of the disclosed alloys by greater than about 60%, 65%, 70%, 75%, 80%, 85%, 90%, or 95%, and less than about 99%, 95%, 90%, 85%, 80%, 75%, 70%, or 65% of the starting thickness. Alternatively, the disclosed alloy may be cold-rolled after obtaining the desired hardness.

In many embodiments, a final annealing may be useful after cold or hot rolling. In these embodiments, an annealing may help to release work hardening. In many embodiments, this annealing may be at about 700-1200° C., for example 800° C., and may be from about 1 to about 30 min, for example about 5 min. This final annealing may be followed by cooling, for example air cooling or water quenching. Air cooling after an optional final annealing may help to restore the tensile ductility (elongation to fracture) by from about 1% to about 40%, and preferably about 5-8% for as-rolled sheets and about 30% for annealed sheets. In many embodiments, the tensile ductility may be improved by greater than about 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25% or 30%, and less than about 40%, 35%, 30%, 25%, 20%, 15%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, or 2%.

In some embodiments, a post rolling annealing step may be above about 800° C., for example about 1000° C. to 1100° C. This high temperature annealing may help to put carbon back into supersaturated solution, which may then aid in enhancing corrosion resistance further.

While multiple embodiments are disclosed, still other embodiments of the present invention will become apparent to those skilled in the art from the following detailed description. As will be apparent, the invention is capable of modifications in various obvious aspects, all without departing from the spirit and scope of the present invention. Accordingly, the detailed description is to be regarded as illustrative in nature and not restrictive.

All references disclosed herein, whether patent or non-patent, are hereby incorporated by reference as if each was included at its citation, in its entirety. In case of conflict between reference and specification, the present specification, including definitions, will control.

Although the present disclosure has been described with a certain degree of particularity, it is understood the disclosure has been made by way of example, and changes in detail or structure may be made without departing from the spirit of the disclosure as defined in the appended claims.

Claims

1. An alloy comprising:

about 13-17 wt % chromium (Cr);

about 2.0-3.5 wt % silicon (Si);

about 0.2-1.0 wt % manganese (Mn);

about 0.3-0.7 wt % titanium (Ti);

about 0.1-0.6 wt % carbon (C); and

wherein the remainder wt % is substantially iron (Fe).

2. The alloy of claim 1, further comprising boron.

3. The alloy of claim 2, wherein the boron concentration is between about 0.01 and 0.5 wt %.

4. The alloy of claim 3, wherein the boron concentration is between about 0.05 to 0.15 wt. %.

5. The alloy of claim 4, further comprising nickel.

6. The alloy of claim 5, wherein the nickel concentration is between about 0.1 and 1.0 wt %.

7. The alloy of claim 6, wherein the carbon concentration is between about 0.3 to 0.6C wt. %.

8. The alloy of claim 7, wherein the carbon concentration is between about 0.4 to 0.6C wt. %.

9. The alloy of claim 5, wherein the carbon concentration is between about 0.5 to 0.6C wt. %.

10. The alloy of claim 1, wherein the alloy concentrations are about 15-16%% Cr, 2.4-2.6% Si, 0.4-0.6% Mn, 0.4-0.6% Ti, 0.5-6% C, and 78-82% Fe.

11. The alloy of claim 1, wherein the alloy concentration is about 14-16% Cr, 2.4-3.2% Si, 0.4-0.6% Mn, 0.4-0.6% Ti, 0.4-0.6% C, and 78-82% Fe.

12. The alloy of claim 1, wherein the alloy concentration is about 14-16% Cr, 2.4-3.2% Si, 0.4-0.6% Mn, 0.4-0.6% Ti, 0.3-0.6% C, and 78-82% Fe.

13. A method of protecting a surface in a corrosive environment, the method comprising the steps of:

providing a surface,

coating or constructing the surface with an alloy, the alloy comprising: about 13-17 wt % chromium (Cr); about 2.0-3.5 wt % silicon (Si); about 0.2-1.0 wt % manganese (Mn); about 0.3-.0.7 wt % titanium (Ti); about 0.1-0.6 wt % carbon (C); and wherein the remainder wt % is substantially iron (Fe),

subjecting the coated surface to a corrosive environment, wherein the environment is at a temperature above about 400° C.

14. The method of claim 13, wherein the alloy comprises all or part of a combustor of a stove.

15. The method of claim 14, wherein the stove is a biomass stove.

16. The method of claim 13, wherein the corrosive environment is a combustor of an energy conversion device.

17. A method of producing an alloy product, the method comprising:

heating the alloy to greater than about 1000° C.;

hot rolling the alloy to reduce its thickness by more than about 10%;

cooling the alloy to less than about 100° C.;

heating the alloy to greater than about 800° C.;

cooling the alloy to less than about 100° C.;

rolling the alloy to reduce its thickness by greater than about 50%; and

increasing the temperature of the alloy to greater than about 800° C.;

maintaining the increased temperature for at least about 5 min.

18. The method of claim 17, wherein the alloy comprises about 13-17 wt % chromium (Cr); about 2.0-3.5 wt % silicon (Si); about 0.2-1.0 wt % manganese (Mn); about 0.3-0.7 wt % titanium (Ti); about 0.1-0.6 wt % carbon (C); and wherein the remainder wt % is substantially iron (Fe).

19. The method of claim 18, wherein the alloy further comprises boron at a concentration between about 0.01 and 0.5 wt %.

20. The method of claim 19, wherein the alloy further comprises nickel at a concentration between about 0.1 and 1.0 wt %.