GNEP

Overview

Cladding and Duct Materials for Advanced Nuclear Recycle Reactors T.R. Allen, J.T. Busby, R.L. Klueh, S.A. Maloy, and M.B. Toloczko

The expanded use of nuclear energy without risk of nuclear weapons proliferation and with safe nuclear waste disposal is a primary goal of the Global Nuclear Energy Partnership (GNEP). To achieve that goal the GNEP is exploring advanced technologies for recycling spent nuclear fuel that do not separate pure plutonium, and advanced reactors that consume transuranic elements from recycled spent fuel. The GNEP’s objectives will place high demands on reactor clad and structural materials. This article discusses the materials requirements of the GNEP’s advanced nuclear recycle reactors program. INTRODUCTION The Global Nuclear Energy Partnership (GNEP) seeks to bring about a significant, wide-scale use of nuclear energy while decreasing the risk of nuclear weapons proliferation and effectively addressing the challenge of nuclear waste disposal. Two primary objectives of GNEP include developing, demonstrating, and deploying: v Advanced technologies for recycling spent nuclear fuel that do not separate pure plutonium, that reduce or eliminate excess stocks of civilian plutonium, and that draw down existing stocks of civilian spent fuel. Such advanced fuel cycle technologies would substantially reduce nuclear waste, simplify its disposition, and in the United States would help ensure the need for only one geologic repository through the end of this century. The most significant repository benefits can be achieved by removing the very long-lived minor actinides and recycling them as part of the fuel for fast 2008 January • JOM

reactors. Removing the fission products cesium and strontium from the high-level waste stream and allowing them to decay separately can achieve further repository benefit. v Advanced reactors that consume transuranic elements from recycled spent fuel. This would be a fast reactor tuned to transmute (burn) actinides in an actinidebearing fuel. This fast reactor, termed advanced burner reactor (ABR) or advanced recycle reactor (ARR), is designed to transform the actinides in a way that makes them easier to store as waste and are not a proliferation concern, while also producing electricity.1 These objectives will place high demands on reactor clad and structural materials. Achieving these goals in an efficient manner requires irradiatHow would you… …describe the overall significance of this paper? This paper provides a concise description of material challenges for Generation IV systems. …describe this work to a materials science and engineering professional with no experience in your technical specialty? This paper describes the operating environment for proposed advanced reactor concepts and the limitations and design challenges of current materials. …describe this work to a layperson? This paper describes the operating environment for proposed advanced reactor concepts and the challenges in developing materials for these systems.

www.tms.org/jom.html

ing fuels in reactors to high burnups (greater than 20%). Thus, the cladding and duct materials for these fuels must be able to withstand significant radiation damage incurred during irradiation (greater than 200 displacements per atom [dpa]). Presently, fuels are limited to a maximum burnup of 20% because of clad material limitations (e.g., void swelling, creep, and fuel clad chemical interaction). If materials were developed and qualified to doses greater than 200 dpa, higher burnups could be achieved without having to replace the cladding and duct. In addition, advanced structural material development may be required to make ABRs commercially viable. Improved structural material performance is one way to improve the economics of fast reactors by potentially allowing both higher operating temperatures (and thus, higher thermal efficiency and power output) and longer lifetimes (reduced replacement costs). Improved material reliability could also result in reduced down time. Superior structural materials will also spur improvements in high-temperature design methodology, allowing for more flexibility in construction and operation. MATERIALS NEEDS FOR GNEP There are many requirements for all nuclear reactor structural materials, regardless of the exact reactor design or purpose. The material must be available, affordable, and it must have good fabrication and joining properties. Good neutronics (low neutron absorption) is an important factor, especially for clad and duct applications. The material must also have suitable thermal properties and because recycle reactors will operate at elevated temperatures, 15

20% CW 316

1 cm

Unirradiated Control

Fluences Beyond FFTF Goal

Figure 1. Swelling in 20% CW 316 stainless steel irradiated to 1.5 s 1023 n/cm2 (E > 0.1 MeV).7

the materials must have good elevatedtemperature mechanical properties, including creep resistance, long-term stability, and compatibility with the reactor coolant. Finally, since the materials will be used in a high-energy and high-intensity neutron field, they must be resistant to irradiation-induced property changes (radiation hardening and embrittlement, swelling, phase instabilities, creep, and helium-induced embrittlement). This is often the most difficult requirement to satisfy. The cladding and duct materials in a fast reactor are exposed to an extremely challenging environment including neutron irradiation, liquid metal corrosion, and fuel interaction combined with significant stresses. The specific environment depends on the component location. Typical coolant temperatures in the reactor core are 400–560°C but coolant temperature at the inlet can range from 250°C to 400°C, depending on reactor design. Typical maximum neutron fluxes (E > 0.1 MeV) in a reactor are up to 4 × 1015 n/cm2s.2 This can result in significant displacement damage accumulation (20–30 dpa in iron per year). To accomplish the GNEP goals, transmutation fuels must be irradiated in the reactor for significant periods of time (5–7 years), resulting in cladding and duct doses of greater than 200 dpa. Such doses combined with other environmental factors can cause degradation of material properties. These property changes may include an increase in the ductile-to-brittle transition temperature (DBTT) and a reduction of fracture toughness from 16

low-temperature irradiation (below 400°C). This increase in DBTT is generally accompanied by an increase in tensile yield stress as well. Another issue is irradiation creep at temperatures above 500°C. This property is stress dependent. As the pressure builds up in the cladding with increasing dose, the creep rate increases. As the temperature increases, the onset of tertiary creep comes sooner. A third issue is helium embrittlement at temperatures above ~500°C. Helium levels of 10–100 appm generated during irradiation can lead to prompt failure caused by intergranular fracture. A fourth issue is void swelling in clad and duct materials. Void swelling is dependent on alloy composition, irradiation temperature, dpa rate, and increased helium generation rates. A fifth issue is fuel clad chemical interaction. The GNEP proposes to use either a metal or oxide transmutation fuel. Fission product accumulation in the fuel during irradiation can migrate to the clad/ fuel interface producing a reaction zone which degrades the clad material properties. The proposed fast reactor for GNEP is a sodium-cooled reactor.

In general, corrosion due to sodium coolant interactions is controlled by keeping oxygen levels very low (<10 ppm). Traditional ferritic/martensitic steels (e.g., Sandvik HT-9) have shown excellent corrosion resistance in sodium but advanced materials have lower chromium concentrations or oxide additions, which may affect their corrosion resistance in sodium. To meet the challenging goals of GNEP, materials must be developed to address all of these extreme operating conditions. Structural components such as the core vessels, internal core supports, and core grid for fast reactor systems have similar materials requirements and challenges. Further, structural components are expected to serve the entire functional life of the reactor, which for commercial reactors may be up to 60 years. This requirement places increased importance on thermal stability, creep strength, and fatigue properties. While the fluences expected at some components are considerably lower than those expected for clad (for example, only 5–10 dpa may be expected at the lower core internal support in current reactor designs2), irradiation damage and degradation must still be considered and evaluated carefully. Structural components have the additional requirement of passing licensing requirements. As load-bearing elements that maintain the integrity of the reactor core, structural materials face considerable scrutiny. While creep, fatigue, corrosion, and irradiation all present a challenge for nuclear materials, they also pose a significant challenge for designers. A key requirement in the design process of any power system is the use of proven, conservative design criteria. High-temperature structural design methodology uses inputs of material properties (tensile, creep, fatigue, compression, toughness, etc.),

Table I. LMFBR Clad/Duct Materials—Development Goals* Performance Parameter

Reference 316 Status

Advanced Alloy Development Goals

Swelling ($V/V) In-Reactor Creep Rate (0.45 kg/m2 at 650°C) Rupture Stress (20,000 h at 650°C)

18% at 2.5 s 1015 n/cm2 >6.2 s 10–7/h–1 36 kg/m2

5% at goal fluence <2 s 10–7 h–1 90 kg/m2

*From Reference 10.

www.tms.org/jom.html

JOM • January 2008

a

b

Figure 2. A comparison of swelling in (a) HT-9 and (b) D-9 fuel bundles irradiated in FFTF to ~75 dpa.7

2008 January • JOM

HISTORY OF U.S. FAST REACTOR MATERIALS DEVELOPMENT Significant experience exists with fast reactor technology, both domestically and internationally, starting with the Experimental Breeder Reactor (EBRI) in the 1950s. In the 1960s, both the EBR-II and Fermi reactors were started in the United States along with the Douneray Fast Reactor in the United Kingdom and the BN-600 in the former Soviet Union. In the 1970s, more liquid metal fast reactors were built including the Bor-60 (Soviet Union), Phenix (France), and JOYO (Japan) reactors. A U.S. Fast Flux Test Facility went critical in 1980, followed by SuperPhenix (France) in 1985 and Monju (Japan) in 1995. With each reactor, there were both significant materials problems and significant progress in understanding materials performance and materials development. Indeed, unexpected material behav-

700

Precipitation Strengthened Nickel-base Alloys

1,200

Rupture Life = 104 hours

600

1,000

Austenitic Stainless Steels

600 400

0

400 300

Austenitic Stainless Steels

200 Ferritic Steels

200

a

Precipitation Strengthened Nickel-base Alloys

500 800

Stress (MPa)

Yield Strength (MPa)

duty cycle (time, temperature, excursions), and other effects (environment, geometry, welds, stress-state, and heatto-heat variations) to assess possible failure modes. These failure modes may include ductility loss, creep rupture, creep-fatigue failure, gross distortion or ratcheting, loss of function through deformation, buckling, creep buckling, and non-ductile rupture. Code qualification and design methodology provide key guidance on material limits to avoid these types of failures. These design rules can be applied to specific applications during project design. High-temperature design methodology is further complicated in nuclear reactor applications as radiation damage may cause further changes in mechanical properties over the lifetime of the reactor. However, there are no established hightemperature design codes that fully incorporate irradiation effects for reactor environments. The current ASME Subsection NH3,4 provides guidance for nuclear reactor structural components, but does not fully account for irradiation effects. Ensuring that materials can meet all these requirements is a significant challenge. Finally, economics provide a key driver for further structural materials development. One of the main obstacles for the commercial deployment of advanced fast reactors is the capital cost. Three key approaches for cost reduction are being pursued in the GNEP fast reactor campaign:1,5 design simplifications, new technologies that allow reduced capital cost, and simulation techniques that help optimize system design. Improved materials will provide opportunities for both simplified design and reduced capital cost.

ior can cause major disruptions in a development program. For example, the first open literature report6 of void formation during neutron irradiation raised concerns about swelling, which dramatically slowed the development of the liquid metal fast breeder reactor (LMFBR) program. If void-induced swelling continued to increase with fluence, then large increases in the size of the core components could be expected. The extent of swelling in fuel clad is shown in Figure 1 for a 20% cold-worked (CW) 316 stainless steel irradiated in EBR-II. Clearly, swelling at this level was intolerable for fast reactor systems and improved materials were required. A decade was required to gain some understanding of the swelling process, and another decade to develop materials with satisfactory performance. Swelling is still not understood to the point of being able to predict it in an arbitrary material. But these programs ultimately did succeed in developing improved materials. Prior to 1974, the fast breeder reactors had focused on CW 316 stainless steel as the primary clad, duct, and structural material. However, testing in EBR-II showed unacceptably high swelling at the high fluences expected in clad and duct materials for both FFTF and the Clinch River Breeder Reactor Project (CRBRP). In response to the initial indications of void swelling, the Atomic Energy Commission initiated the United States National Cladding and Duct Development program (USNCDDP) in 1974. The USNCDDP

100 0

600 200 400 Temperature (°C)

0

800 b

Ferritic Steels

0

200 400 600 Temperature (°C)

800

Figure 3. The (a) yield strength and (b) stress of structural materials generated under USNCDDP (reproduced from Reference 9).

www.tms.org/jom.html

17

, ,

Irradiation Temperature (ºC) Figure 4. The 0.2% offset yield strength as a function of irradiation temperature for several ferritic-martensitic steels irradiated in FFTF.

was a major materials development effort through the early 1980s that was designed to provide advanced materials for the LMFBR program.8 The USNCDDP was administered from the Hanford Engineering Development Laboratory with significant contributions from Oak Ridge National Laboratory (ORNL) and Westinghouse. While the U.S. program was a major domestic effort, it was also part of a wider international effort, collaborating with the United Kingdom, France, and Japan. As noted by J.L. Straalsund et al.,9 the USNCDDP was primarily concerned with providing improved materials for LMFBR service. In particular, USNCDDP was originally “charged with the responsibility for development of the required technology to permit full utilization of the reference material, 20 percent cold-worked Type 316 stainless steel, in early LMFBR core applications.” Table I shows a comparison between the expected behavior of the reference 316 stainless steel and the goals set for alloy development at the beginning of the program.10 Swelling reductions of more than a factor of three were sought with similar improvements expected for creep properties. In the initial portion of the program, three classes of materials were selected for study: austenitic alloys, ferritic alloys, and precipitation-hardened Fe-Ni alloys. Thirty-five alloys were selected for initial study, including 29 commercially available alloys and six developmental alloys. In the first year of the program, a downselect process reduced the number of alloys for further study to 16. By 1980, this matrix 18

of alloys was further reduced to leave only a handful of promising alloys and second-generation candidates. A full evaluation of these three candidate alloys was performed prior to full-power operation of FFTF in early 1980 and a confirmation in early 1983 of reference material performance capabilities for initial-core CRBRP applications. An additional, comprehensive evaluation of reference material performance to commercial plant goal fluence levels was to be complete by 1985. On a more practical level, the USNCDDP focused on irradiation effects in materials. While swelling may have triggered the start of the program, irradiation creep, radiation stability, and the effects of irradiation on mechanical properties were key considerations. In addition, physical metallurgy, alloy stability, fuel-clad interactions, and sodium effects were also examined. Among the austenitic alloys, of particular interest for study were the reference 316 stainless steel alloy and D-9, a 15Cr-15Ni stainless steel stabilized with titanium. The ferritic class included the ferritic/martensitic commercial alloy Sandvik HT-9 and the developmental delta-ferrite alloy D-57. Several precipitation-hardened alloys also received consideration and study in the USNCDDP. These included the developmental alloys D21 (similar to the commercial A-286 alloy), D-66 (analog to Nimonic PE-16), and D-68 (Inconel 706). Each of these alloys received considerable study from 1975 through 1986. At the tail end of the USNCDDP, a relatively new class of materials known as oxide dispersion strengthened (ODS) steels, were selected for applicability studies.11 Straaslund points out that each alloy class was found to have notable strengths and weaknesses.9 For example, the austenitic steels had considerable strength, ductility, and fracture toughness properties. However, under irradiation these alloys exhibit far greater swelling than other alloy systems, as shown in Figure 1. The ferritic steels had greatly reduced swelling compared to the reference CW 316 stainless steel and superior irradiation creep performance, but this came at the expense of reduced high-temperature strength and low-temperature radiation www.tms.org/jom.html

embrittlement concerns. The precipitation-hardened Fe-Ni alloys exhibited superior strength and creep resistance at high temperature but suffered from high-temperature helium embrittlement and phase stability issues under irradiation. The ODS steels showed great promise, but came along too late for near-term consideration. Ultimately, after several experimental fueled subassemblies made from the candidate materials were irradiated in FFTF, the USNCDDP decided that the ferritic steel HT-9 was the nearterm candidate that could best survive to high fluences and fuel burnups in the FFTF. In what were to be the final years of FFTF operation, a large fraction of the FFTF core was replaced with subassemblies made from HT-9 clad and duct. In the United Kingdom, the Nimonic PE-16 alloy was selected for use while Japan and France continued to focus on the advanced austenitics. A variant of the D-9 alloy has been used in the Phenix and SuperPhenix reactors with success to moderate burnups (<100 dpa in the cladding). The ultimate success of these programs is best illustrated in Figure 2.7 This figure shows fuel pin bundles from FFTF after ~75 dpa (~1.5 s 1023 n/cm2). On the left, a fuel bundle with HT-9 clad shows no swelling while the right bundle with D-9 austenitic clad exhibits pin-to-pin variations caused by swelling. In addition to the considerable gains in understanding radiation damage in reactor materials, the USNCDDP also produced a wealth of information for general materials performance. Data

Irradiation Temperature (ºC) Figure 5. Uniform elongation as a function of irradiation temperature for several ferritic-martensitic steels irradiated in FFTF.

JOM • January 2008

B0 MPa–1 dpa –1

Temperature (ºC) Figure 6. Steady-state creep rate as a function of temperature for HT-9 and two austenitic stainless steels.

were generated on mechanical properties, creep performance, sodium compatibility, and phase stability in these developmental alloys. An example is shown in Figure 3, reproduced from Reference 9. This database has provided a solid foundation for subsequent programs. Structural and core materials will require further development to meet the performance and economic demands for ABR and commercial fast reactor applications. However, the USNCDDP provides a sound basis of data and development strategies. Further, the program also provides a template for efficient and successful alloy development that will have benefits to the GNEP program. Materials development has continued since the USNCDDP ended in the 1980s. Commercial alloys such as modified 9Cr-1Mo (Grade 91), NF616 (Grade 92), and HCM12A (Grade 122) were developed; also, low-activation steels were developed for fusion reactor applications.12–14 Variants of D-9 have been further developed, and as noted above, used successfully in the Phenix and SuperPhenix reactors. Additional development has produced the D-9I grade, which shows further reductions in swelling at high fluences. These alloys are currently being studied for use in India. Finally, precipitation-hardened Fe-Ni alloys have also been considered for use in fast reactors for space applications.15

duct, and structural materials program. A detailed description of properties of 9–12 chromium ferritic-martensitic steels for nuclear applications has recently been published.16 Techniques to improve performance have included alternate heat treatments and alternate alloying schemes, including the use of oxide dispersions or other precipitation strengthening. The international fast reactor programs, the fossil power generation industry, and the fusion energy materials programs have driven alloying advances for ferritic-martensitic steels. Structural Materials Studies from the Advanced Fuel Cycle Initiative Program A broad range of irradiation and testing programs on austenitic and ferriticmartensitic steels were initiated during the Advanced Fuel Cycle Initiative (AFCI) program with the intent of using the data to select suitable clad, duct, and target structural alloys for the AFCI accelerator-driven reactor concept.17 As part of this program, testing and analysis were performed on alloys that had been irradiated in the FFTF fast reactor at prototypic fast reactor temperatures and doses. The primary alloy class of interest for AFCI high dose structural components was ferritic-martensitic steels. Similar to the USNCDDP, the ferritic-martensitic steels of interest fell into two groups: commercial ferritic-martensitic steels with the greatest amount of pre-existing irradiation performance data, and developmental ferritic-martensitic steels that were designed to try to overcome the known

Table II. Evolution of Ferritic/Martensitic Steels for Power-Generation Industry25

Generation

Years

0 1

1940–1960 1960–1970

2

1970–1985

ADVANCED MATERIALS

3

1985–95

As noted, materials development continued after the end of the USNCDDP and these improved materials should be analyzed as part of the GNEP cladding,

4

Future

2008 January • JOM

deficiencies of the commercial ferriticmartensitic steels. The specimens were obtained from the Japan–USA Fusion Materials Collaboration and also from the USNCDDP. The alloys tested and analyzed included HT9, T91, and NLF steels, JFMS steel, and the W-series of steels. The latter three steels are developmental reduced activation 9% chromium ferritic-martensitic steels produced in Japan.18,19 Testing and analysis focused on the types of specimens and data that were relevant and readily available. This included tensile specimens that were used to assess strength and ductility properties at the low end of the fast reactor temperature spectrum20,21 where embrittlement of ferritic-martensitic steels is a potential issue, and it included unanalyzed pressurized tube creep data that were used to study creep properties over the entire fast reactor temperature spectrum.22,23 Several key results were obtained from the tensile studies. The first is that when these ferritic-martensitic steels are irradiated at 370°C to 430°C, the majority of the tensile property changes appeared to take place within the first 10 dpa of irradiation. This was followed by minor changes in tensile properties as the dose increased further. Next is that radiation hardening in these alloys can have a strong dependence on irradiation temperature in this temperature range. HT-9 and JFMS, which both contain relatively high amounts of silicon and carbon, undergo strong hardening as the irradiation temperature decreases from 430°C down to 370°C (Figure 4). Uniform elongation is also strongly dependent on irradia-

Steel Modification

105 h Rupture Strength (MPa)

— Addition of Mo, Nb, V to simple Cr-Mo steels Optimization of C, Nb, V Partial substitution of W for Mo and add Cu Increase W and add Co

40 60

www.tms.org/jom.html

100 140

180

Steels

Max. Use Temperature (°C)

T22, T9 EM12, HCM9M, HT9, HT91 HCM12, T91, HCM2S NF616, E911, HCM12A

520–538 565

NF12, SAVE12

650

593 620

19

60 700 °C, 170 MPa Mill Annealed

Creep Strain (%)

50 316 Austenitic Stainless Steel (SS)

40 30

800 H Fusion PCA Advanced-Modified 14Cr/16Ni SS (SA + 5%CW) CE-3 CE-0 AX-7

20 10

AX-5 and AX-8

17-14 CuMo

0 0

2,000

4,000

6,000

tion temperature, as shown in Figure 5, but the differences in post-irradiation ductility between the alloys studied was not very large despite the strong differences in irradiation-induced strengthening. Ductile failure was observed in all the tests. The hardening is attributed to the radiation-induced formation of high density small chromium-rich and silicon-rich precipitates in HT9 and silicon-rich precipitates in JFMS that block almost all dislocation motion. Higher chromium levels have been thought to play a significant role in radiation-induced hardening, but the large increase in strength of the 9% chromium JFMS steel indicates that alloy constituents other than chromium can also play a large role in the tensile properties (and possibly the fracture properties). Creep property data compiled in support of AFCI are currently limited to the well-known commercial ferriticmartensitic steels HT9 and T91. These ferritic-martensitic steels were selected for study during the USNCDDP because of their excellent resistance to swelling compared to austenitic stainless steels. More recent research on these two steels has revealed that over the majority of the fast reactor operating temperature range, these ferriticmartensitic steels have an approximate factor of two lower steady-state creep rates than austenitic steels (Figure 6).23 The most recent research has focused on comparing the differences in creep properties between HT9 and T91. The only notable performance difference occurs as the irradiation temperature increases past 550°C where T91 appears have a smaller initial creep transient and can also handle higher pressures without rapid failure. Swelling, which was evaluated as part of these creep studies, 20

8,000 Time (h)

10,000

12,000

14,000

shows that up to 2.5% can occur in HT9 and T91 irradiated at 400°C up to 225 dpa. At temperatures of 500°C–750°C, no swelling was observed in either steel for doses up to 125 dpa. Advanced Steels from the International Fast Reactor Materials Development Programs The U.S. fast reactor program (1960– 1980) concentrated on HT9, with HT9 chosen for the FFTF core demonstration program. HT9 has some limitations as a cladding and duct material. Specifically, these limitations are in high-temperature creep strength, lowtemperature fracture toughness, and possibly insufficient resistance to FCCI at higher temperatures (FCCI is a potential limitation of any ferritic-martensitic steel). To overcome these limitations in strength and toughness, many different alloys have been developed, although the performance database for most of these advanced alloys after irradiation is not as well developed as for HT9. K.Q. Bagley et al. summarized the European development in 9–12% chromium ferritic-martensitic steels to be used for wrapper (assembly duct) materials in fast reactor systems.24 In the United Kingdom, FV448 was selected as the principal choice. To provide adequate impact properties, the composition of FV448 was carefully chosen to eliminate delta-ferrite and ensure a fully martensitic structure. In France, the focus was on two steels, EM12, a 9Cr2Mo-V-Nb duplex ferritic-martensitic steel and EM10, a fully martensitic 9% chromium unstabilized steel. In Germany, investigations focused primarily on 1.4914, a fully martensitic 12CrMoV-Nb grade similar to FV448. www.tms.org/jom.html

16,000

Figure 7. Creep strain plotted versus time for various Fe-CrNi steels and alloys tested in the mill-annealed condition at 700ºC and 170 MPa. The asterisk denotes rupture.

United Kingdom optimization studies identified prior austenite grain size and tempered strength as essential to controlling impact and toughness properties. Specifically, niobium-carbide precipitates enhanced the grain-refining action (small prior austenite grain size is desirable), and vanadium and niobium enhanced the stability of M2X and M23C6 phases, leading to retention of higher pinned dislocation densities during tempering. German studies found that low nitrogen (<100 ppm) was preferred to achieve low DBTT, but that after irradiation, the beneficial effect of nitrogen is significantly diminished. These heat treatments are designed to provide low DBTT, high upper-shelf toughness, and good ductility, but at the expense of creep strength and are therefore not designed for cladding applications. Advanced Steels from the Fossil Fuel Materials Development Programs The fossil power generation industry has developed ferritic-martensitic steels with higher temperature capability than HT9, which is considered a first-generation steel (Table II). The second-generation modified 9Cr-1Mo ferritic-martensitic steel T91 has the most complete irradiation database other than HT9. T91 shows improved irradiation response properties in several areas as compared to HT9. Two of the third-generation ferritic-martensitic steels that are now ASME code-approved for applications to 620°C are NF616, developed in Japan, and E911, developed in Europe. They have essentially the same composition as the modified 9Cr-1Mo (T91), except that some of the molybdenum is replaced JOM • January 2008

by tungsten in NF616 and a tungsten addition has been made in the E911. The E911 also contains slightly higher nitrogen, and the NF616 contains a boron addition. Another third-generation steel is HCM12A, a 12Cr-0.4Mo2W-Cu-V-Nb-B steel developed in Japan. As operating temperatures are increased, the use of higher chromium steels may be required for oxidation resistance. Only one neutron irradiation experiment on the third-generation steels has been carried out and analyzed.26 In that experiment, two of the new commercial steels, NF616 and HCM12A, along with T91, were irradiated at 300ºC in the mixed neutron spectrum of the high flux reactor (HFR) in Petten, the Netherlands. For these experiments, the properties of the T-91 were superior to the new steels both before and after irradiation.26 Tensile properties indicated that the NF616 showed the least hardening followed by T91, and the HCM12A hardened the most. The ductile-to-brittle transition temperature increased in all three alloys, but the T91 developed a larger shift in ductileto-brittle transition temperature than normally observed for this steel.27 No clear explanation is available for these large increases in DBTT. The authors suggested that temperature variations may have played a role, as the temperature on the specimens was said to vary from 265°C to 312°C. The presence of boron in the two new steels might be a factor, since in the mixed neutron spectrum of HFR, the thermal neutrons can transform the boron in the steel to helium. The presence of copper in HCM12A may have played a role in its hardening. Ion irradiation experiments show similar hardening between T91 and HCM12A.28 Advanced Steels from the Fusion Materials Development Programs In the mid-1980s, programs to develop “low-activation “or “reducedactivation” ferritic steels began in the European Union (E.U.), Japan, the Soviet Union, and the United States. The objective in using such steels is that the more rapid induced-radioactivity decay will allow the steels to be disposed of after service by shallow land burial 2008 January • JOM

rather than the more expensive deep geological storage. For reduced activation, the alloying elements Nb, Ni, Cu, and N needed to be eliminated. Fusion materials programs in Japan, the European Union, and the United States have developed Cr-V and Cr-W-V steels,29–37 to which tantalum is sometimes added as a replacement for niobium. Steels with 7–9% chromium are favored because of the difficulty of eliminating D-ferrite in a 12% chromium steel without the availability of nickel for austenite stabilization; austenite stabilization would then require increasing carbon or manganese to undesirable levels. In the United States, an Fe-9.0Cr2.0W-0.25V-0.07Ta-0.1C (9Cr-2W-VTa) reduced-activation steel was developed at ORNL35 and subsequently showed the best unirradiated and irradiated properties of reduced-activation steels developed elsewhere in the United States, Europe, and Japan.38 In addition, the 9Cr-2W-V-Ta steel had excellent creep-rupture properties. Nanoscale Precipitation The second method for obtaining improved high-temperature strength is through the addition of fine nanometer-sized particles. One example is ODS steels, which have been the focus of several reactor structural material programs. The earliest research on these materials for fast reactor clad and duct applications took place at PNNL toward the end of the USNCDDP.11 Oxide dispersion-strengthened steels have very recently been developed as cladding material in Japan by the Japan Atomic Energy Agency (JAEA).39 These steels derive their elevated-temperature strength from a fine distribution of Y2O3 or TiO2 particles mechanically driven into the alloy, as opposed to carbide precipitation in conventional elevated-temperature steels. This development extends the operating envelope of this material to beyond 700°C, making it especially attractive for use as fuel cladding. The expected radiation resistance features of ODS steels come from the Y2O3 oxide particles (or nanoclusters—atom probe examination indicates the oxide particles dissolve during fabrication and reform as small atom clusters or nanoclusters) in the matrix. In traditional ferritic-martenswww.tms.org/jom.html

itic steels, the precipitates that contribute to the strength of traditional steel will coarsen at higher temperatures. The oxide particles and nanoclusters in ODS steels are stable to very high operating temperatures. These precipitates and nanoclusters block mobile dislocations to improve the high-temperature strength and act as a sink for radiationproduced point defects. Studies on radiation stability of these clusters generally show stability of the oxides, but a limited number of high-temperature, high-dose studies show the particles do decrease in size during irradiation.40 Pulsed magnetic welding was successfully used to weld thin-walled tubing in the ODS steel study for the USNCDDP.11 At the Tokai works of the JAEA, joining technology between the ODS steel cladding and the end plugs has been developed for application in fast reactor fuel pin fabrication. A pressurized resistance welding method is utilized to perform solid-state joining. Higher-temperature operation could also be gained by shifting to nickelbased alloys that have higher strength than ferritic-martensitic steels. Unfortunately, nickel-based materials have traditionally suffered from embrittlement due to extensive gamma prime formation, and also from the formation of bubbles on grain boundaries stabilized by transmuted helium, leaving the material in a weakened state.41 As noted previously, engineering microstructures is one method of mitigating the effects of irradiation damage. Another class of engineered alloy may be appropriate for service in advanced fast reactor systems: nano-precipitation-strengthened steels (also called high-temperature ultra-fine precipitation steels or HT-UPS). This class of materials was developed in the late 1980s, primarily for the U.S. Fusion Materials Program.42,43 These steels have a base chemistry similar to the D-9 alloys discussed previously (15Cr16Ni-2Mn-2Mo), but with other minor alloying changes to help control microstructure. In particular, the carbon levels are higher while silicon levels are the same or slightly lower than typical austenitic steels to help promote the formation of small carbides over intermetallic phases. The HT-UPS steels also have slightly higher titanium and 21

phosphorus contents to produce both carbides and needle-shaped phosphides (Fe-Ti-P). The resulting microstructure is extremely stable and also promotes increased strength and creep resistance. This is illustrated in Figure 7 where creep strain is plotted versus time for various Fe-Cr-Ni steels at 700ºC and 170 MPa. The HT-UPS steels (AX and CE series) show a remarkable improvement over traditional steels at elevated temperatures. The radiation data on these steels is promising, but is limited due to a change in programmatic emphasis in the early 1990s. Further, since only standard alloy production techniques are needed for HT-UPS steels, these materials can be produced in larger quantities and reduced costs over ODS steels. While additional data and development are needed for the HTUPS steels, the improved performance makes them a viable candidate material for advanced fast reactor applications. One other possible method for reducing the likelihood of helium embrittlement is to make the grain boundaries less favorable sites for helium accumulation by engineering the specific grain boundary orientations. Grain boundary engineering (GBE), a technique that utilizes a combination of controlled deformation and annealing steps, has been demonstrated as an effective and economic approach for property improvement of polycrystalline materials.44 Preliminary studies on candidate materials for Generation IV systems indicate that the high-temperature creep-resistance of ferritic-martensitic steel T91 and ductility of austenitic alloy INCOLOY 800H under irradiation were improved by GBE.45,46 However, the thermal stability of GBE structures has to be evaluated in greater detail. GNEP IRRADIATION AND TESTING PROGRAM To develop and qualify materials to a total fluence greater than 200 dpa requires development of advanced alloys and long-term irradiations in fast reactors to test these alloys. Important data required on these advanced alloys include but are not limited to tensile fracture toughness and creep data after irradiation to doses greater than 200 dpa at irradiation temperatures of 350–600°C. The GNEP materials plan to obtain 22

high dose fast reactor irradiation data in the near term involves testing specimens of traditional ferritic/martensitic alloys (HT-9 and T-91) previously irradiated in the FFTF reactor to doses up to 225 dpa at irradiation temperatures from 365–700°C. This includes analysis of an HT-9 duct after irradiation to a total dose of 155 dpa at temperatures from 410°C to 470°C with lower-dose material covering irradiation temperatures from 365°C to 510°C. Such previously irradiated material will provide essential data for traditional ferritic-martensitic steels, but new irradiation data are needed to qualify new advanced alloys. These data will be obtained through international collaborations in fast reactors. The available fast reactors for such irradiations are limited to: v Phenix fast reactor (CEA, France): Irradiations in this reactor can achieve up to 30 dpa per year, but this reactor is being shut down in the beginning of 2009. v BOR-60 LWR (RIAR, Dmitrovgrad, Russia): Irradiations in this reactor can achieve 20–30 dpa per year. This reactor will be operated until at least 2015. v BN-600 fast reactor (Russia): Irradiations in this reactor can achieve up to 30 dpa per year. Since this reactor produces power it is more difficult to use for future materials irradiations. v JOYO (JAEA, Japan) fast reactor: Irradiations in this reactor can achieve up to 20 dpa per year if a high dose location can be obtained. This may be difficult since researchers in Japan have numerous irradiations underway in an effort to test advanced cladding materials and fuels. Discussions are underway for future fuels and materials irradiations for the GNEP program at the JOYO reactor. Fast reactors are being built in China and India but no collaborative agreements are in place for performing irradiations in these reactors. The GNEP materials irradiation plan presently includes two irradiations in the Phenix reactor containing traditional and advanced ferritic/martensitic steels. The irradiations are called MATRIX-SMI www.tms.org/jom.html

and II. This irradiation program is a collaboration between the U.S. Department of Energy (GNEP and Gen IV programs), Eurotrans (DEMETRA), and the European Gen IV program. The Department of Energy (GNEP and Gen IV programs) is contributing specimens of the traditional ferritic/ martensitic alloys (HT9 and T91) and advanced alloys such as EP-823, HCM12A (T122), 3Cr-W-V-Ta, 3Cr-W-V, NF616 (9Cr alloy), 14WT, 12YWT, 14YWT, and Inconel 800H. The total doses obtained in this irradiation will be 9 dpa, 17 dpa, 35 dpa, and 65 dpa at irradiation temperatures from 380°C to 510°C while in direct contact with the sodium. Specimens include tensile, fracture toughness, and transmission electron microscopy specimens. The irradiation will end in January 2009 and specimens will be available for testing by 2010. These data will provide a comparison of irradiated properties of HT9, T91, and newly developed advanced ODS alloys. In addition, collaborative agreements are in progress with RIAR in Dmitrovgrad, Russia, and with JAEA in Oarai, Japan, to perform irradiations in BOR-60 and JOYO, respectively. It is desired to obtain higher-dose irradiation data from these collaborations. References 1. Global Nuclear Energy Partnership Technology Development Plan, Report number: GNEP-TECH-TRPP-2007-00020, Rev 0 (Idaho Falls, ID: Global Nuclear Energy Partnership Technical Integration Office, Idaho National Laboratory, 25 July 2007), p. 177. 2. B. Hill, Argonne National Laboratory, private communication, 2007. 3. R.L. Huddleston and R.W. Swindeman, Materials and Design Bases Issues in ASME Code Case N-47, NUREG/CR-5955, ORNL/TM-12266 (April 1993). 4. Prototype Large Breeder Reactor Phase II Conceptual Design; Vol III Tradeoff Studies, General Electric Report, GEFR00099 (June 1977). 5. E. Hoffman, Estimated Cost for Low Conversion Ratio Burners, Argonne National Laboratory Report, ANL-AFCI-118 (June 2004), p. 15 6. C. Cawthorne and E.J. Fulton, “Voids in Irradiated Stainless Steels,” Nature, 216 (November 1967). 7. F.A. Garner, “Irradiation Performance of Cladding and Structural Steels in Liquid Metal Reactors,” Nuclear Materials, ed. B.R.T. Frost (Weinheim, Germany: VCH Veerlagsgesellschaft mbH, 1996), pp. 420–543. 8. J.J. Laidler and J.W. Bennett, “Core Material Studies Improve Fast Breeder Performance,” Nucl. Eng. Int., 25 (301) (July 1980), pp. 31–36. 9. J.L. Strassland, R.W. Powell, and B.A. Chin, “An Overview of Neutron Irradiation Effects in LMFBR Materials,” J. of Nuclear Materials, 108-109 (1982), p. 299. 10. J.W. Bennett and K.E. Horton, “Material Requirements for Liquid Metal Fast Breeder Reactors,” Met. Trans. A., 9A (1978), p. 143.

JOM • January 2008

11. M.L. Hamilton et al., “Fabrication Technological Development of the Oxide Dispersion Strengthened Alloy MA957 for Fast Reactor Applications,” PNNL-13168 (2000). 12. F. Garner, “Insights on Radiation-Induced Dimensional Instability and Mechanical Properties of Austenitic Alloys Acquired after Closure of the US LMR Program” (Presentation at the GNEP Advanced Materials Workshop, ORNL, July 2007). 13. E.E. Bloom, S.J. Zinkle, and F.W. Wiffen, “Materials to Deliver the Promise of Fusion Power-Progress and Challenges,” J. of Nuclear Materials, 329-333 (2004), p. 12. 14. S.J. Zinkle, “Advanced Materials for Fusion Technology,” Fusion Engg. and Design, 74 (2005), p. 31. 15. T.M. Angeliu, J.T. Ward, and J.K. Witter, “Assessing the Effects of Radiation Damage on Ni-Base Alloys for the Prometheus Space Reactor System,” J. of Nuclear Materials, 366 (2-3) (2007), p. 223. 16. R.L. Klueh and D.R. Harries, High-Chromium Ferritic and Martensitic Steels for Nuclear Applications (West Conshohocken, PA: ASTM, 2001). 17. S.A. Maloy, “Materials Issues in a High Power Spallation Target,” J. of Nuclear Materials, 343 (2005), p. 367. 18. H. Kurishita et al., “Tensile Properties of Reduced Activation Fe-9Cr-2W Steels after FFTF Irradiation,” J. of Nuclear Materials, 212-215 (1994), pp. 730–735. 19. Y. Katoh et al., “Microstructural Changes in Fe10Cr-2Mo Steel by Neutron or Charged Particle Irradiation,” J. of Nuclear Materials, 191-194 (1992), pp. 1204–1208. 20. S.A. Maloy et al., “Tensile Properties of the NLF Reduced Activation Ferritic/Martensitic Steels after Irradiation in a Fast Reactor Spectrum to a Maximum Dose of 67 dpa,” J. of Nuclear Materials, 341 (2005), pp. 141–147. 21. S.A. Maloy et al., “The Effects of Fast Reactor Irradiation Conditions on the Tensile Properties of Two Ferritic/Martensitic Steels,” J. of Nuclear Materials, 356 (2006), pp. 62–69. 22. M.B. Toloczko et al., “Comparison of Thermal Creep and Irradiation Creep of HT9 Pressurized Tubes

2008 January • JOM

at Test Temperatures from ~490°C to 605°C,” Effects of Radiation on Materials: 20th International Symposium, ASTM STP 1405 (2002), pp. 557–569. 23. M.B. Toloczko and F.A. Garner, Stress and Temperature Dependence of Irradiation Creep of Selected FCC and BCC Steels at Low Swelling, ASTM STP 1447 (2004), pp. 454–467. 24. K.Q. Bagley et al., “European Development of Ferritic-Martensitic Steels for Fast Reactor Wrapper Applications,” Nuclear Energy, 27 (5) (1988), pp. 295–303. 25. F. Masuyama, Advanced Heat Resistant Steel for Power Generation, ed. R. Viswanathan and J. Nutting (London: Institute of Materials, 1999), p. 33. 26. M. Horsten et al., Effects of Irradiation on Materials: 19th International Symposium, Proceedings, ASTM STP 1366, ed. M.L. Hamilton et al. (West Conshohocken, PA: ASTM, 2000), p. 579. 27. R.L. Klueh and D.J. Alexander, J. of Nuclear Materials, 187 (1992), p. 60. 28. T.R. Allen et al., “Radiation Resistance of Advanced Ferritic-Martensitic Steel HCM12A,” J. of ASTM International, 2 (8) (September 2005), Paper ID JAI12382. 29. K. Anderko et al., CETA-EinEntwicklungsschritt zu einem Schwach Activierbaren Martensitischen Chromstahl, Kernforschungszentrum Karlsruhe, KfK Reprot 5060 (June 1993). 30. D. Dulieu, K.W. Tupholme, and G.J. Butterworth, J. of Nuclear Materials, 141-143 (1986), p. 1097. 31. D.S. Gelles, Reduced Activation Materials for Fusion Reactors, ASTM STP 1047, ed. R.L. Klueh et al. (Philadelphia, PA: ASTM, 1990), p. 113. 32. N.M. Ghoniem, A. Shabaik, and M.Z. Youssef, Proceedings of Topical Conference on Ferritic Steels for use in Nuclear Energy Technologies (Warrendale, PA: TMS, 1984), p. 201. 33. C.Y. Hsu and T.A. Lechtenberg, J. of Nuclear Materials, 141-143 (1986), p. 1107. 34. H. Kayano et al., J. of Nuclear Materials, 179-181 (1991), p. 445. 35. R.L. Klueh and E.E. Bloom, Nucl. Eng. Design/Fusion, 2 (1985), p. 383. 36. T. Noda et al., J. of Nuclear Materials, 141-143 (1986), p. 1102.

www.tms.org/jom.html

37. M. Tamura et al., J. of Nuclear Materials, 141-143 (1986), p. 1067. 38. M. Rieth, B. Dafferner, and H.D. Röhrig, J. of Nuclear Materials, 233-237 (1996), p. 351. 39. S. Ohtuska et al., Materials Transactions, 46 (3) (2005), p. 1. 40. T.R. Allen et al., “Radiation Response of a 9 Cr Oxide Dispersion Strengthened ODS to Heavy Ion Irradiation,” submitted to J. of Nuclear Materials. 41. R.S. Barnes, “Embrittlement of Stainless Steels and Nickel-Based Alloys at High Temperature Induced by Neutron Radiation,” Nature, 206 (1963), p. 1307. 42. P.J. Maziasz, Microstructural Stability and Control for Improved Irradiation Resistance and for HighTemperature Strength of Austenitic Stainless Steels, ASTM STP 979, ed. B.L. Bramfitt et al. (Philadelphia, PA: ASTM, 1986), p. 116. 43. P.J. Maziasz, “Developing Austenitic Stainless Steel for Improved Performance in Advanced Fossil Power Facilities,” J. of Metals, 41 (7) (1989), p. 14. 44. T. Watanabe and S. Tsurekawa, “The Control of Brittleness and Development of Desirable Mechanical Properties in Polycrystalline Systems by Grain Boundary Engineering,” Acta Mater., 47 (1999), pp. 4171–4185. 45. G. Gupta and G.S. Was, “Interpretation of Improved Creep Properties of a 9Cr-1Mo-Nb-V (T91) Steel by Grain Boundary Engineering,” TMS Letters, 2 (2005), pp. 71–72. 46. L. Tan et al., “Microstructure Tailoring for Property Improvements by Grain Boundary Engineering,” submitted to J. of Nuclear Materials. T.R. Allen is with the University of Wisconsin–Madison, Engineering Physics Department, 1500 Engineering Drive, Madison, WI 53706; J.T. Busby and R.L. Klueh are with Oak Ridge National Laboratory; S.A. Maloy is with Los Alamos National Laboratory; and M.B. Toloczko is with Pacific Northwest National Laboratory. Dr. Allen can be reached at (608) 265-4063; fax (608) 263-7451; e-mail allen@ engr.wisc.edu.

23