An essential prerequisite for maximizing the ‘irra­diation creep resistance’ is to ensure38 the best combination of thermal creep behavior and long­term microstructural stability at high temperature. Hence, the present section would discuss irradiation creep in the same sequence as mentioned above.

The design principles of development of creep — resistant steels are as follows:

• Introduce high dislocation density by either trans­formations or cold work to increase the strength of the basic lattice;

• Strengthen the host lattice by either solid solution strengtheners or defects;

• Stabilize the boundaries created by phase transfor­mations by precipitating carbides along the boundaries;

• Arrest dislocation glide and climb by appropriate selection of crystal structure, solid solution, inter­faces, dislocation interactions, and crystal with low diffusivity;

• Resist sliding of grain boundaries by introducing special type of boundaries and anchoring the boundaries with precipitates;

• Ensure long-term stability of the microstructure, especially against recovery and coarsening of the fine second phase particles;

In the case of 9-12 Cr steels, the martensitic lath structure (Figure 7) decorated with only MX which should39 be stable over long-term service life is the most desired structure. Thermo-Calc evalua­tions show39 that MX can be stabilized at the expense of M23C6 only by reducing carbon to as low a value as 0.02% in 9 Cr-1Mo steel. This value is too low to ensure acceptable high temperature mechani­cal behavior of the steels. In the context of fast reactor core components, the high chromium 9-12% ferritic-martensitic steels assume relevance. Hence, an extensive database25 for a large number of commercial ferritic steels has been generated and

Lath boundary

(a) (b)

Figure 7 The schematic39 of most undesirable (a) and desirable (b) microstructures for design of creep-resistant steels.

Figure 8(a) shows40 the continuous improvement achieved by careful modification of alloying elements, in the thermal creep behavior of successive grades of different commercial ferritic steels.

While understanding thermal creep is essential to narrow down the choice of ferritic steels for use in a fast reactor, ‘benchmarking’ the steels developed under irradiation is an essential stage before actually using the radiation-resistant steels in the reactor. The irradiation creep depends on the stress level, the temperature, and the dose. Figure 8(b) shows41 the comparison of irradiation creep of ferritic steels with competing materials like the austenitics and nickel-based alloys.

It is clear that the point defects generated during irradiation act against the design principles of devel­oping creep-resistant materials, listed earlier. The point defects accelerate the kinetics of dislocation climb, coarsen the precipitates, and generally enhance the diffusivity. In addition, the excess point defects precipitate into either interstitial or vacancy loops, but not randomly. The interaction between point defects and stress leads to the precipitation of intersti­tial loops parallel to the applied stress, while vacancy

(a)

(b)

Figure 9 The mechanisms of stress-induced preferential absorption (a) and stress-induced preferential nucleation (b) during irradiation creep.

loops form in planes perpendicular to the stress. This process (Figure 9(a)) called the stress-induced pref­erential nucleation (SIPN) results in additional creep strain solely due to irradiation. The excess point defects under temperature migrate randomly. But in the presence of an additional factor, that is, stress, the vacancies migrate preferentially to grain boundaries perpendicular to the applied stress, while the intersti­tials toward boundaries parallel to the stress. This is equivalent to removing material from planes parallel to the stress to those which are perpendicular to the applied stress, introducing additional creep strain. This process is called the stress-induced preferential absorption (SIPA) (Figure 9(b)).

The radiation-induced defects also evolve from isolated point defect to loops and voids, which have different types of influence on irradiation creep. Most often, irradiation creep occurs19,42 simultaneously with swelling and sometimes, swelling influences irradia­tion creep. At very small dose levels, swelling enhances creep rates. Beyond a certain dose levels, the creep component reduces and at high dose levels, creep dis­appears, while swelling continues. Figure 10 shows the variation in creep coefficient at various dose levels, and the regimes where swelling has an influence. The dynamics of point defects during irradiation continu­ously evolve with change in structure of dislocation network and loops. At small dose levels, there is a uniform distribution of very fine voids, which act as effective pinning centers for mobile dislocations. Thus the creep rate increases. With increase in dose levels, voids grow and multiply. The chance of interstitials and

Figure 10 Schematic of variation of instantaneous creep coefficient with dose, showing the interplay between irradiation creep and void swelling.

vacancies impinging on the void surface becomes more than their reaching dislocations. The number of inter­stitials reaching a dislocation reduces. Additionally, the defect clusters, that is, the dislocation loops also undergo ‘faulting’ contributing to the density of dis­locations in the matrix. Hence, creep rate reduces, due to two factors: increased dislocation density of the matrix due to unfaulting of dislocation loops and reduced availability of interstitials to dislocations. The above process continues until complete cessation of creep, with swelling continue to take place.

At very high temperatures, the point defect migra­tion along the grain boundaries in preferential routes causes the grain boundary aided creep.

This high temperature limit of ferritic-martensitic steels restricts the application of these steels to at best, wrappers of present generation fast reactors based on oxide fuel. It is necessary to develop materi­als with better high temperature irradiation creep properties and void swelling for clad applications. The future scenario, which envisages the development of metallic fuel to ensure sustainability by breeding, could make use of ferritic steels for both clad and wrapper. This advantage arises due to the lower value ofthe anticipated clad temperatures with metallic fuels (see Chapter 3.14, Uranium Intermetallic Fuels (U-Al, U-Si, U-Mo)), whose choice is mainly to ensure sustainability using high breeding ratio.

4.03.4.2 Irradiation Embrittlement in Ferritic Steels

The stabilized ferritic steels in the normalized and tempered condition have a tempered martensitic structure with a preponderance of monocarbides that impart the necessary creep strength, while the prior austenite grain and lath boundaries are deco­rated with Cr rich M23C6 precipitates which increase the thermal stability of the steel. It is reported that thermal aging at temperatures above 773 K causes gradual but continuous degradation in upper shelf properties in addition to increase in the DBTT. The nature of embrittlement varies for different compo­nents of the reactor. For removable components such as clad, which are subjected to high temperature and pressure, with a residence time of a few years, creep embrittlement is the issue which decides their design and performance, while for permanent sup­port structures increase in hardening and loss in fracture toughness on irradiation are major issues.

The origin of embrittlement is two-fold: segrega­tion of tramp elements to prior austenite grain boundaries which make the grain boundaries deco­hesive and evolution of carbides and intermetallic phases. The latter causes progressive changes in the tempered martensitic microstructure, which deterio­rate the fracture properties of the steel, by introdu­cing irradiation hardening effects.

The increase in the ductile to brittle transition temperature, ADBTT, is known to be related to irradiation hardening, which is generally observed to saturate with fluence. Evidence for a possible maxi­mum in DBTT was observed for the 12Cr steel irra­diated in the range of 35-100 dpa in fast flux test facility (FFTF). Based on observed data in a number of cases it appears that a high fluence and/or high tempera­ture are required before a maximum is observed. This implies that the strength and impact properties are a balance between the point defect production and irradiation-induced precipitation. The precipitation during irradiation hardens the steel and irradiation accelerated recovery and aging soften the steel. The latter process is more important at high fluences and/or higher irradiation temperatures. Hence, hardening in most of these Cr-Mo steels is more than compensated for by the recovery and aging processes, leading to saturation in irradiation hardening above 723 K.

For body centered cubic materials such as ferritic martensitic steels, radiation hardening at low tempera­tures (<0.3 TM) can lead to a large increase in the DBTT and lowering of impact energy for radiation dose as low as 1 dpa (displacement per atom). The minimum operating temperature to avoid embrittle­ment in ferritic martensitic (F/M) steels is ^473- 523 K, while the upper limit is controlled by four different mechanisms: thermal creep, high temperature helium embrittlement, void swelling, and compatibility
with fuel and coolant. Void swelling is not expected to be significant in F/M steels up to damage levels of about 200 dpa.

Extensive evaluation14’15’43-58 of the embrittle­ment behavior of the ferritic steels for different chemistry is shown in Figure 11. The merit in focus­ing on chemistry around 9% chromium is very clear based on the observation of minimum shift in DBTT around this composition, under irradiation. However, higher chromium improves corrosion resistance and ease of reprocessing. Hence, chromium content has to be selected balancing these requirements. It is known44-48 that addition of phosphorous, copper, vanadium, aluminum, and silicon would increase the DBTT while sulfur reduces the upper shelf energy (USE). The 12Cr steels, HT9, show a larger shift (125 K) in DBTT as compared to modified 9Cr-1Mo steel (~54K). Hence, the balance is always between nearly nil swelling resistant 12Cr steels and 9Cr steel which is less prone to embrittlement than 12Cr steels.

Microstructural parameters, like the prior aus­tenite grain size, lath/packet size, carbides, and their distribution influence49,50 the embrittlement behav­ior. Studies on the effects of heat treatment and microstructure on the irradiation embrittlement in 9Cr-1MoVNb and HT9 steels are summarized below:

• Prior austenite grain size (PAGS) influences51 the

DBTT for the 9Cr-1MoVNb steel, but not in

12Cr-MoVW steel. This is attributed to the precipitates in the microstructure controlling the fracture behavior rather than the PAGS, in the 12Cr steel.

• The size of martensitic lath and packet, which is sensitive52 to austenitization temperature, can also affect51 the fracture behavior. Examination of the fracture surface revealed cleavage and regions of ductile tearing along prior austenite grain and lath packet boundaries. Subsurface microcracks and sec­ondary surface cracks were found associated with large boundary carbides. It was suggested that cleav­age fracture initiated in HT9 by propagation of a microcrack from a coarse carbide into the matrix. Propagation was inhibited by the intercepted boundaries, lath or grain and ductile tearing was required53 to continue propagation. The amount of tearing increased with increasing austenitization temperature.

• Tempering for the two normalization tempera­tures had very small effect on the DBTT, for the two steels.

• Irradiation of the two steels at 638 and 693 K resulted37 in an increase in DBTT and a decrease in USE for all conditions with the shift in DBTT for the 12 Cr steel being almost twice that for 9Cr steel.

•

Although the 12Cr steel with the smallest grain size had55 the lowest DBTT after 20 dpa, the effect of tempering was different. In the case

of 12Cr steel, the higher tempering temperature causes coarsening of precipitates thus accelerating fracture.

• The saturation of shift in DBTT with fluence is independent54 of tempering conditions for the 9Cr steel, while for the 12Cr steel, a maximum is observed, probably due to faster growth of preci­pitates during irradiation.

The generation of helium through (n, a) reaction in elements of structural materials is known to cause severe damage to the embrittlement behavior of core component materials. Table 6 lists the shift in DBTT, for 9 and 12CrMo steels, under reactor irra­diation, with and without helium, which demon — strates56 the harmful effect of helium. These results become more pertinent in the case of fusion reactors, where the operating conditions include the genera­tion of helium up to about ^100appmyear~

The increase in the DBTT due to irradiation is a cause of serious concern for use of ferritic steels, since it makes the postirradiation operations very difficult. Several methods have been attempted57,58 to address this problem, which includes modification of the steel through alloying additions, control of tramp elements by using pure raw materials and improved melting practices, and grain boundary engineering (GBE). However, the propensity of the problem is less if the clad thickness is low, which normally is the case to ensure best heat transfer properties. For low thickness components, the triaxial stress necessary for the embrittlement does not develop, which reduces the intensity of this otherwise serious problem of embrittlement in ferritic steels.

An approach to reduce shift in DBTT is an immediate concern in ferritic steels for core com­ponent applications and efforts to overcome this problem by selection of high purity metals, adoption of double or triple vacuum melting for steel making, strict control of tramp and volatile elements, and development of special processing methods, which would improve the nature of grain boundaries (GBE) are in progress.