Skip to main content
EU Science Hub

1.5.1 Irradiation Embrittlement Parameters

Typical parameters varied in studies on irradiation embrittlement are especially fluence, but also chemical composition and irradiation temperature, and – to a lesser extent – flux and microstructure. Chemical composition covers the influence of certain alloying elements (intentionally inside the material) as well as the influence of accompanying elements originating from the manufacturing process (“impurities”, unintentionally inside the material).

Fluence is the parameter of utmost importance – especially in the context of long term operation which is in discussion or already established at many plants in the EU and worldwide. While the influence of neutron irradiation of the reactor pressue vessel (RPV) materials during design lifetime (typically 30 (WWER-440) or 40 (most others) years) is normally covered by irradiation / surveillance programmes, these programmes do not necessarily deliver information for long term operation beyond design lifetime. Thus, a huge amount of research has been performed to study this influence.

1.5.2 Fluence

A set of model alloys representing pressurized water reactor (PWR) RPV steel with systematic variation of the chemical composition was manufactured to study fluence and flux effects together with the influence of certain alloying elements [JRC24564]. For this purpose, the specimens were irradiated at 270 – 275 °C in the LYRA irradiation facility (JRC), at Rovno-1 NPP (WWER-1000, Ukraine), and at Kola-3 NPP (WWER-440, Russia) [JRC30550]. The ductile-to-brittle transition temperature (DBTT) was investigated as a function of fluence. Mechanical properties were tested using the Charpy impact technique. A 1/3 power dependence of the DBTT with fluence can be observed which is in accordance with expectations of the Russian Regulatory Guide and Russian Code, respectively [PNA86] (cited in [JRC30550]). The same fluence dependence is reported by Ahlstrand et al. for WWER-1000 weld material [JRC26623]. The US code, however, assumes a power dependence on fluence with an exponent less than 0.28 [NUR88]. The discrepancy may be explained by the fact that the Russian code takes into account neutrons with a kinetic energy higher than 0.5 MeV whereas the US code uses a threshold of 1 MeV [JRC24564]. Investigations in the framework of the MADAM project dealt extensively with different scales for the neutron fluence. It was found that the correlation factor between the threshold of 0.5 MeV and 1 MeV varies with local position of the irradiated specimen relative to the core and the core structure / core loading itself [Debarberis1998].

The dependence of DBTT shift on neutron fluence is comparable for PWR, WWER-440, and BWR weld materials [JRC30648]. Kryukov et al. recently performed investigations of WWER-440 RPV steel to a neutron fluence far beyond typical end-of-life doses. They found no change in embrittlement mechanism meaning that the used formula and models (see below) are still valid [JRC66610].

In the context of the PISA project irradiation data coming from irradiation campaigns in test / research reactors and surveillance channels in commercial reactors was analysed with regard to the possible occurrence of non-hardening embrittlement. For typical western PWR steels no evidence was found that non-hardening embrittlement occurs for neutron doses up to 1020 n/cm2 (E > 1 MeV). For WWER-1000 base metal, however, results are inconsistent and non-hardening embrittlement cannot be ruled out [JRC46587] [PISA2005] [English2003].

A simplified semi-mechanistic irradiation embrittlement model was set up considering the influence of neutron fluence, neutron flux, and several more effects like the chemical composition [JRC28685] [JRC30340] [JRC30552]. The model, however, needs further improvement when considering high-Ni containing steels [JRC33283].

Verheyen et al. performed irradiation experiments (300 °C, up to 0.2 dpa) of different Fe alloys with various amounts of copper. The material hardens with a square-root dependence on increasing neutron dose in the investigated fluence range [Verheyen2006].

Irradiation of commercial JRQ steel with various neutron doses revealed a behaviour consistent with expectations as described in literature. In addition, it was found that the initial microstructure (i.e. the manufacturing history) and pre-strain have a significant influence on the irradiation embrittlement. For Magnox steel, however, fluence dependence was found to be lower than expected in the investigated fluence range [JRC46587] [PISA2005]. In the FRAME project, JRQ steel was irradiated with various neutron fluences. The derived dependence of the DBTT on neutron fluence is consistent with predictions in national codes and standards. The flux was also investigated [Valo2007a].

WWER-440 cladding materials (ASS) were irradiated up to a dose of 1020 cm–2 (E > 1 MeV) at 20 °C and 300 °C. With increasing neutron fluence, tensile properties increase up to 33 % compared to the non-irradiated specimens. A reduction of the elongation is also observed [JRC34527]. Irradiation of cladding materials from French plants, however, do not show important differences in tensile properties up to a fluence of about 6x1019 n/cm2 at 290 °C [EUR23207].

Compiling irradiation data of RPV cladding specimens revealed that the fluence effect is less pronounced compared to the temperature effect. In this context, the delta-ferrite content plays a role in a way that embrittlement is slightly reduced with increasing delta-ferrite content [Keim2012].

Within the PRIS project, austenitic stainless steels (AISI304(L), AISI316(L)) were irradiated up to a very high fluence (65 dpa) which is far beyond the typical end-of-life fluence occurring at the RPV cladding. All materials show a clear increase in yield strength (up to a factor of 4) and ultimate tensile strength (up to a factor of 2) while elongation is simultaneously reduced (up to a factor of 3) [PRIS2004]. The portability of these results onto RPV cladding must be handled carefully because firstly the materials used for RPV cladding differ to some extent from the investigated materials in the study and secondly within the low-fluence regime effects may occur which are not captured by the high-fluence measurements presented in this project.

Measurements performed during the LONGLIFE project revealed that a saturation of the irradiation embrittlement effect does not occur within the fluence range relevant for long term operation. The majority of test results show that embrittlement does not accelerate ("late blooming effect") beyond a specific neutron fluence threshold except some laboratory steels at lower irradiation temperature [Brumovsky2014b].

1.5.3 Flux

A set of model alloys representing PWR RPV steel with systematic variation of the chemical composition was manufactured to study fluence and flux effects together with the influence of certain alloying elements [JRC24564]. For this purpose, the specimens were irradiated at 270 – 275 °C in the LYRA irradiation facility (JRC), at Rovno-1 NPP (WWER-1000, Ukraine), and at Kola-3 NPP (WWER-440, Russia) [JRC30550]. Flux effects with a reduced fluence-dependent DBTT shift were observed only for high-Ni, high-P, high-Cu containing model alloys.

The flux, however, is in general an uncertainty factor whose quantitative influence still remains unclear [JRC63603]. In experimental studies it was found that neutron flux is only of influence when phosphorous is present in relatively high concentrations [JRC30550]. Investigations of WWER-440 RPV weld materials irradiated with different flux by Debarberis et al. revealed that the flux is of importance in the low to intermediate copper regime and up to moderate phosphorous content. In these regions the DBTT shift was observed to be smaller when the flux is higher [JRC30340]. Recent investigations by Kryukov et al. revealed that a rather low flux increases the DBTT shift in WWER-440 RPV steels when copper is present in concentrations beyond 0.13 wt.-%. After annealing, this effect is reduced [JRC86552].

A simplified semi-mechanistic irradiation embrittlement model was set up considering the influence of neutron fluence, neutron flux, and several more effects like the chemical composition [JRC28685] [JRC30340] [JRC30552]. The model, however, needs further improvement when considering high-Ni containing steels [JRC33283].

In the FRAME project, JRQ steel was irradiated with various neutron fluences. The derived dependence of the DBTT on neutron fluence is consistent with predictions in national codes and standards. The flux was also investigated but was shown to have no influence [Valo2007a].

Investigations performed during the LONGLIFE project confirmed the absence of a flux effect for Western RPV steels even at high fluence. For WWER steels, however, flux effects can be observed when the copper content exceeds 0.13 % [Brumovsky2014b].

1.5.4 Chemical Composition

A set of model alloys representing PWR RPV steel with systematic variation of the chemical composition was manufactured to study the influence of certain alloying elements e.g. nickel (Ni) as well as other chemical elements usually contained in steel as accompanying elements e.g. phosphorous (P) and copper (Cu) together with flux effects [JRC24564]. For this purpose, the specimens were irradiated at 270 – 275 °C in the LYRA irradiation facility (JRC), at Rovno-1 NPP (WWER-1000, Ukraine), and at Kola-3 NPP (WWER-440, Russia) [JRC30550]. The DBTT was investigated for different amounts of the accompanying or alloying elements Cu, P and Ni. Mechanical properties were tested using Charpy impact tests. An increased amount of Ni and P, respectively, increases the DBTT shift [JRC30550]. For all investigated model alloys, increasing amounts of P, Cu, or Ni increase the DBTT shift nearly independent of each other except for low Ni concentrations. A comparison of the model alloy results to commercial steel behaviour revealed good agreement, especially in the middle P (~0.012 wt.-%) group. In this case, Cu and Ni are seen to be key elements in irradiation embrittlement [JRC24564]. The deleterious effect of nickel on the embrittlement behaviour was also reported by Slugen et al. [JRC30141] for WWER-1000 base material (15Kh2NMFAA), WWER-1000 weld material (15Kh2N2MAA), and WWER-440/312 weld material (Sv-10KhMFT). Debarberis et al. performed an intensive comparison between several data and their semi-mechanistic model. As one outcome, they found that especially increased Ni content increases the shift of the DBTT with neutron fluence significantly, independent if the specimen originates from a commercial RPV steel or a model alloy [JRC30546] [JRC30648].

Results of Valo et al. derived in the FRAME project, however, show that responses of model alloys and commercial steels to neutron irradiation differ significantly from each other [JRC32254] with a faster proceeding of embrittlement in model alloys. This is traced back to the more simple structure of the model alloys compared to the rather complex structured commercial steels. As an outcome of the FRAME project, model alloys are hence seen to be inappropriate to model irradiation embrittlement of commercial steels. Different formulas for the DBTT shift in both model alloys and commercial steels are presented with a temperature factor and a chemistry factor describing P, Cu, and Ni content. Especially the formula for commercial steels is demonstrated to be practically applicable with some exceptions [Valo2007b]. The revealed data on commercial steels were compared to trend curve formulas which are part of different national codes and standards. In some cases, the DBTT shift is undesirably underestimated. For the model alloys, the accordance is even worse [Valo2007a].

A simplified semi-mechanistic irradiation embrittlement model was set up considering the influence of Cu and P as alloying elements to a certain extent [JRC28685] [JRC30552]. The implementation of the Ni influence on radiation damage is less straightforward and took more effort, but was done successfully up to a certain limit [JRC30546]. The model, however, needs readjustment, especially when considering high-Ni containing steels [JRC33283].

Ahlstrand et al. reported about irradiation programmes of four commercial weld materials with various amounts of Ni irradiated at Novovoronezh-5 (WWER-1000, Russia, 290 °C irradiation temperature). Consistent with other findings, they found an increased shift of the DBTT with increasing Ni content. A comparison between their and the data from other sources – not further detailed – show good agreement. They also present a model equation for the DBTT shift taking into account the Ni content [JRC26623]. Recent investigations by Kryukov et al. revealed that – when Cu as accompanying element is present beyond 0.13 wt.-% – in WWER-440 RPV steels the flux plays a role for the DBTT shift. After annealing, this effect is reduced [JRC86552].

Verheyen et al. performed irradiation experiments (300 °C, up to 0.2 dpa) of different Fe alloys with various amounts of Cu. The addition of Cu induces much more hardening, depending on the initial Cu content. A plateau is reached even at rather low doses (around 0.05 dpa) with no further hardening with increasing dose [Verheyen2006].

In addition to the role of Cu, Lambrecht et al. performed irradiation experiments in the BR2 material test reactor at 290 °C and 150 bar up to doses of 0.1 dpa, which corresponds to around 40 years of operation, a typical end-of-life for NPPs around the world. Several model alloys with varying amounts of Cu, Ni, and manganese (Mn) and a commercial RPV steel were investigated. While the contribution of Cu to the hardening process saturates at some level consistent with literature, the Mn- and Ni-related features become dominant at higher doses and cause further hardening [Lambrecht2008]. The results indicate that these additional irradiation hardening features may be of relevance for long term operation, especially in low-Cu steels.

The presence of chromium (Cr) as alloying element reduces the radiation sensitivity significantly. This effect saturates at around 2 wt.-% of Cr and is reversed to a negative effect when Cr content exceeds 6 wt.-% [JRC33283].

The presence of certain alloying or accompanying elements influences the efficiency of the annealing process to reduce the effects of irradiation embrittlement. In this context, the presence of P is deleterious meaning that with increasing P content the residual embrittlement is higher. Similar effects are seen with Cu as another accompanying element, with the degree of recovery depending highly on the Cu content [JRC46534].

In the context of the ATHENA project, base and weld metal as used in Magnox reactors was irradiated. The effect of irradiation-induced embrittlement was found to be lower in Magnox steel than expected [EUR23207].

Results from the LONGLIFE project revealed that the irradiation damage rate depends clearly on the Ni, P, and Cu content. The presence of all three elements is in similar way as mentioned above also deleterious on the embrittlement rate when regarding long term operation issues [Brumovsky2014b].

1.5.5 Temperature

With increasing temperature, the mobility of atoms increases enhancing recombination of vacancies and interstitials. This explains why in general, irradiation damage decreases with increasing temperature. A recent literature survey by Ballesteros et al. confirmed this general trend. The flux, however, seems to play a role in the temperature dependence of irradiation embrittlement resulting in discrepancies between LWR and test reactor irradiations [JRC63603]. Irradiation hardening as a function of temperature can be described by a hyperbolic tangent function with low temperature dependence in the range 50–150 °C and 350–400 °C and a strong temperature dependence in between [JRC31151]. At higher temperatures, irradiation embrittlement begins to heal out due to annealing phenomena [Valo2007a].

Debarberis et al. performed investigations and collected data for the temperature range between 220 – 315 °C [JRC31151]. They found a reduced DBTT shift with increasing irradiation temperature. In addition, they included the influence of temperature into their semi-mechanistic model. A comparison with existing data (references in [JRC31151]) revealed good agreement between the respective data sets on the one hand and the extended semi-mechanistic model and experimental data on the other hand.

On the influence of temperature, WWER-440 and BWR specimens show similar behaviour [JRC30648].
Irradiation of commercial JRQ steel at various temperatures revealed a behaviour consistent with expectations as described in literature. In addition, it was found that the initial microstructure (i.e. the manufacturing history) and pre-strain have a significant influence on the irradiation embrittlement [JRC46587] [PISA2005].

Results of Valo et al. derived in the FRAME project, however, show that responses of model alloys and commercial steels to neutron irradiation differ significantly from each other [JRC32254]. Different formulas for the DBTT shift in both model alloys and commercial steels are presented with an irradiation temperature factor and a chemistry factor. Especially the formula for commercial steels is demonstrated to be practically applicable [Valo2007b].

WWER-440 cladding materials (ASS) were irradiated up to a dose of 1020 cm–2 (E > 1 MeV) at 20 °C and 300 °C. With increasing temperature, the dependence of tensile properties on neutron fluence is reduced [JRC34527].

Compiling irradiation data of RPV cladding specimens revealed that the temperature effect is more pronounced than the fluence effect. In this context, the delta-ferrite content plays a role in a way that the irradiation effect is slightly reduced with increasing delta-ferrite content [Keim2012].

Measurements at various temperatures in the course of the LONGLIFE project revealed no changes in temperature dependence at fluence levels relevant for 60 or 80 years of operation [Brumovsky2014b].

1.5.6 Microstructure

WWER-1000 base metal with different initial microstructure was investigated. It was found that in a heat-affected zone near welds irradiation-induced embrittlement is more pronounced than in the base metal itself, additionally influenced by the heat treatment prior and after the welding. In addition for JRQ steel, it was found that the initial microstructure (i.e. the manufacturing history) and pre-strain have a significant influence on the irradiation embrittlement. [JRC46587] [PISA2005].