Reactivity feedback coefficients of a material test research reactor fueled with high-density U3Si2 dispersion fuels

The reactivity feedback coefficients of a material test research reactor fueled with high-density U₃Si₂ dispersion fuels were calculated. For this purpose, the low-density LEU fuel of an MTR was replaced with high-density U₃Si₂ LEU fuels currently being developed under the RERTR program. Calculations were carried out to find the fuel temperature reactivity coefficient, moderator temperature reactivity coefficient and moderator density reactivity coefficient. Nuclear reactor analysis codes including WIMS-D4 and CITATION were employed to carry out these calculations. It is observed that the average values of fuel temperature reactivity feedback coefficient, moderator temperature reactivity coefficient and moderator density reactivity coefficient from 20 °C to 100 °C, at the beginning of life, followed the relationships (in units of Δk/k × 10⁻⁵ K⁻¹) −2.116 − 0.118 ρ_U, 0.713 − 37.309/ρ_U and −12.765 − 34.309/ρ_U, respectively for 4.0 ≤ ρ_U (g/cm³) ≤ 6.0.     

JR curves of the low alloy steel 20 MnMoNi 5 5 with two different sulphur contents in oxygen-containing high temperature water at 240°C

J_R curves of the low alloy steel 20 MnMoNi 5 5 with two different sulphur contents (0.003 and 0.011 wt.%) were determined at 240°C in oxygen-containing high temperature water as well as in air. The tests were performed by the single-specimen unloading compliance technique at load line displacement rates from 1 × 10⁻⁴ down to 1 × 10⁻⁶ mm s⁻¹ on 20% side-grooved 2T CT specimens in an autoclave testing facility at an oxygen content of 8 ppm and a pressure of 7 MPa under quasi-stagnant flow conditions.In the case of testing in high temperature water, remarkably lower J_R curves than in air at the same load line displacement rate (1 × 10⁻⁴ mm s⁻¹) were obtained. A decrease in the load line displacement rate as well as an increase in the sulphur content of the steel caused a reduction of the J_R curves. At the fastest load line displacement rate a stretch zone could be detected fractographically on the specimens tested in air and in high temperature water and consequently J_i could be determined. When testing in high temperature water, the J_i value of the higher sulphur material type decreases from 45 N mm⁻¹ in air to 3 N mm⁻¹, much more than that of the optimized material type from 51 N mm⁻¹ in air to 20 N mm⁻¹ at 1 × 10⁻⁴ mm s⁻¹.     

Results on research of templates from Kozloduy-1 reactor pressure vessel

The studies on the specimens manufactured from the templates cut out from the weld 4 of Kozloduy NPP Unit 1 reactor vessel have been conducted. The data on chemical composition of the weld metal have been obtained. Neutron fluence, mechanical properties, ductile to brittle transition temperature (DBTT) using mini Charpy samples have been determined. The phosphorus and copper content averaged over all templates is 0.046 and 0.1 wt.%, respectively. The fluence amounted up to 5×10¹⁸ n cm⁻² within 15–18 fuel cycles, and about 5×10¹⁹ n cm⁻² for the whole period of operation. These values agree well with calculated data. DBTT was determined after irradiation (T_k) to evaluate the vessel metal state at the present moment, then after heat treatment at the temperature of 475°C to simulate the vessel metal state after thermal annealing (T_an), and after heat treatment at 560°C to simulate the metal state in the initial state (T_k0). As a result of the tests the following values were obtained: T_k, +91.5°C; T_an, +63°C; and T_k0, 54°C. The values of T_k and T_an obtained by measurements were found to be considerably lower than those predicted in accordance with the conservative method accepted in Russia (177°C for T_k and 100°C for T_an). Thus, the obtained results allowed to make a conclusion that it is not necessary to anneal Kozloduy NPP Unit 1 reactor vessel for the second time. The fractographic and electron-microscopic research allowed to draw some conclusions on the embrittlement mechanism.     

Ductile–brittle transition of polycrystalline iron and iron–chromium alloys

Fracture toughness of polycrystalline Fe, Fe–3%Cr and Fe–9%Cr was measured by four-point bending of pre-cracked specimens at temperatures between 77 K and 150 K and strain rates between 4.46 × 10⁻⁴ and 2.23 × 10⁻² s⁻¹. For all materials, fracture behaviour changed with increasing temperature from brittle to ductile at a distinct brittle–ductile transition temperature (T_c), which increased with increasing strain rate. At low strain rates, an Arrhenius relation was found between T_c and strain rate in each material. At high strain rates, T_c was at slightly higher values than those expected from extrapolation of the Arrhenius relation from lower strain rates. This shift of T_c was associated with twinning near the crack tip. For each material, use of an Arrhenius relation for tests at strain rates at which specimens showed twinning gave the same activation energy as for the low strain rate tests. The values of activation energy for the brittle–ductile transition of polycrystalline Fe, Fe–3%Cr and Fe–9%Cr were found to be 0.21, 0.15 and 0.10 eV, respectively, indicating that the activation energy for dislocation glide decreases with increasing chromium concentration in iron.     

Plutonium credit in the low enriched uranium fuel of PARR-1

The amount of plutonium (Pu) isotopes and the resultant savings of ²³⁵U due to their production were calculated in the low enriched uranium (LEU) fuel, being utilized in Pakistan Research Reactor-1 (PARR-1). Further the importance map and relative importance map for different isotopes of Pu were also determined. Equilibrium PARR-1 core was achieved for these calculations. MTR-PC26 package was used to generate the microscopic cross-sections data for 45 elements including fissile/structural materials and also the fission products. Finite difference reactor core analysis code CITATION was employed for the fuel management analysis and static depletion calculations.The results indicated that PARR-1 core has attained its equilibrium state after eleven cycles with each cycle of duration about forty full power (10 MW) days. Further, the results showed that at the beginning of equilibrium cycle (BOEC) of the PARR-1 core, net reactivity addition due to all isotopes of Pu was 4.86 × 10⁻³Δk/k. Amount of ²³⁵U equivalent to this value of reactivity was found to be 15.58 ± 0.021 g. Plots of importance and relative importance maps predicted higher isotopic concentrations of Pu in the fuel elements located in the vicinity of central water box.     

Effect of irradiation by 140 Mev Ag ions on the optical and electrical properties of polypropylene/TiO2 composite

Changes in the optical, structural, dielectric properties and surface morphology of a polypropylene/TiO₂ composite due to swift heavy ion irradiation were studied by means of UV–visible spectroscopy, X-ray diffraction, impedance gain phase analyzer and atomic force microscopy. Samples were irradiated with 140 MeV Ag¹¹⁺ ions at fluences of 1 × 10¹¹ and 5 × 10¹² ions/cm². UV–visible absorption analysis reveals a decrease in optical direct band gap from 2.62 to 2.42 eV after a fluence of 5 × 10¹² ions/cm². X-ray diffractograms show an increase in crystallinity of the composite due to irradiation. The dielectric constants obey the Universal law given by ε α f ⁿ⁻¹, where n varies from 0.38 to 0.91. The dielectric constant and loss are observed to change significantly due to irradiation. Cole–cole diagrams have shown the frequency dependence of the complex impedance at different fluences. The average surface roughness of the composite decreases upon irradiation.     

Fuel irradiation of the first batches produced for the Chinese HTR-10

An irradiation test of four spherical fuel elements (SFE) had been performed in the Russian reactor IVV-2M. The elements were sampled randomly from the first and second product batches which were manufactured for the 10 MW high-temperature gas-cooled test reactor (HTR-10). The maximum burnup of the irradiated fuel elements reached 107,000 MWd/tU and the maximum fast neutron fluence was 1.31 × 10²⁵ m⁻². The release-to-birth rate ratio (R/B) did not increase significantly during irradiation. However, an in-pile heating-up test of element SFE 7 in Capsule 5 led to a failure of approximately 6% of the coated particles. After the test it was estimated that the fuel temperature had very likely been much higher than the intended 1600 °C.     

Power up-grading study for the first Egyptian research reactor

In the present work, power up-grading study is performed, for the first Egyptian Research Reactor (ET-RR-1), using the present fuel basket with 4×4 fuel rods, (17.5 mm pitch), and a proposed fuel basket with 5×5 fuel rods, (14.0 mm pitch), without violating the thermal hydraulic safety criteria. These safety criteria are; fuel centerline temperature (fuel melting), clad surface temperature (surface boiling), outlet coolant temperature, and maximum heat flux (critical heat flux ratio). Different thermal reactor powers (2–10 MW) and different core coolant flow rates (450, 900, 1350 m³ h⁻¹) are considered. The thermal hydraulic analysis was performed using the subchannel code COBRA-IIIC for the estimation of temperatures, coolant velocities and critical heat flux. The neutronic calculations were performed using WIMS-D4 code with 5 — group neutron cross section library. These cross sections were adapted to use in the two-dimensional (2-D) diffusion code DIXY for core calculations. The study concluded that ET-RR-1 power can be upgraded safely up to 4 MW with the present 4×4-fuel basket and with the proposed 5×5-fuel basket up to 5 MW with the present coolant flow rate (900 m³ h⁻¹). With the two fuel arrays, the reactor power can be upgraded to 6 MW with coolant flow rate of 1350 m³ h⁻¹ without violating the safety criterion. It is also concluded that, loading the ET-RR-1 core with the proposed fuel basket (5×5) increases the excess reactivity of the reactor core than the present 4×4 fuel matrix with equal U-235 mass load and gave better fuel economy of fuel utilization.     

The fracture toughness measured on sensitized 304 stainless steel in simulated reactor water

The environmental conditions chemically equivalent to BWR primary water, e.g. 288°C, 0.2 ppm O₂ and/or 98°C, air-saturated, were found to influence considerably the in-water fracture toughness values of furnace-sensitized Type 304 stainless steel.Notched compact tension and three point bend specimens sampled from two heats of standard materials (0.06% C) showed significant reduction in dJ/da values reflecting consistently the effects of loading rate, temperature, dissolved oxygen concentration and degree of sensitization. In particular the crack enhancement with lowering the loading rate was significant. The effect became apparent with dJ dt at and below 1× 10−1 kg·mm/mm²/min (1.6 × 10 J/m²/s) in the typical BWR environment.Based on the results, it is suggested that a critical consideration is needed on the significance of such an environmental effect in the LWR structural safety evaluation, in particular that the probability of instable fracture at the “rings” of sensitized material near welded joints is subject to reviewing.     

Prospects for fission gas extraction from in-reactor UO2 nuclear fuel elements

Due to the many problems encountered in the design of fuel rods for the safe operation of commercial nuclear reactors, caused by the fission gases generated by the fission of fissile material, it was considered opportune to make a theoretical analysis of the feasibility of extraction of fission gases from the fuel rod while in operation.This analysis in the steady state of a Zircaloy-2 sheathed fuel rod containing UO₂ as a fuel, with a 2 mm (2.7 vol.%) diameter porous graphite cylinder inserted in the centre, has demonstrated that a total volume of fission gases (xenon, krypton, and iodine) of about 1.1 × 10⁻⁶ cm³/s (at STP) can be extracted from the fuel rod at a controlled rate, determined by the inherent property of fission gas migration towards the centre of the fuel rod from its place of formation. In this analysis, the fuel rod was assumed to be subjected to irradiation in a reactor the size of a Bruce “A” reactor, operating at 3000 megawatts thermal power. The extracted volume of gas was calculated on a 900 h cycle after the first 90 h of reactor operation had elapsed.     

Diffusion of radiogenic helium in natural uranium oxides

The issue of nuclear waste management – and especially spent fuel disposal – demands further research on the long-term behavior of helium and its impact on physical changes in UO₂ and (U,Pu)O₂ matrices subjected to self-irradiation. Helium produced by radioactive decay of the actinides concentrates in the grains or is trapped at the grain boundaries. Various scenarios can be considered, and can have a significant effect on the radionuclide source terms that will be accessible to water after the canisters have been breached. Helium production and matrix damage is generally simulated by external irradiation or with actinide-doped materials. A natural uranium oxide sample was studied to acquire data on the behavior of radiogenic helium and its diffusion under self-irradiation in spent fuel. The sample from the Pen Ar Ran deposit in the Vendée region of France dated at 320 ± 9 million of years was selected for its simple geological history, making it a suitable natural analog of spent fuel under repository conditions during the initial period in a closed system not subject to mass transfer with the surrounding environment. Helium outgassing measured by mass spectrometry to determine the He diffusion coefficients through the ore shows that: (i) a maximum of 5% (2.1% on average) of the helium produced during the last 320 Ma in this natural analog was conserved, (ii) about 33% of the residual helium is occluded in the matrix and vacancy defects (about 10⁻⁵ mol g⁻¹) and 67% in bubbles that were analyzed by HRTEM. A similar distribution has been observed in spent fuel and in (U_0.9,Pu_0.1)O₂. The results obtained for the natural Pen Ar Ran sample can be applied by analogy to spent fuel, especially in terms of the apparent solubility limit and the formation, characteristics and behavior of the helium bubbles.     

Candu type fuel behavior during rapid overpower transients

Results obtained in the pulse irradiation tests performed on segmented fuel elements in the Romanian Annular Core Pulse Reactor (ACPR) are discussed below. Tests included the effects of initial element internal pressure and a wide range of energy deposition on the fuel element behavior. All tests were conducted in stagnant water at room temperature and atmospheric pressure inside the capsule. The fuel elements were instrumented with thermocouples for cladding surface temperature measurement. Transient histories of reactor power, cooling water pressure, fuel element internal pressure and cladding temperature were recorded during the tests. The fuel elements were subjected to total energy depositions from 70 to 265 cal g⁻¹ UO₂. Cladding failure mechanism and the failure threshold have been established. The fuel failure mechanism is a burst type and is very similar to LOCA failure mechanism even though the rate energy deposition is higher in the ACPR tests. At higher energy deposition brittle cladding fracture near endcap weld region can be produced. The failure threshold is situated between 190 and 200 cal g⁻¹ UO₂ for standard fuel rod (0.2–0.3 MPa internal pressure) and less than 160 cal g⁻¹ UO₂ for pressurized fuel rods (internal pressure between 1 and 3.0 MPa). Pre-pressurization could be an important factor to control the failure threshold energy. The experimental program is still in progress and new experiments are foreseen to be performed in the following period.     

Increased fuel burn up in a CANDU thorium reactor using weapon grade plutonium

Weapon grade plutonium is used as a booster fissile fuel material in the form of mixed ThO₂/PuO₂ fuel in a Canada Deuterium Uranium (CANDU) fuel bundle in order to assure the initial criticality at startup.Two different fuel compositions have been used: (1) 97% thoria (ThO₂) + 3%PuO₂ and (2) 92% ThO₂ + 5% UO₂ + 3% PuO₂. The latter is used to denaturize the new ²³³U fuel with ²³⁸U. The temporal variation of the criticality k_∞ and the burn-up values of the reactor have been calculated by full power operation for a period of 20 years. The criticality starts by k_∞ = 1.48 for both fuel compositions. A sharp decrease of the criticality has been observed in the first year as a consequence of rapid plutonium burnout. The criticality becomes quasi constant after the second year and remains above k_∞ > 1.06 for 20 years. After the second year, the CANDU reactor begins to operate practically as a thorium burner.Very high burn up could be achieved with the same fuel material (up to 500,000 MW·D/T), provided that the fuel rod claddings would be replaced periodically (after every 50,000 or 100,000 MW·D/T). The reactor criticality will be sufficient until a great fraction of the thorium fuel is burnt up. This would reduce fuel fabrication costs and nuclear waste mass for final disposal per unit energy drastically.     

Seismic loads for crack stability assessment, in a review of leak-before-break (LBB) applicability

This study was performed to define the seismic loading conditions for use in the crack stability assessment of a BWR for the applicability of the leak-before-break (LBB) criterion. The LBB has been applied to the design of Class 1 piping in Japanese light-water reactors. Crack penetrated condition with detectable leak in the LBB applicability review is classified into the Level C service condition. Here an S₁-earthquake (maximum design earthquake) is currently assumed, rather than an S₂-earthquake (extreme design earthquake). In order to justify this assumption, the frequency of an S₁-earthquake occurring during coolant leakage due to crack propagation was determined. The frequency of coolant leakage from Class 1 piping must be less than that of the Level C service condition (2.5 × 10⁻² to 1 × 10⁻⁴ per year) in order to assume that an S₁-earthquake is appropriate. Accordingly, the frequency of coolant leakage from Class 1 piping was calculated using a probabilistic fracture mechanics (PFM). The results of this analysis indicate that the frequency of coolant leakage from Class 1 piping is less than the expected occurrence of an S₁-earthquake. As the results, it is concluded that the assumption of the seismic loading employed in an LBB applicability review should be appropriate.     

The damaging effects of nitrogen ion beam implantation on upland cotton (Gossypium hirsutum L.) pollen grains

With the aim to study the effects of an ion beam on plant cells, upland cotton (Gossypium hirsutum L.) cultivar “Sumian 22” pollen grains were irradiated in vacuum (7.8 × 10⁻³ Pa) by low-energy nitrogen ions with an energy of 20 keV at various fluences ranging from 0.26 × 10¹⁶ to 0.78 × 10¹⁶ N⁺/cm². The irradiation effects on pollen grains were tested, considering the ultrastructural changes in the exine and interior walls of pollen grains, their germination rate, the growth speed of the pollen tubes in the style, fertilization and boll development after the pistils were pollinated by the pollen grains which had been implanted with nitrogen ions. Nitrogen ions entered the pollen grains by etching and penetrating the exine and interior walls and destroying cell structures. A greater percentage of the pollen grains were destroyed as the fluence of N⁺ ions increased. Obviously, the nitrogen ion beam penetrated the exine and interior walls of the pollen grains and produced holes of different sizes. As the ion fluence increased, the amount and the density of pollen grain inclusions decreased and the size of the lacuna and starch granules increased. Pollen grain germination rates decreased with increasing ion fluence. The number of pollen tubes in the style declined with increased ion implantation into pollen grains, but the growth speed of the tubes did not change. All of the pollen tubes reached the end of the style at 13 h after pollination. This result was consistent with that of the control. Also, the weight and the diameter of the ovary decreased and shortened with increased ion beam implantation fluence. No evident change in the fecundation time of the ovule was observed. These results indicate that nitrogen ions can enter pollen grains and cause a series of biological changes in pollen grains of upland cotton.     

HTGR core auxiliary cooling system availability and reliability

Using fault tree techniques, a quantitative estimate is made to predict both the start-up availability and operational reliability of the core auxiliary cooling system (CACS) of an HTGR following the postulated, simultaneous occurrence of a design basis depressurization accident (DBDA) and the complete loss of main loop cooling (LOMLC). The effects of a postulated, concurrent loss of offsite power are considered as well. Several potential common mode failures are identified. The limited availability of data presents a problem to numerical evaluation and estimates of uncertainty are at best crude. To provide a basis for measure of this uncertainty, the fault trees were solved using, on a consistent basis, either ‘optimistic’ failure rates, ‘pessimistic’ failure rates, or mean values (the geometric mean).Generally, about 80% of the failure rate data was larger than the ‘optimistic’ value, while only 20% was larger than the ‘pessimistic’ value. Predicated on a variety of assumptions, many of which may be unduly pessimistic, the CACS unavailability following a postulated DBDA and LOMLC has been estimated to be between 4 × 10⁻⁷ and 3 × 10⁻⁵ for the 2000 MW (th) HTGR and between 5 × 15⁻⁷ and 5 × 10⁻⁵ for the 3000 MW (th) HTGR. At the end of 300 hr, the estimated probability that the CACS will not leave sufficient core cooling capacity varies between 9 × 10⁻⁵ and 4 × 10⁻² for the smaller plant and 3 × 10⁻⁴ and 6 × 10⁻² for the larger plant. If it is further postulated that offsite power is concurrently lost, then the estimated mean unavailability at start-up is 3 × 10⁻³ for the 2000 MW (th) plant. The estimated mean probability that the CACS of the smaller plant will not be available at start-up and not be operational after 300 hr is 8 × 10⁻².     

Criticality investigations for the fixed bed nuclear reactor using thorium fuel mixed with plutonium or minor actinides

Prospective fuels for a new reactor type, the so called fixed bed nuclear reactor (FBNR) are investigated with respect to reactor criticality. These are ① low enriched uranium (LEU); ② weapon grade plutonium + ThO₂; ③ reactor grade plutonium + ThO₂; and ④ minor actinides in the spent fuel of light water reactors (LWRs) + ThO₂. Reactor grade plutonium and minor actinides are considered as highly radio-active and radio-toxic nuclear waste products so that one can expect that they will have negative fuel costs.The criticality calculations are conducted with SCALE5.1 using S₈–P₃ approximation in 238 neutron energy groups with 90 groups in thermal energy region. The study has shown that the reactor criticality has lower values with uranium fuel and increases passing to minor actinides, reactor grade plutonium and weapon grade plutonium.Using LEU, an enrichment grade of 9% has resulted with k_eff = 1.2744. Mixed fuel with weapon grade plutonium made of 20% PuO₂ + 80% ThO₂ yields k_eff = 1.2864. Whereas a mixed fuel with reactor grade plutonium made of 35% PuO₂ + 65% ThO₂ brings it to k_eff = 1.267. Even the very hazardous nuclear waste of LWRs, namely minor actinides turn out to be high quality nuclear fuel due to the excellent neutron economy of FBNR. A relatively high reactor criticality of k_eff = 1.2673 is achieved by 50% MAO₂ + 50% ThO₂.The hazardous actinide nuclear waste products can be transmuted and utilized as fuel in situ. A further output of the study is the possibility of using thorium as breeding material in combination with these new alternative fuels.     

Development of the detecting method of helium gas leak from canister

In the concrete cask storage system, spent fuel is installed and weld-sealed in a cylindrical container called a canister. The canister is filled with helium gas and its containment shall be maintained and inspected during storage. The helium gas enhances heat removal from spent fuel. When the helium gas leaks, the effect of helium gas convection is weakened in the canister. Thereof, the temperature on the canister surface changes.In present tests, it was found that temperatures of the center of the top and the bottom on the canister surface change remarkably during the helium gas leak. Therefore, we defined the temperature difference as ΔT_BT. And one can detect helium gas leak using the change of ΔT_BT. ΔT_BT increases monotonously toward a constant value during helium gas leak, even if the inlet air temperature drops. The helium gas leak can be detected at the early stage of the leak by observing both ΔT_BT and inlet air temperature.     

Fission product release and its environment impact for normal reactor operations and for relevant accidents

The radioactive concentration in the primary loop and the radioactive release for both normal operations and accidents for the HTR-10 are calculated and presented in the paper. The coated-particle fuel is used in the HTR-10, which has good performance of retaining fission products. Therefore the radioactive concentration in the primary loop of the HTR-10 is very low, and the amount of radioactive release to the environment is also very small for both normal operation and accident conditions. The radiation doses to the public caused by radioactive release for both normal operations and accidents are given in the paper. The results show that the maximum individual effective dose to the public due to the release of airborne radioactivity during normal operations is only 1.4×10⁻⁴ mSv a⁻¹, which is much lower than the dose limit (1 mSv a⁻¹) stipulated by Chinese National Standard GB8703-86. For depressurization accident and water ingress accident, the maximum individual whole-body doses to man are only 7.7×10⁻² and 2.0×10⁻¹ mSv, thyroid doses only 1.7×10⁻¹ and 1.1 mSv, respectively. They are much lower than the prescribed minimum of emergency intervention level (whole-body dose: 5 mSv, thyroid dose: 50 mSv) for sheltering measures stipulated by the Chinese Nuclear Safety Criterion HAD002/03. The conclusion is that the environmental impact is very small for normal operations and accidents for the HTR-10, and the requirements stipulated in the Chinese Nuclear Safety Criterions are satisfied perfectly.     

Ion irradiation induced nanocrystal formation in amorphous Zr55Cu30Al10Ni5 alloy

Ion irradiation can be used to induce partial crystallization in metallic glasses to improve their surface properties. We investigated the microstructural changes in ribbon Zr₅₅Cu₃₀Al₁₀Ni₅ metallic glass after 1 MeV Cu-ion irradiation at room temperature, to a fluence of 1.0 × 10¹⁶ cm⁻². In contrast to a recent report by others that there was no irradiation induced crystallization in the same alloy [S. Nagata, S. Higashi, B. Tsuchiya, K. Toh, T. Shikama, K. Takahiro, K. Ozaki, K. Kawatusra, S. Yamamoto, A. Inouye, Nucl. Instr. and Meth. B 257 (2007) 420], we have observed nanocrystals in the as-irradiated samples. Two groups of nanocrystals, one with diameters of 5–10 nm and another with diameters of 50–100 nm are observed by using high resolution transmission electron microscopy. Experimentally measured planar spacings (d-values) agree with the expectations for Cu₁₀Zr₇, NiZr₂ and CuZr₂ phases. We further discussed the possibility to form a substitutional intermetallic (Ni_xCu_1−x)Zr₂ phase.