The high heat flux testing of neutron-irradiated divertor mock-ups and the characterisation of neutron-irradiated material samples was recently completed at Forschungszentrum Jülich (FZJ), Germany. This worldwide unique R&D effort has provided essential information on the performance of the ITER divertor, which is now expected to meet the requirements of 1000 cycles of 10MW/m2 for 400s and 100 transient cycles of 20MW/m2 for 10s, even after a neutron irradiation corresponding to about 10 years of full performance plasma operation.

The neutron irradiation was carried out in the fission High Flux Reactor fission reactor located at the Joint Research Centre of Petten, The Netherlands. It is a light watercooled and moderated multipurpose research reactor, which operates at a nominal thermal power of 45 MW.

The effect of energetic neutrons on the performance of the divertor components is an important issue for ITER and future fusion devices as the material thermal conductivity can be reduced leading to a potential reduction of the power handling capabilities. In the divertor there are two critical components, the inner and outer divertor targets, where the plasma is allowed to touch the plasma facing material resulting in surface heat loads in the range of 10 to 20MW/m2. Two plasma facing materials being proposed are, Carbon Fibre Composite (CFC) and tungsten (W). To study the effects of the neutron damage on these components, irradiation experiments PARIDE 1 and 2 were carried out in the mid Nineties in the High Flux Reactor (HFR) in Petten (The Netherlands) with a neutron damage of 0.35 displacements per atom (dpa) in carbon at irradiation temperatures of 350 and 700°C. Post irradiation testing was then performed at FZJ. Subsequently, a new irradiation programme was launched which consisted of two parts:

• PARIDE 3: temperature: ≈200°C; cumulative dpa: ≈0.2 in carbon; ≈0.15 in tungsten.

• PARIDE 4: temperature: ≈200°C; cumulative dpa: ≈1.0 in carbon; ≈0.6 in tungsten.

The 0.2 dpa (carbon) level corresponds to the integrated neutron damage of the divertor plasma facing components after about 10 years of full performance plasma operation, while the 1.0 dpa (carbon) level is that expected for basic machine components in the main reactor chamber of ITER.

The irradiation mock-ups are similar in design to the proposed ITER divertor targets and comprise either CFC or W plasma facing material or armour, joined onto an actively cooled CuCrZr heat sink. A soft pure copper interlayer between the armour and the CuCrZr is used to alleviate the joint stresses. Both monoblock and flat tile geometries were investigated in these experiments. The monoblock consists of armour tiles with a hole into which a cooling tube is inserted and joined to the tiles. The flat tile geometry consists of a hollow CuCrZr rectangular bar, the cooling channel, on to which the flat plasma facing tiles are joined. Testing of the mock-ups, before and after irradiation, under the relevant heat flux conditions, was performed in the electron beam facility, JUDITH, located in a hot cell at FZJ.

In the unirradiated condition both CFC flat tiles and monoblocks surpassed the ITER requirements. The flat tile survived 1000 cycles at both 11.5 and 20 MW/m2 before failing at 23 MW/m2. The CFC monoblock survived 1000 cycles at 19 MW/m2, and 700 cycles at 23 MW/m2. The testing was curtailed at this very high power (about twice higher than the ITER requirements) because of sublimation of the carbon at the high surface temperatures.

After irradiation to a level of 0.2 dpa, two CFC flat tile mock-ups failed at 19.5 MW/m2. The 1 dpa irradiated sample testing was limited to 15 MW/m2 due to surface temperatures of 2500°C causing excessive sublimation. These values are lower than pre-irradiation values but still surpass the ITER steady-state power handling requirements.

Thermal fatigue testing of a CFC monoblock irradiated to 0.2 dpa was carried out at 10 MW/m2 for 1000 cycles and 12 MW/m2 for 1000 cycles without failure. When the power was increased to 14 MW/m2 excessive erosion occurred due to high surface temperatures which made thermal cycling experiments impossible.

W-armoured mock-ups in the flat tile unirradiated geometry were successfully tested for 1000 cycles at 8 MW/m2 and 14 MW/m2 (this is absorbed power density, the surface heat flux is approximately double this value). After irradiation however, a failure occurred at approximately 10 MW/m2. Post-mortem metallography showed detachment of the armour tiles at the W/Cu joint caused by irradiation-induced embrittlement of the pure Cu.

In contrast to the flat tile geometry, the W monoblocks showed no degradation in thermal fatigue performance, even after irradiation, and all survived 1000 cycles at 18 MW/m2.

Thermal conductivities were determined for the CFC material grade NB31 and the corresponding silicon-doped material NS31, produced by Snecma Propulsion Solide, France, before and after irradiation. As expected, the thermal conductivities were reduced significantly by the irradiation and the reduction increased with increasing neutron dose. For all three directions of the anisotropic CFC, the room temperature conductivity was reduced to 17% of the unirradiated value after 0.2 dpa (to 10% after 1 dpa). The effect of neutron irradiation on the thermal conductivity at high temperature was smaller (33% of the unirradiated thermal conductivity was measured at 700°C). Furthermore, a significant recovery of the CFC thermal conductivity (up to 80% of the unirradiated value) was observed during the post-irradiation high heat flux test. The thermal conductivities for the silicon-doped material NS31 and their neutron-induced degradation were comparable with those of NB31.

Pure tungsten and W-1%La2O3 were also tested in the unirradiated state and after irradiation in the PARIDE 3 and PARIDE 4 campaigns. Both materials showed a very small thermal conductivity degradation. At room temperature, it was reduced to 80% and 73% of the unirradiated value after 0.15 and 0.6 dpa, respectively. But at higher irradiation temperatures annealing became effective, and the conductivity degradation was negligible.

Important conclusions have been derived from the results of the testing of irradiated divertor mock-ups and material samples. The effect of the neutron damage on the performance of the mock-ups is greater in a flat tile than in a monoblock geometry, probably because the strain range in the pure Cu interlayer during operation is a factor of three higher in the flat tile configuration and this material rapidly becomes very brittle under irradiation. The thermal conductivity degradation in the CFC was determined and showed a significant recovery after the high heat flux testing. The data obtained in these experiments demonstrate that the effect of about 10 years full performance plasma operation in ITER will only cause an increase of the ITER CFC outer divertor target temperature by, at most, about 400°C under the reference normal operation power load of 10MW/m2. In reality, as the neutron irradiation takes place in ITER the CFC target will be eroded contributing to a reduction in the surface temperature, which will be remain close to its reference value (about 1500 °C) throughout the ITER divertor target lifetime. It is also important to note that the measured degradation of the W thermal conductivity under irradiation should not represent a major JUDITH facility at FZJ issue for the operation of ITER with a W target.

JUDITH is a 60 kW powerful high heat flux testing facility operated at the Forschungszentrum Jülich (FZJ), Germany. It is used to simulate the cyclic thermal loads, which act on  the plasma-facing componentsof a fusion reactor. Being located in a hot cell, it can test irradiated mockups with or without beryllium armour. A new facility, named JUDITH-2, is under construction at FZJ with a larger vacuum chamber and a 200 kW electron beam gun. Its completion is planned by the end of 2004.

For more information see: