In order to qualify new beryllium grades for ITER, several Russian and Chinese materials were tested in the electron beam facility JUDITH-1 and compared to the reference material S65C. In a former campaign, samples from these materials were loaded in thermal shock experiments with single shots and multiple shots. The present work is an extension of this work to other loading scenarios.

Four actively cooled mock-ups were produced in Russia and in China (two by each party). These mock-ups consisted of a water-cooled CuCrZr body with four tiles from different beryllium grades. Both parties used their own joining techniques, but each of the mock-ups also contained beryllium tiles from the other party, as well as from S65C.

Each tile was loaded by the following scenarios on different surface areas:

• simulation of vertical displacement events (VDEs) at 40 MJ/m2, 1 shot, heated area a = 10 × 10 mm2, 50-ms ramp-up, 165-ms steady state

• disruption simulation at 3 MJ/m2, 1 shot, heated area a = 5 × 5 mm2, t = 5 ms

• repetitive test with 1000 shots at 80 MW/m2 (2 MJ/m2), a = 10 × 10 mm2, t = 25 ms. This loading condition is similar to one that was proposed by Sandia National Laboratory for the comparison of different beryllium grades.

Finally, one mock-up by each party underwent a thermal fatigue test with 1000 cycles at 2 MW/m2, 15 s heating, and 15 s cooling (heated area: whole sample surface). Heavy melting was observed in the area of the VDE loading, but no detachment of any of the tiles was found. Following the high-heat-flux experiments in the electron beam facility, post-mortem examinations were performed by optical photography and scanning electron microscopy on the surfaces as well as by metallography. From these analyses, no fundamental differences were found for the damage in the different beryllium grades.