Power deposition on misaligned castellated tungsten blocks in the Magnum-PSI and Pilot-PSI linear devices

For the final design of the ITER divertor it is important to determine whether shaping of each tungsten monoblock to eliminate leading edges is required or not. In order to aid this decision, two experiments were performed in DIFFER's linear plasma devices to study heat loads on misaligned water cooled blocks at glancing incidence. First, a series of tungsten blocks were exposed to a high parallel heat flux (26 MW $m^{-2}$) hydrogen plasma beam in the Magnum-PSI linear device. The blocks were exposed at an oblique angle between $4^{\circ }$ and $7^{\circ }$ with respect to the plasma beam, approaching the low angle of incidence expected in the ITER divertor strike-point regions. One block was vertically misaligned with respect to the others by a value between 0 and $1.2$ mm, and the surface temperature evolution monitored. This was compared to a finite element thermo-mechanical model of the system, to investigate any discrepancies between the two. The importance of Larmor orbit smoothing effects could also be considered, as the Larmor radius was larger than half the misalignment height. Next, a row of identical tungsten blocks were exposed at an angle of $3^{\circ }$ to  ∼500–700 μs pulsed hydrogen plasmas of peak power between 275–625 MW $m^{-2}$ in the Pilot-PSI linear device, to simulate the effect of edge localized mode (ELM) transients. One block was misaligned with respect to the others by 1 mm. In all cases, the model and experiment were found to be in excellent agreement (to  ∼1%), demonstrating clearly that an optical approximation for plasma power loading on misaligned edges is appropriate. A simple model including Larmor smoothing was applied, the results of which give a similar prediction to the optical approximation for the temperature of the misaligned edges, within the error of the temperature measurements, though a small decrease in edge temperature is calculated. Therefore it can be concluded that the application of the optical approximation is generally appropriate as an input physics assumption for the study of shaping alternatives. (paper)