Scaling of the scrape-off layer width during inter-ELM H modes on MAST as measured by infrared thermography

The power load to the divertor surfaces is a key concern for future devices such as ITER, due to the thermal limits on the material surface. One factor that characterizes the heat flux to the divertor is the fall off length in the scrape-off layer (SOL), which recent empirical scalings have shown could be as small as 1 mm. These predictions are based on a multi-machine scaling of the heat flux width fitted using an expression for the divertor heat flux profile which includes a term for the exponential decay in the SOL and diffusion about the last closed flux surface (LCFS) in the private flux region. This expression has been used to fit a database of inter-ELM H mode profiles at the upper divertor and extract the fall off length, λ_q, for a range of different plasma parameters in double null plasmas. The midplane separation between the primary and secondary LCFS (δr_sep) of the double null plasma used in the study are in the range 2 ⩽ δr_sep ⩽ 7 mm and no correlation is seen between the fall off length and the δr_sep. The MAST data shows good agreement with the formula, with the fitted fall off length spanning a range of 5–11 mm in the data base generated. Regression of this data has shown that the fall off length has the strongest dependence on the plasma current (or equivalently, the poloidal magnetic field at the outboard midplane) to the power −0.71. The scaling with the smallest χ² error utilizes the poloidal magnetic field at the outboard midplane (B_pol,omp) and the power crossing the SOL in the relation λ_q(mm)=1.84(±0.48)B_pol,omp^{−0.68(±0.14)}P_SOL^0.18(±0.07) with χ² = 3.46 and R² = 0.56 as a goodness of fit. The equivalent scaling with plasma current is λ_q(mm)=4.57(±0.54)I_p^{−0.64(±0.15)}P_SOL^0.22(±0.08) with χ² = 3.84 and R² = 0.55. The moderate goodness of fit suggests that additional plasma parameters are required to accurately reproduce the observed variation in λ_q. (paper)