Using a simple set of poloidal field coils, one can reach the situation in which the null of the poloidal magnetic field in the divertor region is of second order, not of first order as in the usual X-point divertor. Then, the separatrix in the vicinity of the null point splits the poloidal plane not into four sectors, but into six sectors, making the whole structure look like a snowflake (hence the name). This arrangement allows one to spread the heat load over a much broader area than in the case of a standard divertor. A disadvantage of this configuration is that it is topologically unstable, and, with the current in the plasma varying with time, it would switch either to the standard X-point mode, or to the mode with two X-points close to each other. To avoid this problem, it is suggested to have a current in the divertor coils that is roughly 5% higher than in an “optimum” regime (the one in which a snowflake separatrix is formed). In this mode, the configuration becomes stable and can be controlled by varying the current in the divertor coils in concert with the plasma current; on the other hand, a strong flaring of the scrape-off layer still remains in force. Geometrical properties of this configuration are analyzed. Potential advantages and disadvantages of this scheme are discussed.

No abstract

The snowflake magnetic configuration is characterized by the presence of two closely spaced poloidal field nulls that create a characteristic hexagonal (reminiscent of a snowflake) separatrix structure. The magnetic field properties and the plasma behaviour in the snowflake are determined by the simultaneous action of both nulls, this generating a lot of interesting physics, as well as providing a chance for improving divertor performance. Among potential beneficial effects of this geometry are: increased volume of a low poloidal field around the null, increased connection length, and the heat flux sharing between multiple divertor channels. The authors summarise experimental results obtained with the snowflake configuration on several tokamaks. Wherever possible, relation to the existing theoretical models is described.

This paper summarizes recent progress in the theory of a snowflake divertor, a possible path to reduce both steady-state and intermittent heat loads on the divertor plates to an acceptable level. The most important feature of a SF divertor is the presence of a large zone of a very weak poloidal magnetic field around the poloidal field (PF) null. Qualitative explanation of a variety of new features characteristic of a SF divertor is provided based on simple scaling relations. The main part of the paper is focused on the concept of spreading of the heat flux by curvature-driven convection near the PF null. References to experimental results from the NSTX and TCV tokamaks are provided.

No abstract