So now you understand, we fuse two elements (here we have chosen Deuterium and Tritium - two isotopes of Hydrogen) in plasma form and after fusion they produce Helium and high energy neutrons which get absorbed in the wall of the fusion reactor. The energy of these neutrons are used to produce electricity. But you also have got aware of the problem we are facing - the self-closed magnetic mesh in which our Deuterium-Tritium-Helium plasma is present, need to selectively allow some part of Helium to come out and extract out of the experimental reactor. Thus all our future nuclear fusion reactors will need to be designed and operated in a regime of high energy confinement while particle confinement remains low enough, so that impurities can be exhausted.
To achieve this, we deliberately put some well-planned holes (/divertor) in the reactor to let the Helium go out. But these divertors tear and re-connect and thereby significantly alter few outer layers of our magnetic mesh. The 'Edge' region of our experimental reactor where this change of magnetic field topology takes place, is called 'Scrape-Off Layer (SOL)' and the outermost mesh which remains unaffected is called 'Last Closed Flux Surface (LCFS)'. The hot plasma that leaks through the LCFS, follows the magnetic field lines in SOL region and hits the divertor plate. Unfortunately the plasma remains so hot that no matter what material we use to build the divertor plate, it gets degraded. Hence, for a safe and uninterrupted run of the reactor, we have to find a solution.
What we are struggling for is, we are trying to find a set of optimum parameters such that, the plasma that leaks out from the LCFS, gets randomized (or turbulent), so that it does not hit the divertor plate at one point, rather fall into a wider region on the divertor, thereby decreasing the per-unit-area heat-load on the divertor!