Join Zoom Meeting
https://eu02web.zoom-x.de/j/67474864266?pwd=vpNeMPZbraPSYwBLLNvNL55WQI4aKv.1
Meeting ID: 674 7486 4266
Passcode: DEMOPSD
The density limit following Greenwlad is a very strong design driver for DEMO, for example allowing for fGr=1.6 could allow for a DEMO producing as much electricity as a present EPR (1GW). (see Mattia's slides for more details)
Recent work have revisited past understanding of the density limit on both the turbulent transport (resistive modes) and the radiative collpase. They are nicely summarized and compared against AUG database (at one Ip, B and low Zeff) in Manz NF 2023 https://iopscience.iop.org/article/10.1088/1741-4326/acd9db/meta
On 3 tokamaks (TCV, AUG and JET) a model based on resistive balloning modes Giacomin PRL 2022 https://doi.org/10.1103/PhysRevLett.128.185003 shows that the Greenwlad density at the separatrix (rather than the line average density) multiplied by P_sep^~0.5/B^~0.4 is predicted and allows to explain the density limit of the 3 devices.
During the discussion, it was pointed out (by Clemente) that the 'Giacomin' limit is very close to greenwald for AUG, so the question was on how trustable the lack of additionnnal size dependance is ? as this is key when extrapolating to DEMO. It was reminded by Peter that the fact that the turbulent transport limit is close to the radiative collpase limit on AUG does not prevent both limits to exist and to lead to different constraints when extrapolating to DEMO. Also the database work on AUG shown by Peter was for a given B, Ip, there are no data reaching the density limits at lower B or lower Ip. Nor enough data at higher Zeff. Nicola mentionned various on-going project on database nsep, Tsep work by Ondrej Grover and others.
future experiments should be guided by database work showing the empty corners near density limits at various B, Ip, zeff. Also having as many data as possible in the narraw region where the physics behind the limits can vary.
Future modelling should combine turbulent transport and radiation. Trubulent transport SOL-edge codes (GRILLIX, SOLEDGE3X, GBS etc) should be able to investigate the density limit. As well as integrated mdoelling up to the separatrix in L mode at least. The role of fuelling/pellet and turbulent inward particle transport etc could be disentangled. (ERG to be proposed on this D. Fajardo as far as I know)
HZ, CA et al discussion (27/5): nsep vs nped? how do nsep correlates with line average density as well as the pedestal. Likely today the grad(n) btw sep and ped matters for Er and ExB shear. Fuelling vs pellet (for the same throughput, with pellets one gains on vol. integral density). Int modelling could be used to couple pellet and transport and compare gas puff vs transport?
If density goes up with Psep, what is the next limit? MHD? other?
A EUROfusion science meeting could be dedicated to the density limit to reflect the recent progresses on models and the perspectives?