WIMAS-2 Weekly Meeting

Wednesday 13 May 2020, 10:00 → 11:10 Europe/Berlin

https://kth-se.zoom.us/j/7764082514

Thomas Jonsson

WIMAS-2 Weekly Meeting are technical meetings to plan and coordinate the development of the ETS-6 workflow.

 

Remote participation:

https://kth-se.zoom.us/j/7764082514

Minutes from the Weekly WIMAS-2 Meeting, 13 May 2020

1. Actions from last meeting

See the tab Actions from WIMAS-2 Meetings in:
   https://docs.google.com/spreadsheets/d/1Ps-cGQyKNsYTObCr5Kqu58bXBNQdH45Fjpz5QD0hLB0/edit?usp=sharing

Comments:

Alexei: The question on how to handle non-quasi-neutral sources in a transport solver is a non-trivial problem that require more careful investigations. 

Dave/Thomas: This is not a top priority this year.

2. FEM Solver

Alexei had a question on the conservation properties of the weak formulation. Anna will try to formulate an answer (this was formulated as an action 2 weeks ago).

Pär had comments on how to factorise the transport equations to avoid singularities. The equations in the ETS are already factorised.

ACTION:

1. Anna/Thomas: Add Anna's documentation to the ETS documentation.

3. Transferring ETS-5 developments to the ETS-6

3.1. Hyper-diffusion

Dave commented that we may consider evaluating the advection term using the previous iteration instead of the previous time-step.

ACTION:

1. Unknown: Test hyper-diffusion in the ETS-5 for various solutions.
2. David Coster (to be confirmed): Build a standalone test bench for hyperdiffusion.
3. Team: Design hyper-diffusion for the ETS-6.
4. Team: Implement hyper-diffusion for the ETS-6.

3.2. Internal boundary conditions and pedestal predictions

Question: Could we make use of lessons learned from JINTRAC and ASTRA?

Answer: We seems to have fairly good understanding of how it's been implemented in JINTRAC. Dave also has a report on this. The strategy forward is to test our envisaged solution in ETS-5 and contact the JINTRAC team if needed.

Rui: The solution using a neural network problem cannot be used with pellets.

Dave: It is now important to follow up on Dmitriy's presentation.

ACTION:

1. Team: Design internal boundary conditions for the ETS-6.
2. Team: Implement internal boundary conditions for the ETS-6.
3. Team: Implement Andreas neural network model for the edge in the ETS-6.

4. Any other business

All critical people needed for the WIMAS-2 Sprint 18-19 May 2020 are available. As it's a short sprint we'll focus on WIMAS-2.

Everyone should start preparing for 2020 and their answer to the TSVV call.

1 Actions from last meeting

Minutes from previous meeting:
https://indico.euro-fusion.org/event/295/

Speaker: Thomas Jonsson

2 Active branches

Here we will go through and discuss the ongoing developments on different branches in the ets-core repository.

Speaker: Thomas Jonsson

New active branches

None!

Old active branches

• pinchVelocityComments - need more testing and updates to diagnostic-scripts (from Dmitriy).
• IPramp (Jorge)
• revise_impurity (Denis)
• transport_combiner - work just started.

Merged branches

• negativeValues - as agreed two weeks ago.
  • Note: We now abort if the quasi-neutrality condition generates negative (or too small) densities.
  • Note: We abort if the input to the transport_solver include negative (or too small) densities or temperatures. 
• convertSources
  • Fix error in unit-conversion of transport-solver sources output.
  • Use identifiers for core_sources and core_transport output.
• solverFEM2 - integration of new solver from Anna, see next presentation.
  • Testing for Dirichlet BCs are very good!
  • For gradient BDs (psi-equation) seems to require more work. So don't use solverFEM2 for the psi-equation right now!
  • The option to select this solver is not in the ETS6rc2.xml workflow.

3 FEM Solver

Speaker: Anna Šušnjara (RBI)

TD_FEM - transport equations-only Alternativ formulation.pptx

4 ETS-5 features useful for ETS-6

Speaker: Dmitriy Yadkin (Chalmers)

ets_5_features_for_ets6.pdf

5 Any other business

1. need to plan next week's virtual sprint
2. call for TSVV and ACH's now out -- need to think about this ...