Minutes of the WIMAS-2 Weekly Meeting, 22 April 2020
The meeting included three main parts and there were questions and comments to be investigated in more detail on each of them as summaries below.
Introduction: Active branches
When specifying the boundary condition as a gradient, does the equation have a unique solution?
ACTION:
- "someone": Please clarify.
- Dave (and Jorge): There's a unique solution. But Monte Carlo can have a big effect on the solution; cause bifurcations. Thus, the solution is unreliable. In practice, numerical solutions will either go to zero or infinity.
- Note: Supplying decay length (n/n'=... , T/T'=... ; Robin style) is stable.
- Follow up: Do we want to support derivative boundary conditions?
- Dave: Support it, but be careful.
FEM solver: New developments
In the new formulation, do the differential weak form, including the boundary conditions, satisfy conservation properties?
ACTION:
- ANNA: Could you clarify this?
Transport equations for the IDSs and the ETS
Here we had a long and very interesting discussion with many questions and resulting actions.
First, it was pointed out by Sun Hee Kim that also the equation for the current evolution need clarification. The same goes for the momentum equation.
ACTIONS:
- THOMAS: Schedule a presentation and discussion on the current equation; do we want solve for psi, psi', q...?
- THOMAS: Schedule a presentation and discussion on the momentum equation.
The discussion on transport coefficients for particle and energy conservation were based on a lack of clarity in the definitions of the transport coefficients in the core-transport IDS. There are different options for what convention to use. Pär proposed the convention used in TCI and JINTRACK, while Alexei preferred the one used in ASTRA and the ETS5. The two conventions differ in the convection by a factor <|rho|^2>/<|rho|>. At present the ETS include a mix of these two conventions as TCI and the Transport-solver use different definitions.
In order to device an action plan we need to consider two options:
- We waiting for ITER to decide on the definitions of the transport coefficients and then implement the decision.
- We cannot wait since the ETS6 is presently using two different conventions in the transport solver and in TCI. Thus, we decide to use the TCI convention everywhere - in the transport-solver, transport-combiner, ids2cpo, cpo2ids etc. Once ITER takes a decision, we may need to rewrite our implementation again.
DECISION: After discussing with a few people, the ETS-6 will be adapting to the TCI convention until a decision is reached on IMAS-3039.
If option 1 is chosen all actions below will be waiting until we have a decision from ITER and all we should do now is to add this to our “to-do” list (perhaps a @todo in our source code if this is traces in e.g doxygen).
ACTIONS:
- PÄR and THOMAS: Write tracker on clarifying the definitions of transport coefficients in the core_transport. This tracker could also include adding fields to handle the conversion between different conventions for defining transport coefficients, specifically V’, <|rho|^2> and <|rho|>.
- Completed 23 April, see https://jira.iter.org/browse/IMAS-3039
- THOMAS: Make sure the transport-solver follows the ITER (or temporary WIMAS-2) definition of the convection velocity.
- Implemented in branch pinchVelocity and pinchVelocityComments in ets_core. Testing is ongoing.
- PÄR or THOMAS: Make sure the transport-combiner follows the ITER (or temporary WIMAS-2) definition of the convection velocity when outputting fluxes.
- Ongoing, see branch transport_combiner in ets_core.
- PÄR: Make sure that the IDS output from TCI follows the follows the ITER (or temporary WIMAS-2) definition of the convection velocity.
- PÄR: Document the present convention for the transport coefficients in the ETS5.
- Document added under "Transport equation for IDSs and in ETS", see https://indico.euro-fusion.org/event/295/contributions/1181/attachments/195/447/transport_coefficients_ETSv5.pdf.
- JORGE: Make sure that ids2cpo follows the follows the ITER (or temporary WIMAS-2) definition of the convection velocity.
- It has been confirmed that no modification required.
- NATHAN: Make sure that cpo2ids follows the follows the ITER (or temporary WIMAS-2) definition of the convection velocity.
- It has been confirmed that no modification required.
There are minutes attached to this event.
Show them.
