[FLASH-USERS] Negative 3T Internal Energy when including Nernst

[Hansen, Eddie]

Hi Edouardo,

We have verified that the Nernst term in FLASH works accurately for a semi-analytical problem, but we have just started using this term to model real-world problems. We’re actively investigating which settings work best, and we’ll be able to provide more guidance as we learn more.

If it helps, there are a couple of terms associated with turning on Nernst that your problem may be able to safely ignore…


  1.  Ettingshausen – the qx, ay, and qz heat flux terms in hy_uhd_addNernstFluxes.F90
  2.  The j.E term associated with Nernst in hy_uhd_getFaceFlux.F90, which would require commenting blocks such as lines 675-694 in that file.

It’s possible that turning off one or both of these will avoid the error.

--
Eddie Hansen
Applications Group Leader
Flash Center for Computational Science


From: flash-users <flash-users-bounces at flash.rochester.edu> on behalf of Rover, Edoardo <erover at UCSD.EDU>
Date: Monday, January 27, 2025 at 12:36 PM
To: flash-users at flash.rochester.edu <flash-users at flash.rochester.edu>
Subject: [FLASH-USERS] Negative 3T Internal Energy when including Nernst
Dear FLASH users,

I am attempting to implement extended-MHD in my simulation, and started by enabling the Nernst term in the simulation. However, I get a "Negative 3T Internal Energy" error. Following a previous user email, I attempted to enable the flux limiter options for Nernst, and set them to:
nernstFlMode = "fl_minmax"
nernstFlCoef = 0.08
But the error presented itself once more. I previously attempted to use the "fl_larsen" option as well, and set the flux limiter to "0.09", the same value we are using for the electron conduction flux limiter, but the problem persists.

Setting "use_Nernst" to ".false." makes the simulation run without problems. Preheating the materials or changing the diffusion dt factor (or the radiation transport one) shows no changes, the simulation runs normally if Nernst is set to False, while it crashes if set to True.

For that reason, I think the issue lies entirely in the MHD setup block. Would you have any suggestions on the issue?

Best regards,

Edoardo Rovere

Postdoctoral researcher
UCSD CER
3183 Matthews Ln 92093
La Jolla (CA), US
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20250127/b052f63b/attachment.htm>