Skip to content

Fix gradient handling for viscous fluxes #4

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Open
hheinzer wants to merge 2 commits into
base: master
Choose a base branch
from
Open

Fix gradient handling for viscous fluxes #4

hheinzer wants to merge 2 commits into from

Conversation

@hheinzer
Copy link
Collaborator

@hheinzer hheinzer commented Dec 10, 2025

This PR fixes viscous gradient handling in the FV solver:

  • Viscous fluxes use pre-limiter (unlimited) gradients; limiters are kept for inviscid reconstruction
  • At physical boundaries, viscous face gradients are built from the interior gradient and corrected in the face-normal direction, avoiding averaging with ghost gradients

Tested: built and ran BlasiusBoundaryLayer, no regressions observed; the number of Newton and GMRES iterations decreased slightly.

old:

  Data Output at Iteration          100   
  Time since last I/O:   19.64 s          
  Residuals:  rho:    1.0320544812738925E-005
              m1 :    1.3074980703470162E-004 
              m2 :    6.2375215302630703E-005 
              e  :    3.0815566291081406E-005 
------------------------------------------------------------
 Converged in rho - computation complete!                    
 Iteration number:          101           
 Writing final state to disk...           
 ------------------------------------------------------------
 Computation Time :    148.11169911 s                         
 Netwon Iterations:       3303            
 GMRES Iterations:      16515             
 ------------------------------------------------------------

new:

  Data Output at Iteration          100      
  Time since last I/O:   16.78 s             
  Residuals:  rho:    1.0216841725811122E-005
              m1 :    1.2979553044528200E-004               
              m2 :    6.1733762175799389E-005               
              e  :    3.0354989536618631E-005               
------------------------------------------------------------
 Converged in rho - computation complete!                    
 Iteration number:          101                              
 Writing final state to disk...                              
 ------------------------------------------------------------
 Computation Time :    140.49910301 s                        
 Netwon Iterations:       3283                               
 GMRES Iterations:      16415                                                                            
 ------------------------------------------------------------

@hheinzer
Bugfix: use unlimited gradients for viscous fluxes
356811b 
Compute viscous fluxes using pre-limiter gradients so the
stress/heat-flux uses the raw gradient operator. Limiters are intended
for inviscid reconstruction and can unnecessarily degrade diffusive
accuracy.
@hheinzer
Bugfix: boundary-face gradients for viscous fluxes
f3a6462 
At physical boundaries, construct face gradients from the interior cell
gradient and apply the normal correction implied by the boundary/ghost
state. Avoid averaging in ghost gradients, which are not generally
well-defined and can degrade wall stress/heat-flux accuracy.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

1 participant

@hheinzer