Inquiry on what would cause the solution not to align well with the adaptive meshes

[Mark Mulwana]

Hello Team,

Am performing numerical simulations using Adaptive Mesh Refinement, but on plotting the results, I found out that the solution does not align well with the adaptive meshes, I have making changes in my configuration file (maximum levels of refinement, refinement criteria and so on) but nothing is changing.

What could be the cause?? Thank you

DISCLAIMER: The contents of this email and any attachments are confidential. They are intended for the named recipient(s) only. If you have received this email by mistake, please notify the sender immediately and you are herewith notified that the contents are legally privileged and that you do not have permission to disclose the contents to anyone, make copies thereof, retain or distribute or act upon it by any means, electronically, digitally or in print. The views expressed in this communication may be of a personal nature and not be representative of AIMS-Rwanda and/or any of its Initiatives.

[Kyle Mandli]

Hi Mark,

Could you attach the setrun.py you are using and a couple of plots that show the problem?

Kyle

--
You received this message because you are subscribed to the Google Groups "claw-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to claw-users+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/claw-users/a99d52c7-7cd5-4551-8030-0db3e73368e2n%40googlegroups.com.

[Mark Mulwana]

Hi Kyle,

Actually am in forestclaw so I do not have setrun.py file but I have sent my configuration file and some plots

[user]

example = 0

mapping = 1 # 0 = cubedsphere; 1 = pillowsphere

claw-version = 4

[clawpatch]

# Grid dimensions

mx = 8 # mx_leaf

my = 8 # my_leaf

mbc = 2 # Number of ghost cells

maux = 3

# Clawpatch tagging criteria

# value : value exceeds threshold

# minmax : qmax-qmin exceeds threshold

# difference : difference (e.g. dqx = q(i+1,j)-q(i-1,j)) exceeds threshold

# gradient : gradient exceeds threshold

# user : User defined criteria

refinement-criteria = gradient

# Equations and boundary conditions

meqn = 1 # Number of equations

[Options]

# Refinement

minlevel = 3 # Minimum level, from 2

maxlevel = 8 # Maximum levels of refinement, from 7

regrid_interval = 1 # Regrid every 'regrid_interval' time steps.

refine_threshold = 1 # Use gradient as refinement criteria, from 3

coarsen_threshold = 0.8 #0.001 from 1.5

smooth-refine = T

smooth-level = 4 #from 7

# Time stepping

tfinal = 5.0 #

use_fixed_dt = F # Take a fixed time step

initial_dt = 0.02 # Initial time step for 'minlevel'

max_cfl = 1.0 # maximum cfl

desired_cfl = 0.9 # desired cfl

outstyle = 1 # OutputStyle = 1,2,3 (see below, e.g. [OutputStyle_1]

nout = 20 # Used for all three out styles; has different meaning, though.

nstep = 1 # Only used if outstyle is 3

outstyle-uses-maxlevel = F

subcycle = T # subcycle in time ( = T or F).

# Console and file IO

verbosity = production # verbose

output = T # Set this to F for parallel runs. T is dead slow.

# Debugging adn diagnostics

trapfpe = T # Trap floating point exceptions

mpi_debug = F

conservation-check = F

run-user-diagnostics = F

compute-error = F

report-timing = T

report-timing-verbosity=wall

# Mapping

manifold = T # Manifold

theta = 0 # Rotation angles (degrees)

phi = 0 # Rotation angle

scale = 1 1 1

shift = 0 0 0

[clawpack46]

order = 2 2 # normal and transverse order

mcapa = 1 # mcapa

src_term = 0 # src_term

mwaves = 1 # mwaves

mthlim = 4 # mthlim (is a vector in general, with 'mwaves' entries)

mthbc = 3 3 3 3 # mthbc (=left,right,bottom,top) from 1 1 1 1 to 2 2 2 2

# output

ascii-out = T

vtk-out = F

[clawpack5]

order = 2 2 # normal and transverse order

mcapa = 1 # mcapa

src_term = 0 # src_term

mwaves = 1 # mwaves

mthlim = 4 # mthlim (is a vector in general, with 'mwaves' entries)

mthbc = 3 3 3 3 # mthbc (=left,right,bottom,top), from 1 1 1 1 to 2 2 2 2

# output

ascii-out = T

vtk-out = F

[Mark Mulwana]

Hello Kyle

You asked me to send my setrun.py file and some plots that are showing how the solution is not aligning well with the adaptive meshes which I did, but haven't got any feedback from you, am still waiting for the feedback. Thank you

[Kyle Mandli]

Hi Mark,

Unfortunately I am not an expert in ForestClaw and do not have a good idea as to what may be going wrong.  Hopefully someone with more knowledge can respond.

Kyle

To view this discussion visit https://groups.google.com/d/msgid/claw-users/2332b5b3-9a50-4a46-903e-32bd183732d9n%40googlegroups.com.

[donnac...@boisestate.edu]

Hi Mark, 

It is clear that the solution and the meshes are correct.  But there is likely something wrong with your cubed-sphere mapping in Visclaw.  Maybe you the block numbers are not what you think they are.   The issue isn't a ForestClaw issue, and also not a VisClaw issue (since Visclaw doesn't support the cubed-sphere).   I would try starting with a very simple initial condition (maybe set half the sphere to 0, half to 1) and see if you can get the lined up properly.  My guess is that there is a minor bug somewhere in your Python cubed-sphere mapping.  

[Mark Mulwana]

Thank you so much for your reply, Donna and Kyle

To view this discussion visit https://groups.google.com/d/msgid/claw-users/01609f25-01a8-4d17-9189-86103088cca2n%40googlegroups.com.