search

nufeb / NUFEB

Official release of the NUFEB software
http://research.ncl.ac.uk/nufeb/
GNU General Public License v3.0
18 stars 16 forks source link

example biofilm-het, unexplained behaviour for larger system #4

Closed wmoebius closed 4 years ago

wmoebius commented 4 years ago

Running the example biofilm-het (commit 1574b31403ff1423b99de00f414171d2ed64010d) with slightly modified parameters:

Inputscript-vtk.lammps (modified lines only):

variable EPSratio equal 1.2
fix k1 all kinetics 100 25 10 50 v_diffT v_layer niter 5000
run 500000

atom.in (modified line only):

0.000000e-04   2e-04  zlo zhi

I observe something like an soft 'barrier' at half system height. Part of the biofilm seems to grow beyond that point, part of it seems to be hinded by this 'barrier'. I do not have a biological explanation, which is why I am reporting this as an issue.

shelllbw commented 4 years ago

Hi Wolfram, the upper and bottom surfaces are bounded with frictional and adhesive walls. When changing domain height, you also need to change the height of these two walls with your new zhi in fix wall/gran hooke/history and fix walladh commands (the last two parameters).

Bowen

wmoebius commented 4 years ago

Thanks, I had not realised this. Could I request the feature to set system size at one place and propagate it to the other relevant parameters?

wmoebius commented 4 years ago

I can run a large system now. In Paraview, the colouring by type seems to break down for this larger system a while into the simulation. Is that something you are familiar with?

shelllbw commented 4 years ago

Thanks, I had not realised this. Could I request the feature to set system size at one place and propagate it to the other relevant parameters?

They are actually LAMMPS commands. Allowing such flexibility would be useful in some systems where you want to define additional obstacle walls in the domain.

shelllbw commented 4 years ago

I can run a large system now. In Paraview, the colouring by type seems to break down for this larger system a while into the simulation. Is that something you are familiar with?

Haven't met this issue before, are you using Glyph to represent particles (tutorial, visualisation part), also have a try with Point Gaussian.

valleymouth commented 4 years ago

Maybe a screenshot would help?

wmoebius commented 4 years ago

I believe this was an artifact when using Paraview. Closing now and reopening if it becomes a problem again.