View Issue Details

IDProjectCategoryView StatusLast Update
0001430OpenFOAMBugpublic2014-12-15 22:42
Reporteruser1018Assigned Tohenry  
PrioritynormalSeverityminorReproducibilityalways
Status closedResolutionno change required 
PlatformUnixOSOtherOS Version(please specify)
Summary0001430: zeroGradient condition on velocity makes buoyantBoussinesqPimpleFoam missbehave
DescriptionImposing a zeroGradient condition to the velocity in the example hotRoom, makes the temperature (T) increasing indefinitely in one corner of the mesh.
It seems to happen only to one node of the domain.
The temperature and velocity BC should be independent, I don't see any mathematical explanation but a bug in the code.
Steps To Reproducein the hotRoom example, change in the file 0/U:

    ceiling
    {
        // type fixedValue;
        // value uniform (0 0 0);
      type zeroGradient;
    }
run the example (maybe export results more often to see what is going on)
TagsNo tags attached.

Activities

alexeym

2014-11-14 15:16

reporter  

hotRoom.tar.gz (2,860 bytes)

alexeym

2014-11-14 15:20

reporter   ~0003289

1. Why change BC only for velocity? I.e. if you modify just velocity, you create inconsistency in BCs.
2. Settings of the tutorial case are in fact taken from steady state case, so you need to modify then for transient case.
3. I've attached modified case files, with them it runs till ~23 seconds after that turbulence starts to develop (due to zeroGradient at the ceiling) and I don't think the case mesh is appropriate for the turbulent flow.

henry

2014-12-15 22:42

manager   ~0003319

zeroGradient for U at a wall is incorrect.

Issue History

Date Modified Username Field Change
2014-10-29 13:57 user1018 New Issue
2014-11-14 15:16 alexeym File Added: hotRoom.tar.gz
2014-11-14 15:20 alexeym Note Added: 0003289
2014-12-15 22:42 henry Note Added: 0003319
2014-12-15 22:42 henry Status new => closed
2014-12-15 22:42 henry Assigned To => henry
2014-12-15 22:42 henry Resolution open => no change required