View Issue Details

IDProjectCategoryView StatusLast Update
0000936OpenFOAMBugpublic2013-07-31 14:46
Reporterjherb Assigned Tohenry  
PrioritynormalSeveritymajorReproducibilityalways
Status closedResolutionno change required 
PlatformLinuxOSOpenSuseOS Version11.1
Summary0000936: buoyantPimpleFoam calculates non-physical thick boundary layers and velociities in pipe
DescriptionThe attached simple test case for a (2D) pipe flow simulated with buoyantPimpleFoam results in non-physical thick boundary layers at the walls and therefore in a (much) to high central velocity in the pipe. Using buoyantSimpleFoam to simulate the same case shows results as expected. (Using the stationary result as initialization for the transient calculation does not help.)

(Doubling the number elements in the radial direction does not help either.)
TagsNo tags attached.

Activities

jherb

2013-07-31 00:08

reporter  

jherb

2013-07-31 09:23

reporter  

simple.png (18,858 bytes)   
simple.png (18,858 bytes)   

jherb

2013-07-31 09:23

reporter  

pimple-2.00s.png (26,575 bytes)   
pimple-2.00s.png (26,575 bytes)   

henry

2013-07-31 14:46

manager   ~0002358

There are many issues with the setup of your case, in particular the boundary condicions are not completely consistent and physical which is why you get recirculation from the outlet. Also many solver settings are far from optimal. Even with all the BCs and solver settings optimised it is still a difficult case because the pressure level is VERY large compared to the pressure variation and the density is set to be constant which make the problem very stiff. Nevertheless it is possible to get it running and provide the correct result.

Due to the inconsistent initialisation of p and p_rgh the statement

    // Force p_rgh to be consistent with p
    p_rgh = p - rho*gh;

in createFields.H has a detrimental effect on the start-up of the case and givin the initial field for p_rgh should be uniform it is better not to include this statement. For cases where p is correct and p_rgh is not it is better to include this statement. We will consider better initialisation options.

We can help you correct this case and any others via a support contract.

Issue History

Date Modified Username Field Change
2013-07-31 00:08 jherb New Issue
2013-07-31 00:08 jherb File Added: buoyantPimpleReport-130731.tar.gz
2013-07-31 09:23 jherb File Added: simple.png
2013-07-31 09:23 jherb File Added: pimple-2.00s.png
2013-07-31 14:46 henry Note Added: 0002358
2013-07-31 14:46 henry Status new => closed
2013-07-31 14:46 henry Assigned To => henry
2013-07-31 14:46 henry Resolution open => no change required