View Issue Details
|ID||Project||Category||View Status||Date Submitted||Last Update|
|0002824||OpenFOAM||Bug||public||2018-02-05 08:33||2018-02-07 15:39|
|Platform||Unix||OS||Windows Sub (Ubuntu)||OS Version||16.04.3 LTS|
|Fixed in Version||dev|
|Summary||0002824: violation of mass conservation in compressibleInterDyMFoam|
|Description||Comparing a simple case, where the system is closed with just a piston compressing the gas inside, shows wrong behavior in pressure and density. They both do not rise as expected due to mass loss in the system. With rhoPimpleDyMFoam I do get the correct behavior.|
|Steps To Reproduce||Running the attached case, which has been adapted from a previous issue (https://bugs.openfoam.org/view.php?id=1110)|
on the same topic shows the mentioned behavior. The case has been slightly changed, so that the whole system is closed, hence no inlet/outlet, just a piston compressing the air inside the volume.
Pressure and density do not rise as expected due to mass loss in the system.
|Tags||No tags attached.|
henry_2.0.zip (119,222 bytes)
||Have you tested OpenFOAM-dev?|
Thanks for the prompt answer. No, I didn't check it with the -dev version. Have there been any recent updates regarding this solver?
Yes there have been changes which might affect this behavior. In anycase all tests and changes will be made to OpenFOAM-dev so it is necessary to first test this version.
> shows wrong behavior in pressure and density.
Could you provide more details? What is wrong about the behavior in pressure and density?
||When I evaluate the System after runtime I should get a density of around 1.2kg/m3, from which the solution is still far away with around 1.7kg/m3. This is due to mass loss observed by integrating the density. I wrote this issue here because I couldn't find any problems regarding convergence. Pressure seems to converge fine.|
||I am surprised that mass loss is causing an increase in density, could you clarify?|
||Nono. To clarify. I observe that density as well as pressure does not increase as expected. I mean the setup can be hand calculated using isotropic compression. Reason for this behavior is that mass conservations is violated by losing mass during the process of compressing. I do just not find a reason - at least not by checking the setup and all the solver settings - why there is this mass loss.|
||Oh. the number is 1.17kg/m3.|
Note that compressibleInterDyMFoam operates a corrected incompressible algorithms in order to handle the large phase density differences in two-phase sytems. This means that mass conservation is only obtained by converging the PIMPLE loop sufficiently, proved it does converge. Try tightening tolerances and running more PIMPLE iterations to see if this reduces the convervation error.
rhoPimpleDyMFoam runs a mass conservative compressible algorithm which is appropriate for single-phase, and as such converves mass simply by converging the pressure solution irrespective of convergence of the PIMPLE loop.
||Thanks for the insights. This makes things clearer. Unfortunately I had no success with tightening tolerances nor highering PIMPLE interations.|
||I have isolated the problem and will investigate further.|
||Resolved by commit 852726c25fbe6b4bd4e947bec7ff32bc92bb103e|
|2018-02-05 08:33||arohner||New Issue|
|2018-02-05 08:33||arohner||File Added: henry_2.0.zip|
|2018-02-05 08:52||henry||Note Added: 0009252|
|2018-02-05 09:04||arohner||Note Added: 0009253|
|2018-02-05 09:13||henry||Note Added: 0009254|
|2018-02-05 13:04||arohner||Note Added: 0009255|
|2018-02-05 13:37||henry||Note Added: 0009256|
|2018-02-05 13:53||arohner||Note Added: 0009257|
|2018-02-05 13:54||arohner||Note Added: 0009258|
|2018-02-05 14:05||henry||Note Added: 0009259|
|2018-02-05 14:56||arohner||Note Added: 0009260|
|2018-02-05 15:29||henry||Note Added: 0009261|
|2018-02-07 15:39||henry||Assigned To||=> henry|
|2018-02-07 15:39||henry||Status||new => resolved|
|2018-02-07 15:39||henry||Resolution||open => fixed|
|2018-02-07 15:39||henry||Fixed in Version||=> dev|
|2018-02-07 15:39||henry||Note Added: 0009275|