[Nek5000-users] Vanishing Jacobians while using moving mesh

Tue Feb 21 06:13:34 CST 2012

Hi Pradeep,

There was one thing that I'd intended to mention regarding mesh motion.

The file that I set up uses the most general ALE formulation,
where the mesh velocity is computed through an elasticity solver.
The performance for this solver has not been optimized, but there
are several tricks we can use to make it faster.   At the moment,
I would guess that well over half the time is spent in this solver,
which could be an issue if you plan to look at 3D problems.  As
you get into production runs we should examine the performance and
optimize if necessary.

Regards,

Paul

On Thu, 16 Feb 2012, nek5000-users at lists.mcs.anl.gov wrote:

> Hi Paul,
>
> Thank you so much for your help!
>
> Regards,
> Pradeep
>
> On Thu, Feb 16, 2012 at 7:28 AM, <nek5000-users at lists.mcs.anl.gov> wrote:
>
>>
>> PS -- an example of the results can be found here:
>>
>>  www.mcs.anl.gov/~fischer/ocyl.**gif<http://www.mcs.anl.gov/%7Efischer/ocyl.gif>
>>
>>
>>
>>
>> On Thu, 16 Feb 2012, nek5000-users at lists.mcs.anl.**gov<nek5000-users at lists.mcs.anl.gov>wrote:
>>
>>
>>> Hi Pradeep,
>>>
>>> Attached is a case that appears to work.  Hope this helps.
>>>
>>> Paul
>>>
>>>
>>> On Wed, 15 Feb 2012, nek5000-users at lists.mcs.anl.**gov<nek5000-users at lists.mcs.anl.gov>wrote:
>>>
>>>  Thanks Paul and Josh,
>>>>
>>>> I shall start by making the modification in the rea file and trying visit
>>>> 2.3.2
>>>>
>>>> Regards,
>>>> Pradeep
>>>>
>>>> On Wed, Feb 15, 2012 at 9:39 AM, <nek5000-users at lists.mcs.anl.**gov<nek5000-users at lists.mcs.anl.gov>>
>>>> wrote:
>>>>
>>>>  Pradeep,
>>>>>
>>>>> Which version of Visit are you using?  We have experienced similar
>>>>> issues with version 2.4 when viewing 2D meshes.  If this is the case,
>>>>> try using 2.3.2 and see if you have the same problem.
>>>>>
>>>>> Josh
>>>>>
>>>>> On Tue, Feb 14, 2012 at 8:08 AM,  <nek5000-users at lists.mcs.anl.**gov<nek5000-users at lists.mcs.anl.gov>>
>>>>> wrote:
>>>>>
>>>>>> Hi Neks,
>>>>>>
>>>>>> I'm trying to simulate 2D flow over an oscillating cylinder using
>>>>>> moving
>>>>>> mesh, and tried to set up a test case by looking at the peris example.
>>>>>> I
>>>>>> however am getting vanishing Jacobians after the first few timesteps.
>>>>>>
>>>>>> I am modifying the the boundary velocity of the cylinder in userbc
>>>>>>
>>>>> instead
>>>>>
>>>>>> of userchk (as in the peris example). Would it be possible for you to
>>>>>>
>>>>> look
>>>>>
>>>>>> at my .usr file and tell me what I am doing wrong?
>>>>>>
>>>>>> I have attached the .rea, .usr, and SIZE with this mail.
>>>>>>
>>>>>> Further, when I try to view the mesh of the output (or the xyz fld
>>>>>> outputfile) in visit, visit crashes with the following error:
>>>>>> VisIt's viewer exited abnormally! Aborting the Graphical User
>>>>>> Interface.
>>>>>>
>>>>>> Lastly, would it be possible to simulate moving walls with large
>>>>>> deformations (i.e. where multiple elements get deformed), where the
>>>>>> deformation is known ahead of time? (eg. a flapping aerofoil)
>>>>>>
>>>>>> Thank you for your help!
>>>>>>
>>>>>> Regards,
>>>>>> Pradeep
>>>>>>
>>>>>> Extract from logfile:
>>>>>>
>>>>>>        60 U-PRES gmres:    100  7.5345E+42  1.0000E-02  4.0062E+57
>>>>>> 6.7556E-03  1.5093E-02
>>>>>>  ERROR:  alphad .le. 0 in econjp -1.31175831444004495E+179          16
>>>>>>  ERROR:  alphad .le. 0 in econjp -1.31175831444004495E+179          16
>>>>>>  ERROR:  alphad .le. 0 in econjp -1.31175831444004495E+179          16
>>>>>>  ERROR:  alphad .le. 0 in econjp -1.31175831444004495E+179          16
>>>>>>          60  DNORM, DIVEX  1.18120867781751263E+043
>>>>>> 7.53448962619089522E+042
>>>>>>         60   6.0000E-02  1.6697E-02 Fluid done
>>>>>>     DT/DTCFL/DTFS/DTINIT   0.635E-45   0.635E-45   0.000E+00
>>>>>> 0.100E-02
>>>>>> Step     61, t= 6.0800000E-02, DT= 8.0000000E-04, C=******* 2.2667E-01
>>>>>> 1.8287E-02
>>>>>>             Solving for fluid
>>>>>>
>>>>>>
>>>>>>    3  ERROR:  Vanishing Jacobian near      2th node of element
>>>>>>
>>>>> 3.
>>>>>
>>>>>>
>>>>>>
>>>>>>    2  ERROR:  Vanishing Jacobian near      2th node of element
>>>>>>
>>>>> 1.
>>>>>
>>>>>>  2.27348348709230109E+081 -5.12478988977734591E+080
>>>>>>  -2.60483379627720531E+079  3.21465902039155941E+039
>>>>>>    3 xyz:  -5.30056E+39  -5.84642E+39
>>>>>>    3 xyz:  -5.24850E+39  -6.93109E+39
>>>>>>
>>>>>>
>>>>>>    3  ERROR:  Vanishing Jacobian near      2th node of element
>>>>>>
>>>>> 4.
>>>>>
>>>>>>    2 xyz:   1.25061E+38   4.64957E+39
>>>>>>  9.79493533554062390E+080 -3.01548938509156252E+080
>>>>>>    2 xyz:  -2.46589E-08  -5.64092E-01
>>>>>>
>>>>>>
>>>>>>    3 xyz:  -5.30056E+39  -5.84642E+39
>>>>>>    3 xyz:  -5.24850E+39  -6.93109E+39
>>>>>>    2  ERROR:  Vanishing Jacobian near      3th node of element
>>>>>>
>>>>> 2.
>>>>>
>>>>>>
>>>>>>
>>>>>>    3  ERROR:  Vanishing Jacobian near      2th node of element
>>>>>>
>>>>> 5.
>>>>>
>>>>>>  1.43125817410005932E+039 -6.03607926692375119E+038
>>>>>>  5.54918579647991880E+039 -1.67302572815141788E+039
>>>>>>    3 xyz:  -5.30056E+39  -5.84642E+39
>>>>>>    2 xyz:  -2.46589E-08  -5.64092E-01
>>>>>>    3 xyz:  -5.24850E+39  -6.93109E+39
>>>>>>
>>>>>>
>>>>>>    3  ERROR:  Vanishing Jacobian near      2th node of element
>>>>>>
>>>>> 6.
>>>>>
>>>>>>    2 xyz:  -3.07794E-08  -7.04112E-01
>>>>>>
>>>>>>
>>>>>>    2  ERROR:  Vanishing Jacobian near      7th node of element
>>>>>>
>>>>> 7.
>>>>>
>>>>>>  8.35923896949543198E+038 -3.55455289325888722E+038
>>>>>>    3 xyz:  -5.30056E+39  -5.84642E+39
>>>>>>  1.21747337192549314E+039 -1.79454406120891328E+038
>>>>>>    3 xyz:  -5.24850E+39  -6.93109E+39
>>>>>>    2 xyz:  -5.66434E-08  -1.29581E+00
>>>>>>    2 xyz:  -6.27639E-08  -1.43583E+00
>>>>>>
>>>>>>
>>>>>>    2  ERROR:  Vanishing Jacobian near      9th node of element
>>>>>>
>>>>> 8.
>>>>>
>>>>>>  1.17802538305091326E+039 -4.21489851051369387E+038
>>>>>>    2 xyz:  -6.55671E-08  -1.49996E+00
>>>>>>    2 xyz:  -3.25016E+38   4.93026E+39
>>>>>>  call outfld: ifpsco: F
>>>>>>
>>>>>>       61  6.0800E-02 Write checkpoint:
>>>>>>
>>>>>>       61  6.0800E-02 OPEN:
>>>>>> xyzcyl2.fld01
>>>>>>  Jac error 1, setting p66=4, ifxyo=t
>>>>>>
>>>>>> call exitt: dying ...
>>>>>>
>>>>>>
>>>>>> ______________________________**_________________
>>>>>> Nek5000-users mailing list
>>>>>> Nek5000-users at lists.mcs.anl.**gov <Nek5000-users at lists.mcs.anl.gov>
>>>>>> https://lists.mcs.anl.gov/**mailman/listinfo/nek5000-users<https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> Josh Camp
>>>>>
>>>>> "All that is necessary for the triumph of evil is that good men do
>>>>> nothing" -- Edmund Burke
>>>>> ______________________________**_________________
>>>>> Nek5000-users mailing list
>>>>> Nek5000-users at lists.mcs.anl.**gov <Nek5000-users at lists.mcs.anl.gov>
>>>>> https://lists.mcs.anl.gov/**mailman/listinfo/nek5000-users<https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users>
>>>>>
>>>>>
>>>  ______________________________**_________________
>> Nek5000-users mailing list
>> Nek5000-users at lists.mcs.anl.**gov <Nek5000-users at lists.mcs.anl.gov>
>> https://lists.mcs.anl.gov/**mailman/listinfo/nek5000-users<https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users>
>>
>