[CIG-SHORT] RuntimeError: Determinant of Jacobian (1.74667e-08) for cell 0 is smaller than minimum permissible value (1e-06)!

[Lucas Abraham Willemsen]

Hello Brad,

Thanks for the feedback. Surendra's help and Matt's comment helped me forward already. In response to your suggestions, I think a warning that the nondimensionalization values may be off would be helpful in pointing the user in the right direction.

Lucas
________________________________________
From: cig-short-bounces at geodynamics.org [cig-short-bounces at geodynamics.org] on behalf of Brad Aagaard [baagaard at usgs.gov]
Sent: Monday, March 18, 2013 11:58
To: cig-short at geodynamics.org
Subject: Re: [CIG-SHORT] RuntimeError: Determinant of Jacobian (1.74667e-08) for cell 0 is smaller than minimum permissible value (1e-06)!

Lucas,

Sorry my reply yesterday afternoon didn't go through.

This is an issue that users run into from time to time when the
nondimensionalization scales are off, in this case the length scale.

We will think about some ways we can try to check for consistency in the
parameters and user input to detect poor choices for the
nondimensionalization scales. We will probably end of giving warnings
rather than errors because, as Matt noted, hard cutoffs are fragile.

Regards,
Brad


On 3/17/13 7:05 PM, Matthew Knepley wrote:
> On Sun, Mar 17, 2013 at 10:20 AM, <surendra at caltech.edu
> <mailto:surendra at caltech.edu>> wrote:
>
>
>     On Mar 16, 2013, at 8:37 PM, Lucas Abraham Willemsen wrote:
>
>>
>>     RuntimeError: Determinant of Jacobian (1.74667e-08) for cell 0 is
>>     smaller than minimum permissible value (1e-06)!
>>
>>     This message doesn't really help me a lot. I have no idea why the
>>     Jacobian is bad, when the mesh elements seem to have such a nice
>>     aspect ratio.
>>
>
>     Try changing your non-dimenionalizing parameters.
>
>           normalizer.wave_period = 0.1*s
>
>     Changing the wave period normalizer to 0.1 sec got rid of that error
>     for me.  There are plenty of other setting that aren't set properly,
>     though.
>
>
> Yes, we have a check on the absolute volume of the element, so changing
> the normalization will get
> rid of this error for well-conditioned meshes. Alternatively, we could
> have some idea of the "normal"
> cell size, but this sounds really fragile.
>
>    Thanks,
>
>       Matt
>
>     --Surendra
>
>     _______________________________________________
>     CIG-SHORT mailing list
>     CIG-SHORT at geodynamics.org <mailto:CIG-SHORT at geodynamics.org>
>     http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>
>
>
>
> --
> What most experimenters take for granted before they begin their
> experiments is infinitely more interesting than any results to which
> their experiments lead.
> -- Norbert Wiener
>
>
> _______________________________________________
> CIG-SHORT mailing list
> CIG-SHORT at geodynamics.org
> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>

_______________________________________________
CIG-SHORT mailing list
CIG-SHORT at geodynamics.org
http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short