syfi team mailing list archive

[Martin Sandve Alnæs <martinal@xxxxxxxxx>]

Jeg gjorde det samme. Merk at:

In [1]: 1.41421 / 0.816497
Out[1]: 1.7320455555868546

In [6]: sqrt(2)/sqrt(3)
Out[6]: 0.81649658092772615

så en vanlig facet som burde ha areal 0.5 og dermed facet_D=1,
har facet_D = sqrt(2/3) i stedet.

Martin



On Tue, Mar 17, 2009 at 7:14 PM, Kent Andre <kent-and@xxxxxxxxx> wrote:
>
> Tror den rette måten å debugge dette på er å stappe inn funksjonen 1 på
> randen på UnitCube(1,1,1).
> Jeg gjorde det, jeg får:
>
>
>  facet 0
>  facet_D 0.816497
> A[0] 0
> A[1] 0.235702
> A[2] 0.235702
> A[3] 0.235702
>  facet 3
>  facet_D 1.41421
> A[0] 0.235702
> A[1] 0.235702
> A[2] 0.235702
> A[3] 0
>  facet 0
>  facet_D 0.816497
> A[0] 0
> A[1] 0.235702
> A[2] 0.235702
> A[3] 0.235702
>  facet 3
>  facet_D 1.41421
> A[0] 0.235702
> A[1] 0.235702
> A[2] 0.235702
> A[3] 0
>  facet 0
>  facet_D 0.816497
> A[0] 0
> A[1] 0.235702
> A[2] 0.235702
> A[3] 0.235702
>  facet 3
>  facet_D 1.41421
> A[0] 0.235702
> A[1] 0.235702
> A[2] 0.235702
> A[3] 0
>  facet 0
>  facet_D 0.816497
> A[0] 0
> A[1] 0.235702
> A[2] 0.235702
> A[3] 0.235702
>  facet 3
>  facet_D 1.41421
> A[0] 0.235702
> A[1] 0.235702
> A[2] 0.235702
> A[3] 0
>  facet 0
>  facet_D 0.816497
> A[0] 0
> A[1] 0.235702
> A[2] 0.235702
> A[3] 0.235702
>  facet 3
>  facet_D 1.41421
> A[0] 0.235702
> A[1] 0.235702
> A[2] 0.235702
> A[3] 0
>  facet 0
>  facet_D 0.816497
> A[0] 0
> A[1] 0.235702
> A[2] 0.235702
> A[3] 0.235702
>  facet 3
>  facet_D 1.41421
> A[0] 0.235702
> A[1] 0.235702
> A[2] 0.235702
> A[3] 0
>
>
> Så det er åpenbart noe galt som ikke er riktig.. :)
>
> Kent
>
>
> On ti., 2009-03-17 at 18:55 +0100, SyFi wrote:
>> One or more new changesets pushed to the primary SyFi repository.
>> A short summary of the last three changesets is included below.
>>
>> changeset:   397:ce67fe86b070a51c5c47afbc9bd971840dab81d9
>> tag:         tip
>> user:        "Martin Sandve Alnæs <martinal@xxxxxxxxx>"
>> date:        Tue Mar 17 18:56:52 2009 +0100
>> files:       examples/newsolvers/Poisson3D/cpp/Makefile examples/newsolvers/Poisson3D/cpp/main.cpp site-packages/newsfc/codegeneration/integralcg.py
>> description:
>> Added debug prints of facet tokens and enabled debug code in Poisson3D demo. Maybe this will speed up the hunt.
>>
>>
>> changeset:   396:851fe8859d29c63105d893d83470e0ee1c8c7434
>> user:        "Martin Sandve Alnæs <martinal@xxxxxxxxx>"
>> date:        Tue Mar 17 15:27:34 2009 +0100
>> files:       examples/newsolvers/Poisson3D/cpp/main.cpp site-packages/newsfc/geometry/UFCCell.py site-packages/newsfc/geometry/mappings.py site-packages/newsfc/representation/exteriorfacetintegralrepresentation.py site-packages/newsfc/representation/formrepresentation.py site-packages/newsfc/symbolic_utils/__init__.py
>> description:
>> Reimplemented facet scaling factor, got lost in the sfc->newsfc transition.
>> Still problems with Poisson3D demo though...
>>
>>
>> changeset:   395:320cfabde141ac32ee0a7bfe08b2c2af06dfa469
>> user:        "Martin Sandve Alnæs <martinal@xxxxxxxxx>"
>> date:        Tue Mar 17 12:01:49 2009 +0100
>> files:       examples/newsolvers/HyperElasticity3D/cpp/main.cpp examples/newsolvers/HyperElasticitySVK/cpp/main.cpp examples/newsolvers/Poisson3D/cpp/Makefile examples/newsolvers/Poisson3D/cpp/Poisson.ufl examples/newsolvers/Poisson3D/cpp/SConstruct examples/newsolvers/Poisson3D/cpp/main.cpp examples/newsolvers/Poisson3D/cpp/manufactured_solution.py
>> description:
>> 3D poisson demo for validation.
>> Think I'm doing something wrong, or there's a bug in boundary integrals.
>>
>>
>> ----------------------------------------------------------------------
>> For more details, visit http://www.fenics.org/hg/syfi
>> _______________________________________________
>> SyFi-dev mailing list
>> SyFi-dev@xxxxxxxxxx
>> http://fenics.org/mailman/listinfo/syfi-dev
>
> _______________________________________________
> SyFi-dev mailing list
> SyFi-dev@xxxxxxxxxx
> http://fenics.org/mailman/listinfo/syfi-dev
>