[petsc-users] Question regrading SNES tutorial ex28

[Jed Brown]

Gautam Bisht <gbisht at lbl.gov> writes:

> Hi PETSc team,
>
> Any insights into why the residual equation for U & K are scaled by dx in
> this example (line no. 48-49 and 68 of in ex28.c
> <https://bitbucket.org/petsc/petsc/src/8db46702c31aaef455dabea2fa9d7dd897381b5d/src/snes/examples/tutorials/ex28.c?at=master#cl-8>
> )?

The scaling by hx represents a volume integral, similar to what you
naturally get with finite element methods.  It helps keep the problem
well-scaled during grid refinement.

> Additionally, the comment regarding algebraic equation is incorrect.
>
>  * Algebraic (K):
>
>  *     exp(k-1) + k = u + 1/(1/(1+u) + 1/(1+u_x^2))"
>
>
> should be changed to
>
>  * Algebraic (K):
>
>  *     exp(k-1) + k = 1/(1/(1+u) + 1/(1+u_x^2))"

Thanks; I'll push this shortly.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 818 bytes
Desc: not available
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20150513/192a2f10/attachment.pgp>