[Nek5000-users] explstrs routine

[nek5000-users at lists.mcs.anl.gov]

Hi Josh,

You need the jacobian at least once, since normally:

    B = w3m1*Jac

Does this make sense?

Paul


On Fri, 29 Jul 2011, nek5000-users at lists.mcs.anl.gov wrote:

> Hello Neks
>
> I came across the routine explstrs (and its subroutines) while
> exploring the code (and with the recommendation from Paul).  I
> understand that it is essentially performing something like
>
>   D^T * B * D * [u]
>
> I had a question about the use of the Jacobian in the routine
> expl_strs_e_3d (or _2d).  Following the routine, I can see where it
>
> 1) Computes the gradients w.r.t. r, s, t   (gradl_rst)
>
> 2) Multiplies these gradients by appropriate coordinate derivatives
> (rxm1, tzm1, etc)
>
> 3) Multiplies these results by the quadrature weights. (from w3m1)
>
> 4) Multiplies these results again by the coordinate transformations,
> and then apply the transpose operation (gradl_rst_t)
>
> When I look in coef.f, I see that when dr/dx, ds/dx, etc. are computed
> from dx/dr, dy/ds, etc., the inverse of the Jacobian is not applied
> yet.  So, I can see why jacmi is multiplied along with w3m1 when
> multiplying by the quadrature weights.
>
> My question is, would we not need to multiply again by jacmi when the
> transpose derivative operator is applied (because of the use of rxm1,
> sxm1, etc)?  Is this something that is applied later on, or am I
> missing something and it isn't necessary to apply it again?
>
> Thank you,
>
> Josh
>
> -- 
> 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
> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
>