[Nek5000-users] Data Structure

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

Hi Shriram,

Your interpretation of the layout is correct.  For the most part,
the geometry arrays of interest are

       xm1(i,j,k,e)
       ym1(i,j,k,e)
       zm1(i,j,k,e)

which give the (x,y,z) coordinates for every element e=1,...,nelt
on processor p, p=1,...,P.

(Note that there is no index to reflect "p" --- that index is 
defined implicitly by the distributed memory model; each processor
has it's own unique set of data:  given i,j,k,e,p, one can find
the associated (x,y,z) point.   The inverse problem of finding
i,j,k,e,p for a given (x,y,z) value is much more difficult (and
expensive), but we have routines to handle this.)

Your picture is essentially correct, save that it is not requisite
that the element be oriented with the x-y axis.  It could just as
well have been rotated by 45 or 90 degrees, say.   We retain separate 
arrays for the geometry in order to support curvilinear elements
having deformation and rotation that can accommodate complex shapes.
(A detailed description can be found in Deville, Fischer, Mund,
"High-Order Methods for Fluid Flow", which might be in your library.
I'm hoping to get additional description in a forthcoming reference manual.)

Paul






On Sun, 18 Apr 2010, nek5000-users at lists.mcs.anl.gov wrote:

> Hello,
>
> I am trying to understand the structure of the arrays that store velocities
> and co-ordinates in nek that is available under several subroutines like
> uskchk,usrdat2,usrdat etc.
> Based on my understanding, I found that the velocities, say vx has the form
> vx(lx1,ly1,lz1,e) but I found that there are many arrays that address the
> co-ordinates namely xm0,xm1,xc etc though I am not sure what it denotes. Can
> all these be accessed from any subroutine in the usr file ? I have attached
> herewith a picture where I have mentioned how I thought the data structure
> might be. But, I would really like to know how one would access the
> co-ordinates of say ,a particular element and what xm0,xm1,xc denote for.
> Please correct me if I am wrong about the data structure. Thanks.
>
> Regards
> Shriram Jagannathan
>