[Nek5000-users] extracting data in a smaller computational domain and in a plane

Thu Mar 3 12:39:00 CST 2016

SL,

I have developed a set of routines for this express purpose which I am
happy to share. They are modified versions of the hpts routines that dump
out *fld files of subdomains at arbitrary resolutions.  The routines (and a
few others) can be found at the following link in the file PhilFunc.f:

https://github.com/psakievich/Nek5000

To use you must define your own elements that follow the same format as nek
(npnts_in_x=npnts_in_y=npnts_in_z) and place them in an ascii file named "
ps_hpts.in".  Then in nek call ps_hpts("MYF") where "MYF" is the three
character prefix that will be added to the fld files.

The format of "ps_hpts.in" is also a little different.  The first line
should be:
NPoints, NElm, NPX, NPY, NPZ

where NElm is the number of elements you are creating, and NPX is the
points in the x direction for the given element (akin to lx1), etc. After
that just specify the x,y,z coordinates like the normal hpts.in files.

A couple of additional points:

1) You can crank up the resolution, and use uniform grid spacing to get
much better iso surfaces inside the subvolume you define. (Very nice way to
get high quality visualizations that harness the power of SEM without
interpolating the whole gird onto a finer grid and creating huge data sets)

2) You need to ensure enough memory is allocated for the number of hpts you
are using through the SIZE file like normal.

3) You can use a resolution much higher than the lx1 specified in nek.
However, if you go too high then it may create a seg fault.  I haven't been
able to track down the exact buffer sizes, but they are dependent on LELT,
and LX1 in your size file.  For example, I was to dump out elements of with
32 points in each direction when my lx1 was set to 10 in the SIZE file
(LELT around 1000 I think?). However, when I went higher I received a seg
fault.  I hope to track this down at some point...

4)  If you want the elements to meet each other in VisIT you must specify
redundant boundaries between them (just like in nek) because I am using the
Nek5000 native output routines.  If the elements don't have redundant
boundaries then there will be gaps between them. I would love to get around
this but I haven't had time to write a routine for a new output format.

5) I have only tested these routines in 3D, and I make no promises that
they are bug free. Theoretically, they should work for 2D as well, but I
haven't tried that.

6) DO NOT use the normal hpts/intpnts routines while using the ps_hpts
routines.  They use the same variables and it will create undefined
behavior.

Cheers,

Phil

On Thu, Mar 3, 2016 at 11:00 AM, <nek5000-users-request at lists.mcs.anl.gov>
wrote:

> Send Nek5000-users mailing list submissions to
>         nek5000-users at lists.mcs.anl.gov
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
> or, via email, send a message with subject or body 'help' to
>         nek5000-users-request at lists.mcs.anl.gov
>
> You can reach the person managing the list at
>         nek5000-users-owner at lists.mcs.anl.gov
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of Nek5000-users digest..."
>
>
> Today's Topics:
>
>    1. extracting data in a smaller computational domain and in a
>       plane (nek5000-users at lists.mcs.anl.gov)
>    2. Re: extracting data in a smaller computational domain     and in
>       a plane (nek5000-users at lists.mcs.anl.gov)
>    3. Re: extracting data in a smaller computational domain     and in
>       a plane (nek5000-users at lists.mcs.anl.gov)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Thu, 03 Mar 2016 04:29:05 +0100
> From: nek5000-users at lists.mcs.anl.gov
> To: Nek5000 Users <nek5000-users at lists.mcs.anl.gov>
> Subject: [Nek5000-users] extracting data in a smaller computational
>         domain  and in a plane
> Message-ID:
>         <mailman.16887.1456975753.24727.nek5000-users at lists.mcs.anl.gov>
> Content-Type: text/plain; charset=UTF-8; format=flowed
>
> Hi Neks,
>
> I am trying to do some postprocessing of the Nek5000 results.
>
> I have a 3D domain, however I will be interested in taking a smaller
> piece of this domain. The idea is to generate a data base of snapshots
> containing this smaller part of the domain. For example,
>
> if my computational domain is x in (-15, 20), y in (-20, 20), z in(0,2)
>
> I just want to take a data set of snapshots containing the velocity
> fields in the locations:
>    x in (-2, 5), y in (-5, 5), z in(0,2)
>
> Is there any way to extract this data?
>
> In addition, is there any way to extract the same data but in a 2d
> plane? I mean:
>    x in (-2, 5), y in (-5, 5), z=1
>
> Thank you in advance.
> Cheers
> SL
>
>
> ------------------------------
>
> Message: 2
> Date: Thu, 3 Mar 2016 03:58:48 +0000
> From: nek5000-users at lists.mcs.anl.gov
> To: "nek5000-users at lists.mcs.anl.gov"
>         <nek5000-users at lists.mcs.anl.gov>
> Subject: Re: [Nek5000-users] extracting data in a smaller
>         computational domain    and in a plane
> Message-ID:
>         <mailman.16888.1456977536.24727.nek5000-users at lists.mcs.anl.gov>
> Content-Type: text/plain; charset="us-ascii"
>
> Hi SL,
>
> The easiest way for a profile is to do the Lineout in VisIt and choose
> 'curve' format in 'Set Save Options'.
>
> More accurate way is to setup history/coordinate points in hpts.in file
> and call hpts routine from userchk which will also work for 2D plane points.
>
> Aleks
> ________________________________________
> From: nek5000-users-bounces at lists.mcs.anl.gov [
> nek5000-users-bounces at lists.mcs.anl.gov] on behalf of
> nek5000-users at lists.mcs.anl.gov [nek5000-users at lists.mcs.anl.gov]
> Sent: Wednesday, March 02, 2016 9:29 PM
> To: Nek5000 Users
> Subject: [Nek5000-users] extracting data in a smaller computational
> domain      and in a plane
>
> Hi Neks,
>
> I am trying to do some postprocessing of the Nek5000 results.
>
> I have a 3D domain, however I will be interested in taking a smaller
> piece of this domain. The idea is to generate a data base of snapshots
> containing this smaller part of the domain. For example,
>
> if my computational domain is x in (-15, 20), y in (-20, 20), z in(0,2)
>
> I just want to take a data set of snapshots containing the velocity
> fields in the locations:
>    x in (-2, 5), y in (-5, 5), z in(0,2)
>
> Is there any way to extract this data?
>
> In addition, is there any way to extract the same data but in a 2d
> plane? I mean:
>    x in (-2, 5), y in (-5, 5), z=1
>
> Thank you in advance.
> Cheers
> SL
> _______________________________________________
> Nek5000-users mailing list
> Nek5000-users at lists.mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
>
>
> ------------------------------
>
> Message: 3
> Date: Thu, 03 Mar 2016 05:26:16 +0100
> From: nek5000-users at lists.mcs.anl.gov
> To: nek5000-users at lists.mcs.anl.gov
> Subject: Re: [Nek5000-users] extracting data in a smaller
>         computational domain    and in a plane
> Message-ID:
>         <mailman.16889.1456979182.24727.nek5000-users at lists.mcs.anl.gov>
> Content-Type: text/plain; charset=UTF-8; format=flowed
>
> Thank you Aleks,
>
> that is also very useful. However, it would be nice if I could extract a
> smaller 3d domain from my big computational domain. The idea is to work
> with the entire 3d small computational domain...
>
> I need to do the same for several time steps (velocity snapshots,
> changing in time), so if it would be any way to do it directly from
> Nek5000 it would make the things much easier...
>
> Thanks again.
> SL
>
> El 03-03-2016 04:58, nek5000-users at lists.mcs.anl.gov escribi?:
> > Hi SL,
> >
> > The easiest way for a profile is to do the Lineout in VisIt and choose
> > 'curve' format in 'Set Save Options'.
> >
> > More accurate way is to setup history/coordinate points in hpts.in
> > file and call hpts routine from userchk which will also work for 2D
> > plane points.
> >
> > Aleks
> > ________________________________________
> > From: nek5000-users-bounces at lists.mcs.anl.gov
> > [nek5000-users-bounces at lists.mcs.anl.gov] on behalf of
> > nek5000-users at lists.mcs.anl.gov [nek5000-users at lists.mcs.anl.gov]
> > Sent: Wednesday, March 02, 2016 9:29 PM
> > To: Nek5000 Users
> > Subject: [Nek5000-users] extracting data in a smaller computational
> > domain      and in a plane
> >
> > Hi Neks,
> >
> > I am trying to do some postprocessing of the Nek5000 results.
> >
> > I have a 3D domain, however I will be interested in taking a smaller
> > piece of this domain. The idea is to generate a data base of snapshots
> > containing this smaller part of the domain. For example,
> >
> > if my computational domain is x in (-15, 20), y in (-20, 20), z in(0,2)
> >
> > I just want to take a data set of snapshots containing the velocity
> > fields in the locations:
> >    x in (-2, 5), y in (-5, 5), z in(0,2)
> >
> > Is there any way to extract this data?
> >
> > In addition, is there any way to extract the same data but in a 2d
> > plane? I mean:
> >    x in (-2, 5), y in (-5, 5), z=1
> >
> > Thank you in advance.
> > Cheers
> > SL
> > _______________________________________________
> > Nek5000-users mailing list
> > Nek5000-users at lists.mcs.anl.gov
> > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
> > _______________________________________________
> > Nek5000-users mailing list
> > Nek5000-users at lists.mcs.anl.gov
> > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
>
>
>
>
> ------------------------------
>
> _______________________________________________
> Nek5000-users mailing list
> Nek5000-users at lists.mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
>
>
> End of Nek5000-users Digest, Vol 85, Issue 5
> ********************************************
>

-- 
Phil Sakievich

PhD Candidate - Mechanical Engineering
Arizona State University - Ira A. Fulton School for Engineering of Matter
Transport and Energy
Tempe, Arizona
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/nek5000-users/attachments/20160303/38ccc211/attachment.html>