[Nek5000-users] Setting up solid elements with genbox

Wed Jun 20 11:51:01 CDT 2012

Thanks everyone,

Switching the order of the mesh points in ascending order fixed my right-handed element error. 

Best,
Kyle

----- Original Message -----
From: nek5000-users at lists.mcs.anl.gov
To: nek5000-users at lists.mcs.anl.gov
Sent: Tuesday, June 19, 2012 9:02:17 PM
Subject: Re: [Nek5000-users] Setting up solid elements with genbox

Hello Kyle, 

I do not know much about the set up for your case, however I see an error in the way the .box file is setup. 

In a .box file, the points should be in an ascending order. Hence for the following lines in the your .box file the points in y-direction should be in the reverse order, i.e. start with -3.0 to -1.0. 

Box lower fuel 
-8 8 nelx,nely for Box 
0 8 1. x0,x1,gain (rescaled in usrdat) 
-1.00 -1.04 -1.25 -1.55 -2.00 -2.45 -2.75 -2.96 -3.00 y0,y1,...,yn 
, , , bc's (3 chars each!) 
I ,I ,I ,E bc's (3 chars each!) 

Hope this helps. 

Harish. 

On Tue, Jun 19, 2012 at 6:16 PM, < nek5000-users at lists.mcs.anl.gov > wrote: 

Thanks Justin. I did look at that example, and it is helpful (especially for setting up the boundary conditions in the .box file, which I have now adjusted to be blank for the fluid bc's in the solid domain). 

My issue is that I don't understand the general steps necessary to setup an arbitrary domain. Simply changing the dimensions in the .box file, and using the conj_ht.rea as the base .rea file like this: 

conj_ht.rea 
2 spatial dimension ( < 0 --> generate .rea/.re2 pair) 
2 number of fields 

Box fluid 
-8 8 nelx,nely for Box 
0 8 1. x0,x1,gain (rescaled in usrdat) 
-1.0 -0.96 -0.75 -0.45 0.00 0.45 0.75 0.96 1.00 y0,y1,...,yn 
P ,P ,W ,W bc's (3 chars each!) 
P ,P ,E ,E bc's (3 chars each!) 
Box upper fuel 
-8 8 nelx,nely for Box 
0 8 1. x0,x1,gain (rescaled in usrdat) 
1.00 1.04 1.25 1.55 2.00 2.45 2.75 2.96 3.00 y0,y1,...,yn 
, , , bc's (3 chars each!) 
I ,I ,E ,I bc's (3 chars each!) 
Box lower fuel 
-8 8 nelx,nely for Box 
0 8 1. x0,x1,gain (rescaled in usrdat) 
-1.00 -1.04 -1.25 -1.55 -2.00 -2.45 -2.75 -2.96 -3.00 y0,y1,...,yn 
, , , bc's (3 chars each!) 
I ,I ,I ,E bc's (3 chars each!) 

and then using genbox, changing nelv, deleting the fluid bc's for elements greater than 64, using genmap results in the peculiar error when running nek: 

setup mesh topology 

WARNINGa: Detected non-right-handed element. 
Number 129 C1-4: -0.4000E-01 -0.4000E-01 -0.4000E-01 -0.4000E-01 

This is with the latest repo (843). Maybe I am missing some necessary changes to the .rea file? 

Thanks, 

Kyle 

----- Original Message ----- 
From: nek5000-users at lists.mcs.anl.gov 
To: nek5000-users at lists.mcs.anl.gov 
Sent: Tuesday, June 19, 2012 3:17:21 PM 
Subject: Re: [Nek5000-users] Setting up solid elements with genbox 

Hi Kyle, 

Did you take a look at the conj_ht case in the examples folder? It has something pretty close to what you want. 

Justin 

On Tue, Jun 19, 2012 at 1:18 PM, < nek5000-users at lists.mcs.anl.gov > wrote: 

Hi all, 

I am attempting to setup a simple 2D conjugate heat transfer problem using genbox. I use the example "turbChannel" as my model. I want to setup a fluid box in between two solid boxes. Therefore, my 2DturbChannel.box file appears: 

base.rea 
2 spatial dimension ( < 0 --> generate .rea/.re2 pair) 
2 number of fields 
Box 
-8 8 nelx,nely for Box 
0 8 1. x0,x1,gain (rescaled in usrdat) 
-1.0 -0.96 -0.75 -0.45 0.00 0.45 0.75 0.96 1.00 y0,y1,...,yn 
P ,P ,W ,E bc's (3 chars each!) 
P ,P ,W ,E bc's (3 chars each!) 
Box 
-8 8 nelx,nely for Box 
0 8 1. x0,x1,gain (rescaled in usrdat) 
1.00 1.04 1.25 1.55 2.00 2.45 2.75 2.96 3.00 y0,y1,...,yn 
P ,P ,E ,W bc's (3 chars each!) 
P ,P ,E ,W bc's (3 chars each!) 
Box 
-8 8 nelx,nely for Box 
0 8 1. x0,x1,gain (rescaled in usrdat) 
-1.00 -1.04 -1.25 -1.55 -2.00 -2.45 -2.75 -2.96 -3.00 y0,y1,...,yn 
P ,P ,W ,E bc's (3 chars each!) 
P ,P ,W ,E bc's (3 chars each!) 

Running genbox on this file produces a .rea file with 192 fluid elements. I first try to change "nelv" in the .rea file to 64. Running genmap on this file produces a "PERIODIC MISMATCH" error, likely due to the 192 boundary conditions for the fluid field. So I then delete all of the boundary conditions for elements greater than 64 in the fluid (leaving thermal boundary conditions for all 192 elements). Genmap now works, but running nek with this .rea/.map combination gives the error: 

ERROR READING INITIAL CONDITION/DRIVE FORCE DATA 
ABORTING IN ROUTINE RDICDF. 

I thought that this procedure would work, given previous discussion on the problem of conjugate heat transfer in Nek5000 ( http://lists.mcs.anl.gov/pipermail/nek5000-users/2010-February/000121.html ). Perhaps I generated the .map file from the incorrect form of the .rea file? My question: is there a simple way to deal with solid elements in this configuration without using prenek? 

Thanks, 

Kyle 
_______________________________________________ 
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 

_______________________________________________
Nek5000-users mailing list
Nek5000-users at lists.mcs.anl.gov
https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users