3D lat example request

Report all good and bad behavior here
Tian
Posts: 18
Joined: Mon Mar 10, 2014 1:19 pm
Security question 1: No
Security question 2: 92

3D lat example request

Post by Tian » Wed May 28, 2014 10:55 am

I am writing a 3D burn up model of a single assembly, for the lattice I have two choices, lattice type 11 and lattice type 9. For type 11 I found that the output power is not in W but in W/cm, and for type 9 I come cross overlap problems when I put cells together. Is it possible for me to have two examples for these two types of lattices respectively?

Thank you

User avatar
Jaakko Leppänen
Site Admin
Posts: 2387
Joined: Thu Mar 18, 2010 10:43 pm
Security question 2: 0
Location: Espoo, Finland
Contact:

Re: 3D lat example request

Post by Jaakko Leppänen » Wed May 28, 2014 12:15 pm

Unfortunately I don't have very good examples available. Lattice type 9 is used in the HTTR input example:

http://virtual.vtt.fi/virtual/montecarl ... ORE/input/

but the geometry is very complicated. It might be better if you posted your inputs here and we'll try to figure out what the problem is.

What do you mean that the output power is in W/cm? Basically the units depend on normalization.
- Jaakko

Tian
Posts: 18
Joined: Mon Mar 10, 2014 1:19 pm
Security question 1: No
Security question 2: 92

Re: 3D lat example request

Post by Tian » Wed May 28, 2014 12:27 pm

What I mean is that in that 3D lattice calculation, the printed output power seems only matches the power in unit of W/cm rather than W, looks like Serpent is performing a 2D calculation rather than 3D. And also the output power looks related to the number of elements in z direction. I am not quite sure about what is wrong in the geometry description.

Here I attached the input file type 9 and type 11 respectively.

Thanks
Attachments
3Dlat.tar.gz
(3.07 KiB) Downloaded 1277 times

User avatar
Jaakko Leppänen
Site Admin
Posts: 2387
Joined: Thu Mar 18, 2010 10:43 pm
Security question 2: 0
Location: Espoo, Finland
Contact:

Re: 3D lat example request

Post by Jaakko Leppänen » Wed May 28, 2014 12:43 pm

Thank's. I'll take a look at the inputs and get back to you. The problem with normalization is probably related to incorrect material volumes. Serpent calculates the volumes of material regions automatically, but this only works for simple 2D geometries. If you enter the material volumes manually in the material cards using the "vol" entry, you should get the correct power for the given power density.
- Jaakko

User avatar
Jaakko Leppänen
Site Admin
Posts: 2387
Joined: Thu Mar 18, 2010 10:43 pm
Security question 2: 0
Location: Espoo, Finland
Contact:

Re: 3D lat example request

Post by Jaakko Leppänen » Wed May 28, 2014 1:01 pm

Input "lat11" looks OK. Do you get the expected value for power if you enter the volumes of material regions manually? Note that you can check the volumes using the "-checkvolumes" command line option (Sec. 1.2 in Serpent manual).

Input "lat9" has some geometry errors, but basically you should get the same results as with lat11, if you define each axial zone as a separate complete two-dimensional universe:

Code: Select all

lat 140 1 0.0 0.0 18 18 1.27

3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 
3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3
3 3 4 3 3 3 55 3 3 3 3 55 3 3 3 4 3 3
3 3 3 55 3 3 3 3 3 4 3 3 3 3 55 3 3 3
3 3 3 3 4 3 3 55 3 3 55 3 3 4 3 3 3 3
3 3 3 3 3 55 3 3 3 3 3 3 55 3 3 3 3 3
3 3 55 3 3 4 3 3 3 3 3 3 4 3 3 55 3 3
3 3 3 3 55 3 3 3 3 3 3 3 3 55 3 3 3 3
3 3 3 4 3 3 3 3 3 3 3 3 3 3 3 3 3 3
3 3 3 3 3 3 3 3 3 3 3 3 3 3 4 3 3 3
3 3 3 3 55 3 3 3 3 3 3 3 3 55 3 3 3 3
3 3 55 3 3 4 3 3 3 3 3 3 4 3 3 55 3 3
3 3 3 3 3 55 3 3 3 3 3 3 55 3 3 3 3 3
3 3 3 3 4 3 3 55 3 3 55 3 3 4 3 3 3 3
3 3 3 55 3 3 3 3 4 3 3 3 3 3 55 3 3 3
3 3 4 3 3 3 55 3 3 3 3 55 3 3 3 4 3 3
3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3
3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3

surf 1400 sqc 0.0 0.0 11.43
cell 1044 104 fill 140 -1400
cell 1040 104 water     1400
and stack them top of each other using lattice type 9. Also in this case you'll need to set the material volumes manually.
- Jaakko

Tian
Posts: 18
Joined: Mon Mar 10, 2014 1:19 pm
Security question 1: No
Security question 2: 92

Re: 3D lat example request

Post by Tian » Wed May 28, 2014 4:54 pm

Thanks. Now everything looks make sense and volume check also looks fine.

Tian
Posts: 18
Joined: Mon Mar 10, 2014 1:19 pm
Security question 1: No
Security question 2: 92

Re: 3D lat example request

Post by Tian » Wed May 28, 2014 6:15 pm

By the way, I also kept coming across "Segmentation fault" while running lat 11, what may the reason be? The version is 2.1.20. Also another question is about the power density, in the input I set the powdens 26E-9, when I use ompi doing the calculation, with 10 nodes, the printed output of first step gives like

Code: Select all

TOT_POWDENS               (idx, [1:   6]) = [  1.33290E+02 0.00027  1.33290E+02 0.00027  0.00000E+00 0.0E+00 ];
but the TOT_POWER is what I expected, is it anything wrong or it is only due to ompi calculation?

User avatar
Jaakko Leppänen
Site Admin
Posts: 2387
Joined: Thu Mar 18, 2010 10:43 pm
Security question 2: 0
Location: Espoo, Finland
Contact:

Re: 3D lat example request

Post by Jaakko Leppänen » Thu May 29, 2014 6:11 pm

Could you try running the calculation with the -DDEBUG compiler option enabled? It could reveal more information about the cause of the segmentation fault.

Do you get the same result for power density if you run without MPI?
- Jaakko

Tian
Posts: 18
Joined: Mon Mar 10, 2014 1:19 pm
Security question 1: No
Security question 2: 92

Re: 3D lat example request

Post by Tian » Mon Jun 02, 2014 3:29 pm

I run the same input with 2.1.20 ompi , 2.1.20 and 2.1.18 respectively, I think it is due to the new version as I got the expected output power density in 2.1.18 but in 2.1.20 the output power density is much larger, I did not got the reason why.

Here is the output in 2.1.18, ran without ompi,

Code: Select all

TOT_POWER                 (idx, [1:   6]) = [  9.14200E+06 0.0E+00  9.14200E+06 0.0E+00  0.00000E+00 0.0E+00 ];
TOT_POWDENS               (idx, [1:   6]) = [  2.59640E-02 0.0E+00  2.59640E-02 0.0E+00  0.00000E+00 0.0E+00 ];
Here is the output from 2.1.20, ran without ompi,

Code: Select all

TOT_POWER                 (idx, [1:   6]) = [  9.14200E+06 4.1E-09  9.14200E+06 4.1E-09  0.00000E+00 0.0E+00 ];
TOT_POWDENS               (idx, [1:   6]) = [  1.33261E+02 0.00027  1.33261E+02 0.00027  0.00000E+00 0.0E+00 ];
And here below is the result of 2.1.20 with ompi, 10 nodes,

Code: Select all

TOT_POWER                 (idx, [1:   6]) = [  9.14200E+06 4.0E-09  9.14200E+06 4.0E-09  0.00000E+00 0.0E+00 ];
TOT_POWDENS               (idx, [1:   6]) = [  1.33260E+02 0.00026  1.33260E+02 0.00026  0.00000E+00 0.0E+00 ];
All these results are only from the first burn up steps.

Also I attached the input for checking, could you check whether it is new version problems or there was something run in my input file?

User avatar
Jaakko Leppänen
Site Admin
Posts: 2387
Joined: Thu Mar 18, 2010 10:43 pm
Security question 2: 0
Location: Espoo, Finland
Contact:

Re: 3D lat example request

Post by Jaakko Leppänen » Mon Jun 02, 2014 3:55 pm

I think you forgot to include the input?
- Jaakko

Post Reply