Result is zero flux for MCNP6 *F4 tally

In summary, the conversation discusses a problem with finding flux in a cell bounded by two spheres and a cone. The code is producing a warning and a tally result of zero, and the surfaces and cells are defined in the code. The speaker also mentions issues with overlapping cells and the need for materials and cross sections. They also express difficulty with learning MCNP and mention an email list for questions.
  • #1
93
6
I am trying yo find the flux in a cell which is bounded by two concentric spheres and a cone. When I run the code I get a warning that no cross section tables are called for in this problem and a tally result of zero. The way I defined the surfaces and cells is below if anyone sees where I have made a mistake please let me know. I have also attached the code, output and a screen shot of the geometry plot.

1 Mev photon flux at 100 cm to determine KERMA
C Simulate an isotropic point source of 1 MeV photons
C Cell Cards
109 0 -10 IMP:P,e=1
110 0 -11 10 IMP:P,e= 1 $ cell volume inbetween 97.5 and 102.5 shells
120 0 -12 11 IMP:P,e=0 $ Graveyard
130 0 -11 -13 IMP:p,e=1 $ creates a volume bounded by the conical surface as well as concentric spheres

C Surface Cards
10 SO 97.5 $ Creates a 97.5 cm sphere
11 SO 102.5 $ Creates a 102.5 cm sphere
12 SO 200 $ creates graveyard
13 KY 0 0.0625 1 $ creates a conical surfface

C Data Cards
MODE p e $photon and electron transport
C tally card
*F4:p 130
 

Attachments

  • Screenshot (11).png
    Screenshot (11).png
    75.2 KB · Views: 495
  • F4_KERMA.txt
    2.2 KB · Views: 308
  • out_F4_kerma.txt
    37 KB · Views: 276
Engineering news on Phys.org
  • #2
I want to add that I get a result when I *F4 tally on cells 109 and 110.
 
  • #3
All regions must be covered exactly once. You have some cells overlapping. I fixed it one possible way. The old cells are commented out with the new cells shown right after.

You can detect these kinds of things with the plotter if you plot through a surface that involves the offending surfaces. In this case, you need to plot through a surface that intersects the cone. You will see the tell-tale red dotted line. Whenever you have something whacky happening in MCNP, first look for the red dotted line. Change the plot between xy, xz, yz. Change the origin to several places within your model.
 

Attachments

  • KERMA.txt
    2.4 KB · Views: 336
  • #4
You do not have any materials defined.
Each cell should have a material number as the second number, but all of your cells are void (material 0).
No materials, no cross sections.
 
  • #5
DEvens said:
All regions must be covered exactly once. You have some cells overlapping. I fixed it one possible way. The old cells are commented out with the new cells shown right after.

You can detect these kinds of things with the plotter if you plot through a surface that involves the offending surfaces. In this case, you need to plot through a surface that intersects the cone. You will see the tell-tale red dotted line. Whenever you have something whacky happening in MCNP, first look for the red dotted line. Change the plot between xy, xz, yz. Change the origin to several places within your model.
Thank you so much! I am finding the learning curve with this to be very steep and the documentation not so good.
 
  • #6
rpp said:
You do not have any materials defined.
Each cell should have a material number as the second number, but all of your cells are void (material 0).
No materials, no cross sections.

Yes the protocol I am following has the flux measured in vacuo
 
  • #7
khary23 said:
Thank you so much! I am finding the learning curve with this to be very steep and the documentation not so good.

You are not wrong.

I spent a month learning it entirely on my own and learned very little. Then I took a 1 week class at Los Alamos. It made all the difference. Now I'm in an office with about eight other people who know the code, several much better than me. And we still occasionally spend a couple hours puzzling over some aspect of the code.
 
  • #8
DEvens said:
You are not wrong.

I spent a month learning it entirely on my own and learned very little. Then I took a 1 week class at Los Alamos. It made all the difference. Now I'm in an office with about eight other people who know the code, several much better than me. And we still occasionally spend a couple hours puzzling over some aspect of the code.

In my opinion the documentation is horrible. I taught myself python as well as other languages in a few weeks, but this has been rough. It makes me feel better that I am not the only one whi has had difficulties.
 
  • #9
We seem to get many questions on use of MCNP6. When reviewing our archives I see that many of those questions in the past went unanswered.

Isn't there a user forum that focuses on that software?
 
  • #10
anorlunda said:
We seem to get many questions on use of MCNP6. When reviewing our archives I see that many of those questions in the past went unanswered.

Isn't there a user forum that focuses on that software?

There is an email list. You must be registered with RSICC to subscribe to it. And sometimes the questions go unanswered there also. It's all volunteer, just as it is here. Occasionally the developers will answer a question on that email list.
 
  • Like
  • Informative
Likes khary23 and anorlunda
  • #11
anorlunda said:
We seem to get many questions on use of MCNP6. When reviewing our archives I see that many of those questions in the past went unanswered.

Isn't there a user forum that focuses on that software?
I would love auch a forum.
 
  • #12
khary23 said:
Yes the protocol I am following has the flux measured in vacuo
Then the warning is correct. You only have a vacuum, so the code is giving you a warning that there are no materials (or cross sections).

The code is very general and models a lot of different physics, so there is a fairly steep learning curve to it.
However, I would not say that it isn't documented well. It is extremely well documented, especially compared to other codes I've used. All the information is there, it just may not be obvious where to look, or what options you should use.

For beginners, I would suggest starting with one of several "get started" guides on the internet. @Astronuc has put together a list in one posting that is pinned to the "nuclear engineering" page
https://www.physicsforums.com/threads/mcnp-resources.959258/
My personal preference is the Shultis and Faw guide from Kansas State
https://www.mne.k-state.edu/~jks/MCNPprmr.pdf
It is also very helpful to start with a sample problem. There are many samples in the MCNP samples distributed with the code and on the internet. (One trick I use is to do a google search on the specific MCNP input card names, but your case doesn't really use any unique card names.)

Going back to your original problem, I would focus on the geometry errors pointed out by @DEvens.
First, all of the geometry should be covered by non-overlapping cells. You should also pay attention to the boundary conditions.

Second, you have defined a region with zero importance (called the "graveyard"), but your tally is outside that region. Once the particle hits a region of zero importance, the code will quit tracking the particle and you will never get to the tally region. Define a non-void region all the way to the tally region, then put a voided region beyond the tally region.

There are many people on the forum willing to help with particular questions, but understand that most people do not have time to download your problem, run the case, and debug it for you.
 
  • Informative
  • Like
Likes DEvens and anorlunda
  • #13
rpp said:
Then the warning is correct. You only have a vacuum, so the code is giving you a warning that there are no materials (or cross sections).

The code is very general and models a lot of different physics, so there is a fairly steep learning curve to it.
However, I would not say that it isn't documented well. It is extremely well documented, especially compared to other codes I've used. All the information is there, it just may not be obvious where to look, or what options you should use.

For beginners, I would suggest starting with one of several "get started" guides on the internet. @Astronuc has put together a list in one posting that is pinned to the "nuclear engineering" page
https://www.physicsforums.com/threads/mcnp-resources.959258/
My personal preference is the Shultis and Faw guide from Kansas State
https://www.mne.k-state.edu/~jks/MCNPprmr.pdf
It is also very helpful to start with a sample problem. There are many samples in the MCNP samples distributed with the code and on the internet. (One trick I use is to do a google search on the specific MCNP input card names, but your case doesn't really use any unique card names.)

Going back to your original problem, I would focus on the geometry errors pointed out by @DEvens.
First, all of the geometry should be covered by non-overlapping cells. You should also pay attention to the boundary conditions.

Second, you have defined a region with zero importance (called the "graveyard"), but your tally is outside that region. Once the particle hits a region of zero importance, the code will quit tracking the particle and you will never get to the tally region. Define a non-void region all the way to the tally region, then put a voided region beyond the tally region.

There are many people on the forum willing to help with particular questions, but understand that most people do not have time to download your problem, run the case, and debug it for you.

Thank you for the pointers!
 
  • #14
khary23 said:
In my opinion the documentation is horrible. I taught myself python as well as other languages in a few weeks, but this has been rough. It makes me feel better that I am not the only one whi has had difficulties.

Hi,
remember that MCNP is not a informatic language but a nuclear physics code!
ok first you must try to have a good geometry with no cookies cutter. But it just the beginning!
After that you must think like a nuclear physicist.
In your example you are in "mode p e" but with no material. It makes no sense! How to create electrons with photons in vacuum ? Impossible!
Another example you want to calculate a kerma. Remember what a kerma: Kinetic energy released per unit mass. But how transfer an energy in vacuum with no interaction (because no material) ?
Another example : you calculated a "*F4" and weighted with DE/DF. Your DE/DF are the mass energy-absorption coefficient. Strange! Why not used a tally type F6 (energy deposition over a cell).
In the of Astronuc post (https://www.physicsforums.com/threads/mcnp-resources.959258/) you can see that there is not only the document of MCNP6.2 pdf (LA-UR-17-29981.pdf!
There is also a large number of publications citing MCNP (try https://scholar.google.fr/scholar?hl=fr&as_sdt=0,5&q=MCNP6&btnG= )
It's quite logical to have as much publication when a code can do as much type of calculation (and different particles: photons, electron neutrons, heavy ions, ...)
It is strange to write "the documentation is horrible".
but it's just my opinion
PSR
 
  • #15
PSRB191921 said:
Hi,
remember that MCNP is not a informatic language but a nuclear physics code!
ok first you must try to have a good geometry with no cookies cutter. But it just the beginning!
After that you must think like a nuclear physicist.
In your example you are in "mode p e" but with no material. It makes no sense! How to create electrons with photons in vacuum ? Impossible!
Another example you want to calculate a kerma. Remember what a kerma: Kinetic energy released per unit mass. But how transfer an energy in vacuum with no interaction (because no material) ?
Another example : you calculated a "*F4" and weighted with DE/DF. Your DE/DF are the mass energy-absorption coefficient. Strange! Why not used a tally type F6 (energy deposition over a cell).
In the of Astronuc post (https://www.physicsforums.com/threads/mcnp-resources.959258/) you can see that there is not only the document of MCNP6.2 pdf (LA-UR-17-29981.pdf!
There is also a large number of publications citing MCNP (try https://scholar.google.fr/scholar?hl=fr&as_sdt=0,5&q=MCNP6&btnG= )
It's quite logical to have as much publication when a code can do as much type of calculation (and different particles: photons, electron neutrons, heavy ions, ...)
It is strange to write "the documentation is horrible".
but it's just my opinion
PSR
Maybe horrible is overdoing it a bit. I just find that things could be explained better or maybe there is just so much material that its difficult to find what I'm looking for.

So what I am looking for is the Air-kerma strength which is defined as
"Air-kerma strength, SK , is the air-kerma rate, in vacuo and due to photons of energy greater than delta at distance D, multiplied by the square of this distance, D^2."
This is why I have my cells in a vacuum. I understand what you are saying about mode e and have taken it out. I have also tried the F6 tally and get the "1 tally volumes or areas were not input nor calculated" fatal error.
 
Last edited:
  • #16
Also wanted to thank you for the link to the publications.
 

Suggested for: Result is zero flux for MCNP6 *F4 tally

Back
Top