MCNP6: Getting a "10 particles got lost" error

In summary, the geometry of the input file appears to be incorrect, and changing the z limits in the fill cell does not fix the issue.
  • #1
a1234
78
6
MCNP6 gives me a "10 particles got lost" error when I try to run the attached input file modeling a 3x3 fuel lattice surrounded in coolant. As I understand it, this error is usually related to the geometry/surface definitions of each component, but I'm unsure of what the source of the error is. The output file says "there appears to be no cell on the other side of the surface from cell 11 at that point," which makes me think the issue stems from the way the lattice was defined.

I'd appreciate it if somebody could help me figure out the source of the issue, or perhaps help me interpret the output file. Thanks.
 
Engineering news on Phys.org
  • #2
a1234 said:
when I try to run the attached input file
I'm not seeing any file attachment. Maybe the format of the file is preventing it from uploading. If so, change the suffix to ".txt" to upload it, and mention what the suffix should be changed back to in order to run it.
 
  • Informative
Likes Oldman too and Alex A
  • #3
The Geometry.txt file extension should be changed to .i to run it in MCNP.
 

Attachments

  • outq.txt
    47.6 KB · Views: 198
  • Geometry.txt
    1.6 KB · Views: 175
  • Like
Likes berkeman
  • #4
In universe 2, the outside of the cladding isn't defined as a cell. ie +(surface)4. Fixing this does not fix everything, but it does help.

Your last line for the fill is 31113 instead of 33333. I don't understand lattices, is this correct? (changing it does not seem to help)

I'll take another look tomorrow.
 
  • Informative
Likes berkeman
  • #5
Redefined cell 11 as being +4 and being material 3 (coolant), but still getting the same error.
Yes, the last line in the fill should have been 3 3 3 3 3.
 
  • #6
In this, cell 11 is still the lattice cell. I just made a new cell number for the missing volume.

Ok, mcplot to the rescue. Running with ipz on the command line let's us click to see a default plot in the interactive viewer. We can then click on an area to see what cell it's defined as. The command "pz 0" then shows us a slice. The is immediately wrong, there are only two fuel rods. When you specified the lattice you put the z's first. The right thing to do is probably be to not specify z limits in cell 11 at all, it's a 2D lattice. I see examples in the manual that do this so I think this is, at worst, fine. I then put the z limits into the cell being filled. This was helpful, not a total fix, I'm not 100% it isn't just a mcplot issue.

Next we have the problem that the 5x5 lattice isn't big enough to fill the volume you've told it to fill. This means there is a shell of volume around the core that isn't assigned to any cell. Fixing this is just means tweaking the geometry until it matches the real thing, if this is based on a real thing.

I think at that point the leaks stopped. I had a source overflow that needed me to change the initial k guess, but I've messed with a lot of stuff in order to get it running on my system and to make things easily visible in mcplot. Btw, your control rod tube is very very narrow. I don't think I needed to do anything else to make it work.
 
  • Informative
  • Like
Likes a1234 and berkeman
  • #7
Thank you so much for the help, I was able to get it to work!
 
  • Like
Likes berkeman

FAQ: MCNP6: Getting a "10 particles got lost" error

What does the "10 particles got lost" error mean in MCNP6?

The "10 particles got lost" error in MCNP6 is an indication that 10 particles in the simulation have exited the geometry or have been terminated before reaching the end of the calculation. This can happen due to a variety of reasons such as incorrect geometry definitions, improper boundary conditions, or insufficient number of particles in the simulation.

How can I troubleshoot the "10 particles got lost" error in MCNP6?

To troubleshoot this error, you can first check the input file for any errors or inconsistencies in the geometry definitions. You can also try increasing the number of particles in the simulation or adjusting the boundary conditions to prevent particles from exiting the geometry prematurely. Additionally, reviewing the MCNP6 manual or seeking help from experienced users can provide further guidance on troubleshooting this error.

Can the "10 particles got lost" error impact the accuracy of my MCNP6 simulation?

Yes, the "10 particles got lost" error can affect the accuracy of your simulation as it indicates that some particles have not reached their intended destinations in the geometry. This can lead to incorrect results and affect the overall reliability of your simulation. It is important to address this error and ensure that all particles are properly tracked throughout the simulation.

Is there a way to prevent the "10 particles got lost" error in MCNP6?

While it is not possible to completely eliminate this error, there are steps that can be taken to reduce its occurrence. These include carefully defining the geometry, using appropriate boundary conditions, and ensuring a sufficient number of particles in the simulation. Additionally, regularly reviewing and debugging the input file can help catch any potential issues that may lead to this error.

Are there any alternative codes or methods to avoid the "10 particles got lost" error in MCNP6?

There are other Monte Carlo codes available that can be used as alternatives to MCNP6, such as FLUKA and Geant4. These codes have different algorithms and may have different error messages and ways of handling lost particles. Additionally, using variance reduction techniques and optimizing the geometry can also help minimize the occurrence of the "10 particles got lost" error in MCNP6.

Similar threads

Back
Top