Importance of Meshing

User Support
1

Hello Everyone,

I’m a bit confused about the applicability of meshing in OpenMC. I was tailoring the mesh dimensions in ‘Post Processing Example’ and what I found was rather strange. When I changed the dimensions from [100,100] to [50,50] and [150,150] the relative error increased in the latter case and decreased in the former. Though the k-effective value and other values were same but I think that is because the number of particles (5000) were same and the fact that the change in values with mesh were very small to be seen in the output.

Other than an improvement in capturing the flux.mean and fission.mean images, I do not find any reason of increasing the mesh dimensions (making it more discrete).

So I tried uploading the images of distribution of relative error with [50,50] and [150,150] dimensions but the server dismisses it. Please help me understand the reason behind this and let me know if and where I’m wrong.

Thank You in Advance
Shubham

2

Answering an old message that never received a response – the behavior that you are describing is expected. If you keep the number of particles constant and increase the number of mesh cells (i.e., smaller volume for each mesh cell), this means that each mesh cell will have fewer collisions / fewer tracks on average, thereby leading to an increase in the reported variance. As the number of mesh cells is decreased, the volume of each mesh cell is greater and therefore there are more collisions / more tracks on average, leading to a lower variance.

Best regards,
Paul