In summary, the conversation discusses an issue with a particle lost error in an MCNP model and the need to properly define cells in the 3-D space. It is suggested to use the interactive plotter and the "ip" option to identify any geometry errors. The conversation also mentions using the "pz 0" command to get a cross section and suggests defining reflectors with simple surfaces instead of macrobodies to potentially resolve the issue with the void cell.
#1
AlexFi
19
5
TL;DR Summary
Help identify gap in geometry or something
I keep getting particle lost error even though there were no hole in the lattice.
Can someone identify any mistake in my code?
The first thing to do is fire-up the plotting package that comes with MCNP. Look for those dotted lines around some cells. These indicate geometry errors.
In an MCNP model, every part of 3-D space must be filled with exactly one cell. There must be no gaps that are not part of a defined cell. And no point can be part of more than one cell.
#3
rpp
106
71
To give a few more hints, search for "interactive plotter" in the manual and use the "ip" option on the command line.
#4
Alex A
Gold Member
234
153
I ran mcnp5 ip inp=file
i means process the input file, p means plot the geometry in the interactive plotter.
I click on the "Click here or picture or menu" and type "pz 0" to get a cross section through the reactor. Nothing is coloured in, everywhere I click on the plot is cell 99. This strongly suggests an error in the definition of cell 99, the void cell. Without having much time to look into it :5:6 where 5 and 6 are macrobody surfaces "Top reflector" and "Bottom reflector" feels wrong. Defining those reflectors with simple surfaces rather than macrobodies might make defining the void cell easier.