search

mit-crpg / OpenMOC

A method of characteristics code for nuclear reactor physics calculations.
https://mit-crpg.github.io/OpenMOC/
Other
149 stars 84 forks source link

ray tracing generation does not cover the whole interface #436

Open xycjscs opened 5 years ago

xycjscs commented 5 years ago

I am developing OpenMOC based on one history version. I am not sure if the following issue has been fixed by one commit.

Taking the pin-cell.py as the reference, 2X1 pins model is simulated with 128 azimuthal angles. The dimension of each cell is 12.6*12.6 cm, and the ray trace width is 0.01 cm. When I accumulate all ray traces crossing the interface between these two pins, ray traces do not cover the whole interface through different azimuthal angles' sweeping. Some azimuthal angle covers most of the interface, while others did not. For different azimuthal angles, the following figure shows different corresponding lengths of the interface covered by all tracked ray tracing. https://www.deepgalaxy.tech:5001/sharing/JcB8A1tCA

This may do not affect the cell averaged solutions, however, changes the surface solutions, especially for discontinuity factors related works.

GiudGiud commented 5 years ago

Surface areas, like volumes, are dependent on the quadrature angles and spacings in track-based integration. As you reduce the track spacing you should see the surface area converge to the analytical value.

I'm in Raleigh beginning of August if you want to talk about it

xycjscs commented 5 years ago

I think I may understand. Even the cell areas are accumulated based on the quadrature angles and spacings (segments). Hence, this 12.6X12.6 cm^2 cell has a different area with different azimuthal angles. Welcome to Raleigh. Appreciate your help. I hope to get some suggestions from you at that time. Thank you very much.