Infinite Bounding Box

User Support
1

Hi all,

I am building a reactor geometry with outer boundaries given by two ZPlane’s and one ZCylinder, and am obtaining a bounding box with lower_left equal to [-inf, -inf, 0.0] and upper_right equal to [+inf, +inf, 5.3]. Particles seem to track fine over the geometry (none are lost, k_eff values are reasonable), so I am wondering if there are any practical consequences of having an infinite bounding box in the x and y directions? In other words, am I safe to move forward with my analysis, or is this indicative of a problem? The model is kind of complicated, and I could not reproduce the problem with only three surfaces (2 * ZPlane and 1 * ZCylinder), so I’m pretty sure this is related to how I’ve defined the geometry.

I’ve attached a figure of the geometry along the XZ plane for y = 0. The red corresponds to solid UO2, the gray to graphite, and the white to the image background.

Thanks!

-April

Untitled.png
Untitled.png671×911 18.9 KB

2

Hi April,

I suspect that the issue is with those slanted surfaces. If you have any cells that are bounded by a cone or a general plane, there’s no easy way to calculate the bounding box for them so we end up defaulting to just an infinite box in that case. In terms of practical consequences, the bounding boxes are not used for anything critical, so it should be safe to proceed with your analysis. If you wanted to do a stochastic volume calculation or something else that required specifying a bounding box, you’d just have to specify it manually.

Best,
Paul

3

Thanks for the clarification, Paul!

-April