Coordinator
May 12, 2012 at 8:57 PM

As mentioned in the
previous discussion, I would like to run a simulation in the xz plane (all "Y" information in a single bin), where the goal is to capture 2D images of radiance pointing in different polar angles
in the XZ plane. I'm not sure that we have this angle directly available in our existing detectors.
In these results, you can see RadianceOfXAndYAndZAndThetaAndPhi, where each image represents a different theta, or polar angle off of the Uz=1 axis. The source is a collimated point source pointing downward, and
you can see the forwarddirected photons in the top left image. Theta goes (by definition) from 0 (Uz=1) to Pi (Uz=1), and data is collected in a
single phi bin.
Here's my challenge: I now want to tilt the direction of the source toward the positive xdirection. However, since theta is only defined to go from 0:Pi, I will not be able to capture the "backward" (negativexpointing) photons
from Pi:2*Pi.
Is the solution (using existing detectors) to break up Phi into four quadrants, such that the negativexpointing" photons are in either Pi<phi<Pi/2 or Pi/2<phi<Pi and the positivexpointing ones are in Pi/2<phi<0 and 0<phi<Pi/2?
Presumably this would let me separate out angles the XZ plane that were negative vs positive x at a given theta.


Developer
May 12, 2012 at 9:47 PM

I'm not sure I completely understand your question. If theta spans [0,pi] and phi spans[pi,pi], then
all angles are covered within the unit sphere. So any Direction Cosine that describes the angular orientation of the photon will be capture in some bin, none will be lost.


Coordinator
May 12, 2012 at 10:47 PM
Edited May 12, 2012 at 10:48 PM

I didn't mean to imply I was losing photons. When I said:
"I will not be able to capture the "backward" (negativexpointing) photons from Pi:2*Pi"
I should have said:
"I will not be able to discriminate the "backward" (negativexpointing) photons from Pi:2*Pi from the "forward" ones"
In any case, consider a 2D geometry in XZ. Say we name the polar angle in that plane "alpha". I want to tally Radiance(X,Z,Alpha), where Alpha goes from 0:2*Pi.
I did what I proposed above, and the results are making intuitive sense, but I'd like to gain some more confidence in this mathematically.

