Message: Re: Optical Photons: alternative ways of defining wrapping material? Not Logged In (login)
 Next-in-Thread Next-in-Thread
 Next-in-Forum Next-in-Forum

Feedback Re: Optical Photons: alternative ways of defining wrapping material? 

Forum: Processes Involving Optical Photons
Date: 10 Dec, 2008
From: Gumplinger Peter <Gumplinger Peter>

Dear Estela,

My instructions of how to make option 2) mirror option 1) was not complete. When the photons go from the air-gap back to the interface with the scintillator, this order of volumes does not yet have a G4OpSurface defined for it, and hence the interface will be treated as default; e.g. perfectly polished. You need to define a second logical border surface with the reverse order of physical volumes and attach to it the same G4OpSurface.

> I had understood from the post #882 on 27-03-2008 (this forum) that only 
> dielectric-metal surfaces take the 'REFLECTIVITY' into account.

This/My post is not correct in this respect. The reflective surface in the 'backpainted' options of the UNIFIED model do take the reflectivity into account.

You can still use option 1) - the simulation of tracking photons should be faster - if you need to have a physical Al casing for the X-rays to see and interact in, but you now have to, of course, implement it as a 'physical volume'. The wrapping in option 1) "lives" only inside the G4OpBoundaryProcess.

> I checked anyway if I could reproduce with method 2) the low light collection
> from my initial results from method 1).

When you declare in method 1) dielectric_metal and UNIFIED you are defining a surface with no air gap. The UNIFIED specific reflection parameters are used and so is your specified reflectivity.

> first I didn't understand well why I need to put the surface of my reflector
> as completely unpolished, since I would rather consider it as perfectly
> polished.

Sorry, but this is a limitation of method 1) The code is not (yet) flexible to allow for anything but a diffuse reflector as the wrapping. If you need a polished wrapping, I am afraid, only your method 2) will work at the moment.

> I was very surprised to see that I get actually I higher light collection
> that ever before: about 45%. Do you know why should it be so?

Much depends on your 'light collection' counting. If the photons must leave the scintillator to be counted then with perfectly polished surfaces light internally reflected on the sides may also stay trapped on the interface with your PMT. A diffuse reflector helps against these eternal bounces.

From Wolfgang's posting:

> ad b) I think you will not really see Alu->Air, because the photons are 
> reflected at the Alu surface without entering the physical Alu volume. (Peter,
> is this correct?) But in the case of total internal reflection of e.g.
> Sci->Air, my photon statistics suggests that the photons really enter the
> physical Air volume before they are reflected back into the physical Sci
> volume (like evanescent waves?).

There is a zero-length reallocation step taken when photons reflect via total internal reflection or off an external metal reflector - you see this step with your SteppingVerbose. This is required by the G4Navigator but it does not mean the photons have in any way penetrated the opposite medium like in reality with the evanescent wave. G4OpBoundaryProcess catches this zero-length step and does not again spring into action. Still, no photon will have actually 'come from' the Al. By the way, if you do not give an index of refraction to any medium, photons will be absorbed should they find themselve in such a medium.

Estela, are you getting the same result with method 2) - including an Air-Sci surface, and diffuse Al, as you get with method 1)? What is the CPU time difference?

I hope I have answered all you questions.

Peter

Inline Depth:
 1 1
 All All
Outline Depth:
 1 1
 2 2
 All All
Add message: (add)

 Add Message Add Message
to: "Re: Optical Photons: alternative ways of defining wrapping material?"

 Subscribe Subscribe

This site runs SLAC HyperNews version 1.11-slac-98, derived from the original HyperNews