Message: Re: Behaviour of unified model for backpainted surfaces Not Logged In (login)
 Next-in-Thread Next-in-Thread
 Next-in-Forum Next-in-Forum

Feedback Re: Behaviour of unified model for backpainted surfaces 

Forum: Processes Involving Optical Photons
Re: Question Behaviour of unified model for backpainted surfaces (Nicolas Di Vara)
Date: 25 Jul, 2012
From: Gumplinger Peter <Gumplinger Peter>

Hi Nicolas,

> I have been trying to trace photons impinging on a polishedbackpainted
> surface, and I have a doubt on the exact mechanism of interaction at
> this boundary.

Please, see a concise flow diagram at:

http://hypernews.slac.stanford.edu/HyperNews/geant4/get/AUX/2012/05/23/23.20-70533-ces_in_geant4_revised.png

> - The photon comes at a polished interface and gets reflected, refracted
> or totally internally reflected following the Snell law (in this case
> the parameters into play should be the index of refraction of the volume
> and that of the wrapping surface).

Yes, but Snell Law is applied according to the unified model's ground/roughness paramenter - SigmaAlpha - after sampling a FacetNormal. The 'polished' in polishedbackpainted refers to the wrapping. It implies that the wrapping is a perfectly smooth mirror with only specular spike reflection taking place.

> - In the case it gets refracted, it
> "enters" in the thin slice of the border surface.

Yes.

> Here, depending on the
> probabilities specified for every type of reflection (lobe, specular,
> etc.), the photon is reflected back to the volume.

No, here is your misunderstanding!

> Though, in the case of a polishedfrontpainted surface the photon gets
> simply and only reflected.

Yes,and the reflection is Specular Spike for polishedfrontpainted.

> As an example: - let us imagine a photon coming from a medium with index
> of refraction 1.8 to a wrapping with index of refraction 1 and a surface
> defined as polishedbackpainted

 (and completely diffusive).

???

> In the case
> the photon impinges on the surface with an angle higher than that for
> total internal reflection it gets reflected back.

Yes, but the angle is measured wrt the FacetNormal (remember the polished refers to the wrapping!). The reflection also takes the unified reflection parameters into account and can be specular spike, specular lobe, backscatter or lambertian according to the assigned probabilities.

> Otherwise it enters
> the wrapping and is diffused.

Refraction is calculated wrt the FacetNormal.

> - if, as a countercheck, we set the index
> of refraction of the wrapping surface at 1.8 (the same as the bulk
> volume), the photon always crosses the surface and gets diffusely
> reflected by the painted surface.

No, due to polished in 'polishedbackpainted' it gets specular spike reflected by the painted surface (diffuse - Lambertian - if you had specify groundbackpainted instead).

The current version of G4OpBoundaryProcess does not shortcut the Snell Law calculation when Rindex1 == Rindex2 (where Rindex2 is that of the gap in 'backpainted' or the medium across the boundary). It does short cut the calculation if Medium1 == Medium2 for a regular dielectric_dielectric interface.

Hope this clarifies the situation, Peter

 Add Message Add Message
to: "Re: Behaviour of unified model for backpainted surfaces"

 Subscribe Subscribe

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