Message: Re: questions about optical modeling Not Logged In (login)
 Next-in-Thread Next-in-Thread
 Next-in-Forum Next-in-Forum

Feedback Re: questions about optical modeling 

Keywords: unified optical modeling ground
Forum: Documentation and Examples
Re: Question questions about optical modeling (Aaron Maschinot)
Date: 14 Sep, 2006
From: Gumplinger Peter <Gumplinger Peter>

> But what happens when a photon with an energy not explicitly specified
> in the array is encountered? Does GEANT4 perform some sort of a spline
> in between listed optical values?

Yes, but only a linear interpolation is done between listed values.

>  What happens above (below) the highest (lowest) momentum
>  in the array?

A warning is written:

                G4cout << "\nWarning: G4MaterialPropertyVector::GetProperty";
                G4cout << "\n==> attempt to Retrieve Property below range"
                G4cout << "\n==> attempt to Retrieve Property above range"

and the value associated with the min/max photon energy is returnd.

In the case of G4Scintillation, the array FASTCOMPONENT/SLOWCOMPONENT specifies the range of optical photons produced. It should not be wider than the arrays for the index of refraction etc. for all the materials encountered by these photons, or else, a lot of warnings are produced. For the G4Cerenkov process, the array for the refraction index of the radiator defines the range of optical photons produced. Again, all other optical propery arrays should have the same or larger range on both ends.

Now to your second question:

>   Am I correct in thinking that the above listed variables need to be
>   specified only when thin coverings are used?

No, they are needed if any kind of reflection is possible, even for total internal reflection and Fresnel reflection.

If you choose the UNIFIED model, you need to specify the probability of one of the kinds of reflection that can happen in this model at any dielectric_dielectic surface (regardless of whether there is or isn't also a wrapping). If you don't, UNIFIED assumes 'Lambertian Reflection' and models the reflected photon direction from G4LambertianRand without any reference to sigma_alpha. What you want is 100% LobeReflection whose parameter in the UNIFIED model is 'sigma_alpha' for sampling the FacetNormal. When the photon is found to be refracted none of the UNIFIED Reflection-Types come into play and the FacetNormal is sampled from sigma_alpha alone; just as you suspected.

The logic is such that all probabilities are set to 0% and Lambertian is the default. Now, if you only set:

G4double SpecularLobe[num] = {1.0, ...... ,1.0};


you'll have 100% SPECULARLOBE, no other type of reflection.

As well, you only need the efficiency array for dielectric_metal surfaces and then also only if you want to tag a photon as "detected".

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

 Add Message Add Message
to: "Re: questions about optical modeling"

 Subscribe Subscribe

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

[ Geant 4 Home | Geant 4 HyperNews | Search | Request New Forum | Feedback ]