Optical photon velocity

tome · 2020-12-07T19:09:19.406Z

Hi,

I’m injecting relativistic muons (2 GeV) in water. Water has a n=1.33 index of refraction.

For simplification I turned off most processes
/process/inactivate Electromagnetic all
/process/activate Cerenkov

I can see the Cherenkov optical photons being produced and propagated with a velocity c. Shouldn’t it be c/1.33 ? See below.

Or am I missing something ?
Many thanks in advance.
Bernardo Tomé

G4Track Information: Particle = opticalphoton, Track ID = 112, Parent ID = 1

Step# X(mm) Y(mm) Z(mm) KinE(MeV) dE(MeV) StepLeng TrackLeng NextVolume ProcName
0 0 0 1.7e+03 3.62e-06 0 0 0 WCD_10010 initStep

AlongStepDoIt (after all invocations):
++List of invoked processes

  1) CoupledTransportation

++G4Step Information 
  Address of G4Track    : 0x7f87f26e1470
  Step Length (mm)      : 2256.179710119461
  Energy Deposit (MeV)  : 0
  -----------------------------------------------------------------------
    StepPoint Information               PreStep            PostStep
  -----------------------------------------------------------------------
     Position - x (mm)   :                    0   507.6507772455736
     Position - y (mm)   :                    0   1395.593722444087
     Position - z (mm)   :    1698.515744581532                   0
     Global Time (ns)    :  0.03835567841198741   7.564161114516256
     Local Time (ns)     :                    0   7.525805436104268
     Proper Time (ns)    :                    0                   0
     Momentum Direct - x :   0.2250045840624524  0.2250045840624524
     Momentum Direct - y :   0.6185649645657849  0.6185649645657849
     Momentum Direct - z :  -0.7528282153071921 -0.7528282153071921
     Momentum - x (MeV/c): 8.137314257714439e-078.137314257714439e-07
     Momentum - y (MeV/c): 2.237046647941492e-062.237046647941492e-06
     Momentum - z (MeV/c): -2.722611095038221e-06-2.722611095038221e-06
     Total Energy (MeV)  : 3.616510433163372e-063.616510433163372e-06
     Kinetic Energy (MeV): 3.616510433163372e-063.616510433163372e-06
     Velocity (mm/ns)    :           299.792458          299.792458
     Volume Name         :            WCD_10010           WCD_10010
     Safety (mm)         :                    0               5e-10
     Polarization - x    :   0.2573464190273144  0.2573464190273144
     Polarization - y    :   0.7074766019992675  0.7074766019992675
     Polarization - Z    :   0.6582170449307644  0.6582170449307644

dsawkey · 2020-12-15T16:09:48.684Z

Hi,

I generated Cerenkov photons using extended example OpNovice2, the macro below, and Geant4 version 10.7, and find
fVelocity = 225.40786315789472
which is correct.

/control/verbose 2
/tracking/verbose 2
/run/verbose 1
/process/optical/verbose 0
/run/numberOfThreads 1

/process/optical/processActivation Cerenkov      true
/process/optical/processActivation OpAbsorption  true
/process/optical/processActivation OpBoundary    true
/process/optical/processActivation Scintillation false
/process/optical/processActivation OpRayleigh    false
/process/optical/processActivation OpMieHG       false
/process/optical/processActivation OpWLS         false

/opnovice2/boxProperty RINDEX    0.000002 1.33    0.000008 1.33
/opnovice2/boxProperty ABSLENGTH 0.000002 1000000 0.000005 2000000 0.000008 3000000

/opnovice2/worldProperty RINDEX    0.000002 1.01    0.000008 1.01
/opnovice2/worldProperty ABSLENGTH 0.000002 1000000 0.000005 2000000 0.000008 3000000

/opnovice2/surfaceProperty TRANSMITTANCE 0.000002 0.0 0.000008 0.0
/opnovice2/surfaceProperty REFLECTIVITY  0.000002 1.0 0.000008 1.0
/opnovice2/surfaceProperty EFFICIENCY    0.000002 0.0 0.000008 0.0
/opnovice2/surfaceProperty RINDEX        0.000002 1.4 0.000008 1.5

/opnovice2/surfaceSigmaAlpha 0.2

/run/initialize
#
/gun/particle mu-
/gun/energy 2 GeV
/gun/position 0 0 0 cm
# will be normalized
/gun/direction .9 .44 0

/opnovice2/surfaceModel unified
/opnovice2/surfaceType dielectric_metal
/opnovice2/surfaceFinish polished
/run/beamOn 1

tome · 2020-12-16T13:10:37.252Z

Hi,
Many thanks for the feedback. I ran the OpNovice2 example with Geant4 version 10.5 and your macro and indeed found the correct velocity ! What’s puzzling in my code is how can the muons emit the Cherenkov photons at the correct angle (I checked that), meaning that the water refractive index is properly assigned, but these propagate at c ? Any hint ?

Thanks again,
Bernardo Tomé

dsawkey · 2020-12-16T14:38:23.034Z

Check to see that the material parameter GROUPVEL is present for the material. This should be automatically calculated when RINDEX is set.

You can call G4MaterialPropertiesTable::DumpTable().

In realtime, with gdb, get to the G4SteppingManager::Stepping frame, and print:

(gdb) p *((fStep->fpPostStepPoint->fpMaterial->fMaterialPropertiesTable)->GetProperty(9,false))

$38 = {<G4PhysicsVector> = {_vptr.G4PhysicsVector = 0x7ffff52110e8 <vtable for G4PhysicsOrderedFreeVector+16>, type = T_G4PhysicsOrderedFreeVector, edgeMin = 1.9999999999999999e-06, 
    edgeMax = 7.9999999999999996e-06, invdBin = 0, baseBin = 0, verboseLevel = 0, numberOfNodes = 2, dataVector = std::vector of length 2, capacity 2 = {225.40786315789472, 225.40786315789472}, 
    binVector = std::vector of length 2, capacity 2 = {1.9999999999999999e-06, 7.9999999999999996e-06}, secDerivative = std::vector of length 0, capacity 0, useSpline = false}, <No data fields>}

tome · 2020-12-16T15:41:10.429Z

Hi,

Indeed the GROUPVEL was not present. I found it had to do the way I was inserting the RINDEX values. I was doing this way,

MPT->AddProperty(“RINDEX”, new G4MaterialPropertyVector());
for (int ie=0; ie<nEntries; ie++) {
MPT->AddEntry(“RINDEX”, e[ie],RINDEX);
}

which apparently does not work.

Now, if I do :

G4MaterialPropertyVector* mpv = new G4MaterialPropertyVector();
for (int ie=0; ie<nEntries; ++ie) {
  mpv->InsertValues(e[ie], RINDEX);
}    

MPT->AddProperty("RINDEX", mpv);

I can now see that the material has the GROUPVEL parameter present (and the speed of optical photons in water is indeed c/1.33 !).
So, I would guess that the GROUPVEL is computed when AddProperty is called. In the first case the MPV is still empty at that stage…

But isn’t the syntax
MPT->AddEntry(“PROPERTY”, energy, value)

generally correct ? At least I was able to properly set the index of refraction …

Thanks !
Bernardo Tomé

tome · 2020-12-16T17:18:11.782Z

Hi again,

Looking at the code (G4MaterialPropertiesTable.cc) I find that the parameter GROUPVEL is indeed computed when AddProperty is called :

void G4MaterialPropertiesTable::
AddProperty(const char *key, G4MaterialPropertyVector *mpv)
{
// Provides a way of adding a property to the Material Properties
// Table given an G4MaterialPropertyVector Reference and a key
G4String k(key);
// if the key doesn’t exist, add it
if (std::find(G4MaterialPropertyName.begin(),
G4MaterialPropertyName.end(), k) ==
G4MaterialPropertyName.end()) {
G4MaterialPropertyName.push_back(k);
}
G4int index = GetPropertyIndex(k);
MP[ index ] = mpv;

// if key is RINDEX, we calculate GROUPVEL -
// contribution from Tao Lin (IHEP, the JUNO experiment)
if (k==“RINDEX”) {
CalculateGROUPVEL();
}
}

So, calling

MPT->AddProperty(“RINDEX”, new G4MaterialPropertyVector());

and then inserting the RINDEX values, with
MPT->AddEntry(“RINDEX”, e[ie],RINDEX);

will not compute the GROUPVEL because the method CalculateGROUPVEL() returns 0 if the RINDEX MPV exists but has no entries.
Maybe this should be handled in a different way ?..

Thanks !
Bernardo

dsawkey · 2020-12-18T13:34:53.702Z

You are correct. AddEntry() with RINDEX doesn’t update GROUPVEL. I’ll consider this a Geant4 bug. Thanks.

tome · 2020-12-18T15:58:14.877Z

Ok. Thanks again for the prompt feedback !