Forum: Space Applications Not Logged In (login)
Show subscribers

Space Applications

The email gateway for this forum is: space_app-g4hn@slac.stanford.edu


Inline Depth:
 0 0
 1 1
Add message: (add)

Question Normalization of an isotropic flux in space  by valentina <valentina>,   03 Mar, 2009
Dear all,

I am simulating the interaction of a space telescope with the cosmic ray environment and, after a long discussion with my colleagues, I am a bit confused on the normalization process that translates the number of counts in my detector in unity of counts/cm^2 sec.

So I really appreciate if somebody could help me to clarify this topic!

I show here what is my normalization way.

I have an isotropic proton flux F:

F = prot /cm^2 sec sr

I run the protons from a sphere of radius R within an half cone angle q to increase the statistics. My spacecraft is placed within the emission angle, that is the spacecraft is placed within the sphere of radius r, with

r = tan(q) x R

The GPS macro is:

/gps/particle proton

/gps/pos/type Surface

/gps/pos/shape Sphere

/gps/pos/radius 50.0 m ---> R

/gps/pos/centre 0.0 0.0 0.0 mm

/gps/ang/type iso

/gps/ang/mintheta 0 rad

/gps/ang/maxtheta 0.01 rad ---> q

while the spectral shape is made via /gps/hist .

At the end of the run, I detect a number of counts C on the focal plane. If N is the number of emitted protons, we call the fraction of detected/emitted particles Eff:

Eff = C/N

In order to trasform the number C to a rate P in counts/sec, I apply this formula:

P = Eff x F x (4 x PI) x PI x r^2 counts/sec

Then I divide the count rate P by the detecting area Adet, and the final flux B is given by:

B = P/Adet counts/cm^2 sec

The same normalization can be achieved using the simulation time:

D = F x A_sphere x omega prot/sec

where:

- A_sphere = 4 x PI x R^2 (the source sphere)

- omega = 2 x PI x (1 - cos(q)) (solide angle of the emission cone)

Then we calculate the simulated time:

Time = N/D (sec)

The flux B on the focal plane is given by:

B = C/(Time x Adet) counts/cm^2 sec

My normalization way is based on what shown by Sullivan (1971).

Now my question is: is this the right way?

Thank you all for the help!

Best regards, Valentina

None Re: Normalization of an isotropic flux in space  by Paul Nicholas Colin Gloster <Paul Nicholas Colin Gloster>,   03 Mar, 2009
Re: Question Normalization of an isotropic flux in space (valentina)
Valentina Fioretti submitted:
|----------------------------------|
|"[..]                             |
|                                  |
|I have an isotropic proton flux F:|
|                                  |
|F = prot /cm^2 sec sr"            |
|----------------------------------|

Dear Valentina Fioretti,

As you called that a flux, was that supposed to be
F = #protons / (centimeter^2 second steradian)
instead of what you seemed to type which was equivalent to
#protons second steradian / centimeter^2?


|------------------------------------------------------------------------|
|"[..]                                                                   |
|                                                                        |
|At the end of the run, I detect a number of counts C on the focal plane.|
|If N is the number of emitted protons, we call the fraction of          |
|detected/emitted particles Eff:                                         |
|                                                                        |
|Eff = C/N                                                               |
|                                                                        |
|In order to trasform the number C to a rate P in counts/sec, I apply    |
|this formula:                                                           |
|                                                                        |
|P = Eff x F x (4 x PI) x PI x r^2 counts/sec"                           |
|------------------------------------------------------------------------|

Eff can be treated as a dimensionless ratio (well strictly, its unit is count per proton). The unit of F, assuming F = #protons / (centimeter^2 second steradian), is count / (centimeter^2 second steradian).
Therefore, the unit of Eff x F is also count / (centimeter^2 second steradian).
For the unit of the P to be count/second, the unit of (4 x pi) x pi x r^2 would need to be centimeter^2 steradian.
It is not clear how you obtained the expression (4 x pi) x pi x r^2.



|-------------------------------------------------------------------------|
|"Then I divide the count rate P by the detecting area Adet, and the final|
|flux B is given by:                                                      |
|                                                                         |
|B = P/Adet counts/cm^2 sec"                                              |
|-------------------------------------------------------------------------|

If the unit of P is count/second then the unit of B is count/(cm^2 second), which you seemed to realize but you typed something else.


|--------------------------------------------------------------------|
|"The same normalization can be achieved using the simulation time:  |
|                                                                    |
|D = F x A_sphere x omega prot/sec                                   |
|                                                                    |
|where:                                                              |
|                                                                    |
|- A_sphere = 4 x PI x R^2 (the source sphere)                       |
|                                                                    |
|- omega = 2 x PI x (1 - cos(q)) (solide angle of the emission cone)"|
|--------------------------------------------------------------------|

The solid angle subtended by the detector at the source of the protons depends at least on the detector surface that faces the source and the distance between the detector and the source. You do not seem to have informed us of the distance. Omega = 2 x pi x (1 - cos(q)) may or may not apply for your situation.
For the correct value of Omega, the unit of F x A_sphere x Omega is indeed protoncount/second.



|--------------------------------------|
|"Then we calculate the simulated time:|
|                                      |
|Time = N/D (sec)"                     |
|--------------------------------------|

Agreed.



|-------------------------------------------|
|"The flux B on the focal plane is given by:|
|                                           |
|B = C/(Time x Adet) counts/cm^2 sec"       |
|-------------------------------------------|

Yes (with the unit being count/(centimeter^2 second)).


|----------------------------------------------------------------|
|"My normalization way is based on what shown by Sullivan (1971).|
|                                                                |
|Now my question is: is this the right way?                      |
|                                                                |
|[..]"                                                           |
|----------------------------------------------------------------|

Bear in mind that the flux reaching the detector (as opposed to merely the spacecraft) would be affected by the corresponding geometric factor of the detector, which can be affected by the particles' trajectories through the instrument (therefore different energies may have different geometric factors) and the geometry of the instrument.

It seems that you referred to Sullivan, J. D., 1971, "Geometrical factor and directional response of single and multi-element particle telescopes". I do not have a copy of this and I do not know enough about what you are trying to simulate so I do not say much more. You may wish to read
Sullivan, J. D.,
1972,
"Erratum: Geometrical factor and directional response of single and multi-element particle telescopes, Nucl. Instr. and Meth. 95 (1971) 5-11"
,
HTTP://DX.DOI.org/10.1016/0029-554X(72)90450-8
, "Nuclear Instruments and Methods", Volume 98, Issue 1, Page 187.

Regards,
Colin Paul Gloster
Feedback Re: Normalization of an isotropic flux in space  by Giovanni Santin <Giovanni Santin>,   03 Mar, 2009
Re: Question Normalization of an isotropic flux in space (valentina)
Dear Valentina

the paper by Sullivan is indeed a good starting point for normalisation issues.

I tend to agree on your 1st final formula.

I see a problem in your macro, as you use "iso" for the angular distribution. In fact, if you want to generate an isotropic source in a volume starting from a surface, you have to bias the angular distribution with a cosine-law function, that means using "cos" in GPS.

Some time ago I tried to summarise all this in some diagrams, you can find a presentation at: http://geant4.in2p3.fr/2007/prog/GiovanniSantin/GSantin_Geant4_Paris07_Normalisation_v07.ppt

There, the final formula I get at slide #10 is equivalent to your first one, if you substitute Xs=C/Adet , q=thetamax , F=Phi

At the last slide, I include a couple more useful references.

Also, it seems to me that the 2nd formula you propose is a factor 2 different from the 1st one (am I right?). Indeed, that might come from the fact that you should integrate the cosine-biased emission, not an isotropic one.

In my calculations, also, if I limit the max emission angle theta to a certain q, I normally get a sin2(q) instead of (1 - cos q), as a result of the integration of the cos-biased distribution.

I hope this can help you, and I am interested in comments from your side

Kind regards, Giovanni

Question Re: Normalization of an isotropic flux in space  by valentina <valentina>,   03 Mar, 2009
Re: Feedback Re: Normalization of an isotropic flux in space (Giovanni Santin)
Dear Giovanni,

thank you for your quick reply!

The fact that you agree with my approach calm me down a lot.

>I see a problem in your macro, as you use "iso" for the angular distribution. In >fact, if you want to generate an isotropic source in a volume starting from a >surface, you have to bias the angular distribution with a cosine-law function, >that means using "cos" in GPS.

Yes, you'r right. The reason is that I am running a cross-checking simulation, and my colleagues use "iso"... I will change it.

>Also, it seems to me that the 2nd formula you propose is a factor 2 different >from the 1st one (am I right?). Indeed, that might come from the fact that you >should integrate the cosine-biased emission, not an isotropic one.
>In my calculations, also, if I limit the max emission angle theta to a certain q, >I normally get a sin2(q) instead of (1 - cos q), as a result of the integration >of the cos-biased distribution.

Yes, the 2nd formula is a more intuitive but approximated approach. Since q is this case a very small angle, the approximation does not affect the result.

From the 1st formula, the simulation time is given by:

Time_1 = N / [F(prot/(cm2 s sr)) x (4 x PI) x (PI x (R x tan(q))^2)]

From the 2nd formula, the time is given by:

Time_2 = N / [F x (4. x PI x R^2) x (2 x PI x (1 - cos(q)))]

For q = 0.01 rad, Time_1 is about Time_2.

If a take q = 0.9, Time_1 is equal to Time_2 for the following formulas:

Time_1 = N / [F(prot/(cm2 s sr)) x (4 x PI) x (PI x (R x sin(q))^2)]

Time_2 = N / [F x (4. x PI x R^2) x (2 x PI x (1/2 - (cos(q)^2)/2))]

(I take the cosin-law function)

Apart from the differences with these two formulas, my problem is that another normalization approach has been presented to me.

Since the intersection of infinite cones originating from the source sphere (radius R) would give the surface of the small sphere of radius r, the simulation time is given by:

Time_B = N/[F(prot/(cm2 s sr)) x (4 x PI) x (4 x PI x (R x sin(q))^2)]

Note the factor 4, that gives the area of the sphere surface and not of the plane of radius r.

The simulation time is 4 times lower than "my" simulation time.

Consider of having an isotropic flux, integrated over 4PI, of:

F = 2.31 #protons / (cm^2 sec)

If we run a simulation with a simple geometry made of a total absorbing cube placed within the small sphere (of radius r), what is the proton count rate that you expect on the cube surface?

If you use the last normalization, you obtain 2.31 counts/(cm^2 sec). My objection is that the proton count rate on the cube surface is not the flux integrated over 4PI, but (using the cosin law) over PI, because every cube side sees only PI sr.

What do you think of this different normalization? And, the most important, how one could prove the accuracy of one normalization approach respect to the other?

Thank you Giovanni!

Best regards, Valentina

Note Re: Normalization of an isotropic flux in space  by Juan <Juan>,   04 Mar, 2009
Re: Question Re: Normalization of an isotropic flux in space (valentina)
Dear Valentina

    If a take q = 0.9, Time_1 is equal to Time_2 for the following formulas:
    Time_1 = N / [F(prot/(cm2 s sr)) x (4 x PI) x (PI x (R x sin(q))2)]
    Time_2 = N / [F x (4. x PI x R2) x (2 x PI x (1/2 - (cos(q)2)/2))]

Notice this seem normal to me as both equations are the same

Time_1 = N/F x 4 PI^2 x R^2 x sin(p)^2
Time_2 = N/F x 4 PI^2 x R^2 x 2 x 1/2 x (1-cos(p)^2) =
      = N/F x 4 PI^2 x R^2 x sin(p)^2

From what Giovanni shows in his presentation, if I understand well, when you have a sphere of radius r as sensitive volume inside a sphere of radius R which is surrounded by an isotropic flux F

Nr = F 4PI^2 R^2

is the number of particles per second traversing the big sphere and entering in this volume. So in G4 if you make beamOn Nr you simulate the number of particles entering in the big volume in one second.

When you only need the particles hitting the little sphere, to save time, you look only the particles in the big sphere having a incident angle lower than q where r = R sin(q) this gives the rate of particles entering in the little volume per second in page 11 of Giovanni's presentation.

Nr = F 4PI^2 R^2 sin(q)^2

For me this is equal to

Nr = F 4PI^2 r^2

Is like if you put the source on the surface of the little sphere without limitation of angle q. And if you do beamOn Nr you simulate the number of particles entering the little sensitive volume in one second. beamOn 100xNr will simulate 100 seconds, etc.

If I want to know how many particles will enter a little cube of side L per second. What I will do is to put the cube inside a sphere of radius

r = sqrt(3)/2 L

Which is 1/2 of the big diagonal of your cube. So you have the smaller sphere surrounding your cube. As in precedent case I will use

Nr = F 4PI^2 r^2

Then, doing beamOn Nr, I will count in my GEANT4 program the number of particles entering the cube per second Nc. The geometric factor of the cube omega can be found as

Nc = F omega

dividing both equations :

Nc/Nr = F/F omega/(4PI^2 r^2)

omega = Nc/Nr (4PI^2 r^2)

This geometrical factor gives you the number or particles entering in the cube per second when you know the external flux. The geometric factor is independent of the particle energy. Now, if you have a detector perhaps you have to take in account the detection efficiency which depends of the incoming energy.

Juan Cabrera

Ok Re: Normalization of an isotropic flux in space  by valentina <valentina>,   04 Mar, 2009
Re: Note Re: Normalization of an isotropic flux in space (Juan)
Dear Juan,

thank you for your reply!

> 
>     If a take q = 0.9, Time_1 is equal to Time_2 for the following formulas:
>     Time_1 = N / [F(prot/(cm2 s sr)) x (4 x PI) x (PI x (R x sin(q))2)]
>     Time_2 = N / [F x (4. x PI x R2) x (2 x PI x (1/2 - (cos(q)2)/2))]
> 
> Notice this seem normal to me as both equations are the same
> 
> Time_1 = N/F x 4 PI^2 x R^2 x sin(p)^2
> Time_2 = N/F x 4 PI^2 x R^2 x 2 x 1/2 x (1-cos(p)^2) =
>       = N/F x 4 PI^2 x R^2 x sin(p)^2
> 

Yes, they are the same in the exact formulation. Giovanni made me notice in the previous message that I was using an approximated formulation, that doesn't work for larger angles.

> Then, doing beamOn Nr, I will count in my GEANT4 program the number of
> particles entering the cube per second Nc. The geometric factor of the
> cube omega can be found as
> 
> Nc = F omega
>

Since F is [#protons / (cm^2 sec sr)], here you forgot to mention the cube area (the detecting area):

Nc = F x omega x A_cube

> dividing both equations :
> 
> Nc/Nr = F/F omega/(4PI^2 r^2)
> 

Nc/Nr = F/F (omega x A_cube) /(4PI^2 r^2)

> omega = Nc/Nr (4PI^2 r^2)
> 

omega = (Nc/Nr) x (4PI^2 r^2) x A_cube

If you try to run this simulation test, you obtain omega = PI, that is just what I am saying.

I am glad to see that you agree!

Shifting from the simulation to a real case, if you have an isotropic flux:

F = # protons/ (cm^2 sec sr)

What is the flux that you expect on the cube? From the two approaches that I have presented, two are the options:

1) F x omega = F x PI (my normalization way)

2) F x 4PI

Or, if you have an ideal (detection efficiency = 1) detector in space exposed to a 4PI isotropic flux, what is the detected count rate in counts/(cm2 sec)?

1) F x PI?

2) F x 4PI?

What do you think?

Thanks!

Warm regards, Valentina

None Re: Normalization of an isotropic flux in space  by Juan <Juan>,   04 Mar, 2009
Re: Ok Re: Normalization of an isotropic flux in space (valentina)
Dear Valentina,

I should call the geometric factor Gf (cm2 sr) not omega which is used for solid angle (sr). This can do a bad interpretation of the formula I wrote

Nc = F Gf

Which gives the number of particles per second entering in the cube.

From Giovanni presentation page 10 Method 1, if I'm not wrong, the Gf can be calculate multiplying PI x Surface. For a cube the surface presented to the flux is 6 L^2.

Then, the number of particles per second is

Nc = F (PI 6 L ^2)

Now, if you want to have the number of particles entering in the cube per cm2 per second you divide by the cube area

Nc/cm2 = F PI

Which is what you want.

Juan

Question gps sphere source   by Daniel Braun <Daniel Braun>,   09 Oct, 2008
Hallo,

I need to use a sphere as a particle source with the gps class.

and I want the particles to shoot just in the inside of the sphere maximally tangential to the sphere. They also should be distributed even on the surface of the sphere.

What is the best way to realize this?

Thanks for any help.

Regards Daniel

Feedback Re: gps sphere source   by Giovanni Santin <Giovanni Santin>,   10 Oct, 2008
Re: Question gps sphere source (Daniel Braun)
Hello Daniel,

> I need to use a sphere as a particle source with the gps class.
> 
> and I want the particles to shoot just in the inside of the sphere

GPS is indeed a good choice, it will make your life easy.
For generating particles on the surface of a sphere, you can insert this in your macro:

/gps/pos/type Surface
/gps/pos/shape Sphere
/gps/pos/radius 2.0 m
/gps/pos/centre 0. 0. 0. cm

> They also should be distributed even on the surface of the sphere.

The vertices will be automatically uniformly distributed on the surface.

For the angular distribution:
During the run, for each point, randomly generated on the surface, GPS orients +z of the "generation frame" as normal to the surface, directed towards inside.

Normally, then, for generating particles inside the sphere to reproduce an isotropic angular distribution, one has to specify a "cosine law" biasing, and let the theta go from 0 to 90 degrees:

/gps/ang/type cos
/gps/ang/mintheta 0 deg
/gps/ang/maxtheta 90 deg

I am not sure what you mean with 

> maximally tangential to the sphere. 

If you really want the particles to only start parallel to the surface, you can just modify the mintheta to 90 degrees:

/gps/ang/mintheta 90 deg
/gps/ang/maxtheta 90 deg

If instead you mean only normal to the surface (i.e. directed to the centre of the generation sphere) you set maxtheta to 0 degrees.

/gps/ang/mintheta 0 deg
/gps/ang/maxtheta 0 deg

I hope this helps
Kind regards,
Giovanni

ESA / ESTEC
Space Environments and Effects
e-mail: Giovanni.Santin@esa.int

None Re: gps sphere source   by Daniel Braun <Daniel Braun>,   10 Oct, 2008
Re: Feedback Re: gps sphere source (Giovanni Santin)
Hallo Giovanni, thank you very much for your answer.

I tried everything like you explained it but I think there still must be a fault somewhere in my macro.

To check if the particles are all going inside the sphere I made a Detector (green Sphere in the pictures) which has the size of the spherical source. I gave the particles a very small energy so they should stop immediately if they have contact with the Material of the detector. I suspected to get a lot of very short trajectories. Especially for the case of mintheta = 0 and maxtheta = 0. But instead I get trajectories which I can't explain.

In the case of mintheta = 0 and maxtheta = 0 they seem to come just from the middle and in the case of mintheta = 0 and maxtheta = 90 it looks like they are coming form one hemisphere.

I attached some pictures and my macro.

Pleas tell me where I make a mistake.

Regards

Daniel

/control/verbose 0

/control/saveHistory

/run/verbose 0

/event/verbose 0

/vis/scene/create

/vis/open OGLIX

/vis/viewer/zoom 0.4

/vis/scene/add/trajectories

/vis/modeling/trajectories/create/drawByCharge

/vis/modeling/trajectories/drawByCharge-0/default/setDrawStepPts true

/vis/modeling/trajectories/drawByCharge-0/default/setStepPtsSize 2

/vis/scene/endOfEventAction accumulate

/vis/viewer/set/viewpointThetaPhi 0 0 deg

/vis/scene/endOfEventAction accumulate 10000000

/gps/particle proton

/gps/pos/centre 0 0 0

/gps/pos/type Surface

/gps/pos/shape Sphere

/gps/pos/radius 116 mm

/gps/pos/rot1 0 0 1

/gps/ene/type Mono

/gps/energy 0.00000570E-02 eV

/gps/ang/type cos

/gps/ang/mintheta 0 deg

/gps/ang/maxtheta 90 deg

# or /gps/ang/maxtheta 0 deg

/run/beamOn 3000

   Attachment:
      http://hypernews.slac.stanford.edu/HyperNews/geant4/get/AUX/2008/10/10/03.59-14287-0_0.eps
      http://hypernews.slac.stanford.edu/HyperNews/geant4/get/AUX/2008/10/10/03.59-30858-0_90.epsOn Fri, 10 Oct 2008 09:08:24 GMT, Giovanni Santin wrote:

None Re: gps sphere source   by Daniel Braun <Daniel Braun>,   30 Oct, 2008
Re: None Re: gps sphere source (Daniel Braun)
Hello,

I also found a mistake in my geometry.

My Detector was lightly shifted which caused a asymmetry.

After removing the rotation of the source and the translation of the detector, everything looks fine.

Thanks for your help Giovanni

Regards

Daniel

Feedback Re: gps sphere source   by Giovanni Santin <Giovanni Santin>,   16 Oct, 2008
Re: None Re: gps sphere source (Daniel Braun)
Hi Daniel

I could indeed reproduce some problematic particle generation cases.
After a few trials with your macro, it seems to me that the 
problem is caused by the command

/gps/pos/rot1 0 0 1 

If I remove it GPS behaves as expected, and for example with maxtheta=0
I see all particles directed towards the centre of the generation sphere.

I will have a look at the GPS code and doc to better understand this 
"feature" of the rot1/rot2 commands.

Thanks for reporting this
Kind regards,
Giovanni

ESA / ESTEC
Space Environments and Effects
e-mail: Giovanni.Santin@esa.int
Feedback Re: gps sphere source   by Giovanni Santin <Giovanni Santin>,   10 Oct, 2008
Re: Question gps sphere source (Daniel Braun)
Hello Daniel,

> I need to use a sphere as a particle source with the gps class.
> 
> and I want the particles to shoot just in the inside of the sphere

GPS is indeed a good choice, it will make your life easy.
For generating particles on the surface of a sphere, you can insert this in your macro:

/gps/pos/type Surface
/gps/pos/shape Sphere
/gps/pos/radius 2.0 m
/gps/pos/centre 0. 0. 0. cm

> They also should be distributed even on the surface of the sphere.

The vertices will be automatically uniformly distributed on the surface.

For the angular distribution:
During the run, for each point, randomly generated on the surface, GPS orients +z of the "generation frame" as normal to the surface, directed towards inside.

Normally, then, for generating particles inside the sphere to reproduce an isotropic angular distribution, one has to specify a "cosine law" biasing, and let the theta go from 0 to 90 degrees:

/gps/ang/type cos
/gps/ang/mintheta 0 deg
/gps/ang/maxtheta 90 deg

I am not sure what you mean with 

> maximally tangential to the sphere. 

If you really want the particles to only start parallel to the surface, you can just modify the mintheta to 90 degrees:

/gps/ang/mintheta 90 deg
/gps/ang/maxtheta 90 deg

If instead you mean only normal to the surface (i.e. directed to the centre of the generation sphere) you set maxtheta to 0 degrees.

/gps/ang/mintheta 0 deg
/gps/ang/maxtheta 0 deg

I hope this helps
Kind regards,
Giovanni

ESA / ESTEC
Space Environments and Effects
e-mail: Giovanni.Santin@esa.int

Question dose depth curve  by Daniel Braun <Daniel Braun>,   12 Sep, 2008
Hello everybody,

I tried to get a Dose depth curve in an Aluminium Sphere and compared it with the one I get from www.Spenvis.oma.be.

But for a thin shield I get not enough Energy deposited and for a thick shield I get to much Energy deposited.

Does anybody know what the Problem is

My Physiks are

// // // $Id: ExN03PhysicsList.cc,v 1.21 2007/07/02 13:22:08 vnivanch Exp $ // GEANT4 tag $Name: geant4-09-01-patch-02 $ // //

//....oooOO0OOooo........oooOO0OOooo........oooOO0OOooo........oooOO0OOooo...... //....oooOO0OOooo........oooOO0OOooo........oooOO0OOooo........oooOO0OOooo......

#include "ExN03PhysicsList.hh"

#include "G4ProcessManager.hh"
#include "G4ParticleTypes.hh"

//....oooOO0OOooo........oooOO0OOooo........oooOO0OOooo........oooOO0OOooo......

ExN03PhysicsList::ExN03PhysicsList():  G4VUserPhysicsList()
{
  defaultCutValue = 1*mm;
  SetVerboseLevel(1);
}

//....oooOO0OOooo........oooOO0OOooo........oooOO0OOooo........oooOO0OOooo......

ExN03PhysicsList::~ExN03PhysicsList() {}

//....oooOO0OOooo........oooOO0OOooo........oooOO0OOooo........oooOO0OOooo......

void ExN03PhysicsList::ConstructParticle()
{
  // In this method, static member functions should be called
  // for all particles which you want to use.
  // This ensures that objects of these particle types will be
  // created in the program. 

  ConstructBosons();
  ConstructLeptons();
  ConstructMesons();
  ConstructBaryons();
}

//....oooOO0OOooo........oooOO0OOooo........oooOO0OOooo........oooOO0OOooo......

void ExN03PhysicsList::ConstructBosons()
{
  // pseudo-particles
  G4Geantino::GeantinoDefinition();
  G4ChargedGeantino::ChargedGeantinoDefinition();

  // gamma
  G4Gamma::GammaDefinition();

  // optical photon
  G4OpticalPhoton::OpticalPhotonDefinition();
}

//....oooOO0OOooo........oooOO0OOooo........oooOO0OOooo........oooOO0OOooo......

void ExN03PhysicsList::ConstructLeptons()
{
  // leptons
  G4Electron::ElectronDefinition();
  G4Positron::PositronDefinition();
  G4MuonPlus::MuonPlusDefinition();
  G4MuonMinus::MuonMinusDefinition();

  G4NeutrinoE::NeutrinoEDefinition();
  G4AntiNeutrinoE::AntiNeutrinoEDefinition();
  G4NeutrinoMu::NeutrinoMuDefinition();
  G4AntiNeutrinoMu::AntiNeutrinoMuDefinition();
}

//....oooOO0OOooo........oooOO0OOooo........oooOO0OOooo........oooOO0OOooo......

void ExN03PhysicsList::ConstructMesons()
{
 //  mesons
  G4PionPlus::PionPlusDefinition();
  G4PionMinus::PionMinusDefinition();
  G4PionZero::PionZeroDefinition();
  G4Eta::EtaDefinition();
  G4EtaPrime::EtaPrimeDefinition();
  G4KaonPlus::KaonPlusDefinition();
  G4KaonMinus::KaonMinusDefinition();
  G4KaonZero::KaonZeroDefinition();
  G4AntiKaonZero::AntiKaonZeroDefinition();
  G4KaonZeroLong::KaonZeroLongDefinition();
  G4KaonZeroShort::KaonZeroShortDefinition();
}

//....oooOO0OOooo........oooOO0OOooo........oooOO0OOooo........oooOO0OOooo......

void ExN03PhysicsList::ConstructBaryons()
{
  //  barions
  G4Proton::ProtonDefinition();
  G4AntiProton::AntiProtonDefinition();
  G4Neutron::NeutronDefinition();
  G4AntiNeutron::AntiNeutronDefinition();
}

//....oooOO0OOooo........oooOO0OOooo........oooOO0OOooo........oooOO0OOooo......

void ExN03PhysicsList::ConstructProcess()
{
  AddTransportation();
  ConstructEM();
  ConstructDecay();
}

//....oooOO0OOooo........oooOO0OOooo........oooOO0OOooo........oooOO0OOooo......

#include "G4ComptonScattering.hh"
#include "G4GammaConversion.hh"
#include "G4PhotoElectricEffect.hh"

#include "G4MultipleScattering.hh"

#include "G4eIonisation.hh"
#include "G4eBremsstrahlung.hh"
#include "G4eplusAnnihilation.hh"

#include "G4MuIonisation.hh"
#include "G4MuBremsstrahlung.hh"
#include "G4MuPairProduction.hh"

#include "G4hIonisation.hh"

#include "G4LowEnergyIonisation.hh" 
#include "G4LowEnergyBremsstrahlung.hh" 

//....oooOO0OOooo........oooOO0OOooo........oooOO0OOooo........oooOO0OOooo......

void ExN03PhysicsList::ConstructEM()
{
  theParticleIterator->reset();
  while( (*theParticleIterator)() ){
    G4ParticleDefinition* particle = theParticleIterator->value();
    G4ProcessManager* pmanager = particle->GetProcessManager();
    G4String particleName = particle->GetParticleName();

    if (particleName == "gamma") {
      // gamma         
      pmanager->AddDiscreteProcess(new G4PhotoElectricEffect);
      pmanager->AddDiscreteProcess(new G4ComptonScattering);
      pmanager->AddDiscreteProcess(new G4GammaConversion);
      //pmanager->AddProcess(new G4eBremsstrahlung,   -1, 3,3);

    } else if (particleName == "e-") {
      //electron
      pmanager->AddProcess(new G4MultipleScattering,-1, 1,1);
      pmanager->AddProcess(new G4eIonisation,       -1, 2,2);
      pmanager->AddProcess(new G4eBremsstrahlung,   -1, 3,3);      

    } else if (particleName == "e+") {
      //positron
      pmanager->AddProcess(new G4MultipleScattering,-1, 1,1);
      pmanager->AddProcess(new G4eIonisation,       -1, 2,2);
      pmanager->AddProcess(new G4eBremsstrahlung,   -1, 3,3);
      pmanager->AddProcess(new G4eplusAnnihilation,  0,-1,4);

    } else if( particleName == "mu+" || 
               particleName == "mu-"    ) {
      //muon  
      pmanager->AddProcess(new G4MultipleScattering,-1, 1,1);
      pmanager->AddProcess(new G4MuIonisation,      -1, 2,2);
      pmanager->AddProcess(new G4MuBremsstrahlung,  -1, 3,3);
      pmanager->AddProcess(new G4MuPairProduction,  -1, 4,4);       

    } else if ((!particle->IsShortLived()) &&
	       (particle->GetPDGCharge() != 0.0) && 
	       (particle->GetParticleName() != "chargedgeantino")) {
      //all others charged particles except geantino
      pmanager->AddProcess(new G4MultipleScattering,-1, 1,1);
      pmanager->AddProcess(new G4hIonisation,       -1, 2,2); 

    }
  }
}

//....oooOO0OOooo........oooOO0OOooo........oooOO0OOooo........oooOO0OOooo......

#include "G4Decay.hh"

void ExN03PhysicsList::ConstructDecay()
{
  // Add Decay Process
  G4Decay* theDecayProcess = new G4Decay();
  theParticleIterator->reset();
  while( (*theParticleIterator)() ){
    G4ParticleDefinition* particle = theParticleIterator->value();
    G4ProcessManager* pmanager = particle->GetProcessManager();
    if (theDecayProcess->IsApplicable(*particle)) { 
      pmanager ->AddProcess(theDecayProcess);
      // set ordering for PostStepDoIt and AtRestDoIt
      pmanager ->SetProcessOrdering(theDecayProcess, idxPostStep);
      pmanager ->SetProcessOrdering(theDecayProcess, idxAtRest);
    }
  }
}

//....oooOO0OOooo........oooOO0OOooo........oooOO0OOooo........oooOO0OOooo......

void ExN03PhysicsList::SetCuts()
{
  if (verboseLevel >0){
    G4cout << "ExN03PhysicsList::SetCuts:";
    G4cout << "CutLength : " << G4BestUnit(defaultCutValue,"Length") << G4endl;
  }

  // set cut values for gamma at first and for e- second and next for e+,
  // because some processes for e+/e- need cut values for gamma
  //

G4double lowlimit=250*eV;
  G4ProductionCutsTable::GetProductionCutsTable()
    ->SetEnergyRange(lowlimit, 100.*GeV);

  // default cuts for world volume
  G4double cutValue = 1*mm;
  SetCutValue(cutValue,"gamma");
  SetCutValue(cutValue,"e-");
  SetCutValue(cutValue,"e+");

  if (verboseLevel>0) DumpCutValuesTable();
}

//....oooOO0OOooo........oooOO0OOooo........oooOO0OOooo........oooOO0OOooo......

From my Detector Construction

I use aluminium as shield and Silicon as detector Material defined like this.

new G4Material("Aluminium", z=13., a=26.98*g/mole, density=2.70*g/cm3); new G4Material("Silicon" , z=14., a= 28.09*g/mole, density= 2.33*g/cm3);

I'm new with geant4 and don't have a clue where the mistake could be.

Thanks for any help

Kind regards

Daniel

None Re: dose depth curve  by michel maire <michel maire>,   12 Sep, 2008
Re: Question dose depth curve (Daniel Braun)
On Fri, 12 Sep 2008 13:45:02 GMT, Daniel Braun wrote:
> Hello everybody,
> 
> I tried to get a Dose depth curve in an Aluminium Sphere and compared it
> with the one I get from www.Spenvis.oma.be.
> 
> But for a thin shield I get not enough Energy deposited and for a thick
> shield I get to much Energy deposited.
> 

  You may be interested by examples/extended/electromagntic/TestEm12

       Michel

None Re: dose depth curve  by Daniel Braun <Daniel Braun>,   12 Sep, 2008
Re: None Re: dose depth curve (michel maire)
Thank you for your answer.

I had a look at the example you recommended but as I get it right, my Situation is a bit different. It is shown in the geometry file I attached in Case 3.

I'm shoot a fluence of particles ,I calculated with Spenvis, from the outside on the sphere.

And I want to get a curve like the one I calculated with Spenvis (Ionizing and non-ionizing dose models for simple geometries). I also attached the curve (sd2_dose).

I modeled the same geometry in Geant4 and shoot the same particles on it. But the result isn't the same. But I guess it should be.

Is there a way to create a model with G4 that delivers the same results as Spenvis?

   Attachment:
      http://hypernews.slac.stanford.edu/HyperNews/geant4/get/AUX/2008/09/12/09.28-98770-geometry.gif
      http://hypernews.slac.stanford.edu/HyperNews/geant4/get/AUX/2008/09/12/09.28-68716-sd2_dose.jpg

None Re: dose depth curve  by Hugh Evans <Hugh Evans>,   15 Sep, 2008
Re: None Re: dose depth curve (Daniel Braun)
On Fri, 12 Sep 2008 16:46:49 GMT, Daniel Braun wrote:
> Thank you for your answer.
> 
> I had a look at the example you recommended but as I get it right, my
> Situation is a bit different. It is shown in the geometry file I
> attached in Case 3.

It must be remembered that the shieldose spherical shell method is derived from an analytical expression from the slab geometry. (See the various Shieldose references from the SPENVIS website).

Reproducing the spherical shell results is prone to difficulties.

Another difference in the simulation to consider: shieldose gives results for infinitesimally small targets (Si, SiO2, etc.). The target must be "insignificantly" small, yet you have to be able to hit it enough to get good statistics.

Additionally, the dose for the slab geometries is calculated from the shielded flux spectrum exiting the slab thickness folded with the dE/dX curve of the target - the target itself is not included in the simulation.

> 
> I'm shoot a fluence of particles ,I calculated with Spenvis, from the
> outside on the sphere.

If it is a simple result that you wish, then I'd recommend using Mulassis (also available to run on SPENVIS, and source code/local installation is available from the QinetiQ site). I've made comparable simulations to SHIELDOSE using Mulassis (particularly for protons, electrons requires some playing with the energy biasing to get a good statistical sampling of the effects of the various energy ranges).

> 
> And I want to get a curve like the one I calculated with Spenvis
> (Ionizing and non-ionizing dose models for simple geometries). I also
> attached the curve (sd2_dose).

You do realise that for the slab or spherical geometry that you have to run the complete simulation for each of the shielding thicknesses in the curve? This can be a bit tedious (and slow) :-(.

> 
> I modeled the same geometry in Geant4 and shoot the same particles on
> it. But the result isn't the same. But I guess it should be.
> 
> Is there a way to create a model with G4 that delivers the same results
> as Spenvis?
> 

Yes. but the simulation isn't an exact match to the SHIELDOSE geometry.

Regards, Hugh

None Re: dose depth curve  by Daniel Braun <Daniel Braun>,   16 Sep, 2008
Re: None Re: dose depth curve (Hugh Evans)
Thank you for your Answer.

Is there a simpler way to check the G4 results?

I have to know whether I use the right physics and the right gps settings and so on.

Thanks for your help.

Greetings

Daniel

None Re: dose depth curve  by Hugh Evans <Hugh Evans>,   17 Sep, 2008
Re: None Re: dose depth curve (Daniel Braun)
Hi Daniel,

The quickest way to get the space environment flux spectra into your simulation is to use the GPS macros created by SPENVIS for its version of Mulassis. These GPS macros you can then add to your own software (the /analysis/normalise factor must also be included to normalise the results of the simulation to those of the environment, but it may require adjustments to account for variations between the geometry of the GPS source in your own code and what's in Mulassis).

The quickest way to check that the G4 simulation is giving comparable results to SHIELDOSE is by running the SHIELDOSE slab geometry and the corresponding Mulassis Slab geometry for a given shielding thickness, you can then compare the dose results directly with the Mulassis simulation on SPENVIS.

As for the Physics: the SHIELDOSE model does include nuclear interactions for protons; for electrons the "em" physics should be sufficient; to a reasonable approximation the "em" physics is sufficient for protons as well, but you can add "binary cascade" to include nuclear interactions to the results.

Regards, Hugh

None Re: dose depth curve  by Daniel Braun <Daniel Braun>,   22 Sep, 2008
Re: None Re: dose depth curve (Hugh Evans)
Thank you very much for your answer.

I tried now to compare the results of Geant 4 and Mulassis and it matched very well.

Regards

Daniel

None SPE Primary Particles   by Steven Avery <Steven Avery>,   21 Apr, 2008

I'm using the radioprotection example but I would like to only use solar particle events (SPE) as my primary particle. The example only has ascii files for galactic cosmic rays (GCR) data. How do I set the example to only generate SPE events?

Feedback Re: SPE Primary Particles   by Giovanni Santin <Giovanni Santin>,   21 Apr, 2008
Re: None SPE Primary Particles (Steven Avery)

Dear Steven

A few suggestion:

I would recommend to use the General Particle Source (GPS) to generate your particle spectra. It's fairly simple to use, you can find a link to its detailed documentation from http://geant4.web.cern.ch/geant4/support/index.shtml

Regarding SPE spectra, it depends if you want to simulate individual solar events or the total proton fluence in a certain mission period.

For the latter, for example, you can get the total fluence predicted by the ESP model through the SPENVIS interface: http://www.spenvis.oma.be/ After you have registered (it's open and free) you look at the MULASSIS module. At the end of the user input process, you can get directly a Geant4 macro with the spectrum in GPS format.

[If you prefer to stay with the PrimaryGenerator from the radioprotection examples, you can get the ESP differential spectra from SPENVIS anyway. But I would recommend the GPS because of the many features it offers]

Please have a look as well at the 2 PPTs on "Primary particle" at http://geant4.in2p3.fr/2007/prog/prog.htm

Don't hesitate to contact me for further details

Best regards

Giovanni

Question Grazing angle proton scattering  by valentina <valentina>,   13 Feb, 2008

Hi, I have a question about the simulation of grazing angle proton scattering.

In the article:

"Low Angle Scattering of Protons on the XMM-Newton Optics and Effects on the On-Board CCD detectors"

Nartallo et al. 2001

Nartallo uses the multiple scattering to simulate the interaction of protons with a mirror reflecting surface layer (in this case XMM-Newton optics).

As a conseguence of:

"Grazing Angle Proton Scattering: effects on Chandra and XMM-Newton X-ray telescopes"

Dichter and Woolf, 2003

, where the authors show that the multiple scattering under predicts the expected proton flux at the focal plane,

Nartallo in the article:

"Update on the use of Geant4 for the simulation of low energy protons scattering off X-ray mirrors at grazing incidence angles"

implements a new physics process, G4FirsovScattering, in the G4 Monte Carlo code, to treat the scattering of low energy protons at grazing incidence angles.

Now the question is: someone knows where is G4FirsovScattering? I can't find it in G4 library!

Thanks!

Valentina

None Generating a Pulse height spectra through GEANT  Keywords: Getting a Pulse height spectra from GEANT
by rajani <rajani>,   02 Feb, 2007

Hi All!!

In a normal space experiment, the particles ( suppose photons here) interact with the detector, giving rise to the absorption or transmission of the particle. In GEANT, we can trace the physics processes it undergoes, the energy it carries after that & the step at which it does that. In a Lab- experiment, what we get out is a Pulse Height Spectra that tells the variation of no of counts with the voltage, My concern is, Can GEANT in any way tell me about the Pulse height spectra? If yes, Can you tell me how & the reference? Thanking You

Rajani

None Re: Generating a Pulse height spectra through GEANT  Keywords: Getting a Pulse height spectra from GEANT
by Maria Grazia Pia <Maria Grazia Pia>,   02 Feb, 2007
Re: None Generating a Pulse height spectra through GEANT (rajani)

Dear Rajani,

Geant4 provides tools to generate a detector response, such as, for instance, pulse height spectra. This simulation domain is called "digitisation"; the Geant4 digits_hits package is responsible for it.

I suggest you to read the corresponding sections in the Geant4 Application Developer Guide, where the details of how it works are illustrated.

You may also want to have a look at the Geant4 Advanced Example gammaray_telescope, that shows how to use the digitisation mechanism in the experimental environment of gamma-ray astronomy. You can find it in geant4/examples/advanced/gammaray_telescope.

Best wishes,

Maria Grazia Pia

Question Sunyaev-Zeldowich effect and GEANT4  Keywords: Cosmology, Astroparticle Physics
by Forough <Forough>,   22 Nov, 2006
   Hi,
   We are going to use GEANT4 in scattering of electrons
   from ionized atoms, so-called Sunyaev-Zeldowich effect
   in cosmology and astroparticle physics.
   Up to the best of our knowledge, this topic has not been
   worked until now. We are pleased to receive your comments
   about the application of GEANT4 in Sunyaev-Zeldowoch effect.
   Could you please help us? 
   Thanks if you reply.
   Sincerely,
   Forough Nasseri
Question hemisphere generation using General Particle Source Module  Keywords: hemisphere in gps
by Manju Sudhakar <manjus@isac.gov.in>,   28 Aug, 2006

Hi,

I would like to know how to generate a hemisphere using General Particle Source Module. It is easy to do so in the G4UserPrimaryGeneratorAction.cc, but I want to use gps because my energy is in the form of a user defined histogram. I checked the queries posted in this forum regarding generation of a hemisphere using gps, and I found out that "biasing" should be used and the extended example "exgps" has a macro "test32.g4mac, which gives the methodology to do so. But, I donot know how the biasing works and the gps manual doesnot give any details. Could someone please help me out in this.. It is really important, as I want to model the space-craft background due to photons generated on the inner surface of a hemisphere that is placed behind the space-craft, ie, the photons are incident from the hemisphere inner surface and impinging on the back of the detector.

Please could someone tell me how to go about this ??

Thanks, Manju Sudhakar.

None Physics Process to include to take care of x-ray polarization  by Manju Sudhakar <manjus@isac.gov.in>,   30 May, 2006

Hi,

I am working on a first step x-ray polarimeter design.I want to measure the polarization of low energy x-rays after scattering from a target. So, inorder to do that, I need to include, in the physics list a process that takes care of polarization. There is a physics process "G4LowEnergyPolarizedComptonScattering", but it is not used in any of the examples ..

And another thing is that in the http://geant4.esa.int/events/g4suw03/ site, there are a number of presentations that mention polarized photoelectric, Rayleigh scattering, pair production etc. And in the photoelectric presentation , there is a page where they show the actual track of the photon, and it is mentioned "G4LowEnergyPolarizedPhotoelectric".

I am using GEANT version 4.7.0.p01.
In the directory /geant4/geant4.7.0.p01/source/processes/electromagnetic/lowenergy/include, 
the only low energy polarization physics process is G4LowEnergyPolarizedComptonScattering
is there. Also, there is no PolarizationManager. In the paper by Ralph Dollan, Karim Laihem 
and Andreas Schalicke "Monte Carlo based studies of a polarized positron 
source for International Linear Collider (ILC)" (arXiv:physics/0512192 v1), they talk about 
polarization framework in GEANT4 .. They talk about "proposed implementation of polarized processes" 
 ... so are these yet to be implemented ?? In the latest version of GEANT, 
are these implemented ??

So, in the version of GEANT that I am using, 4.7.0.p01, the only physics process that is available is "G4LowEnergyPolarizedComptonScattering".

Could you please give me more information about the proposed implementation of polarization ? Is it already implemented ?

Thanks in advance, Manju Sudhakar.

None Re: Physics Process to include to take care of x-ray polarization  by Maria Grazia Pia <Maria Grazia Pia>,   02 Jun, 2006
Re: None Physics Process to include to take care of x-ray polarization (Manju Sudhakar)

Please see the reply to a similar posting in the Electromagnetic Processes section.

Best wishes,

Maria Grazia Pia

None Geant4 applications at IEEE NSS 2006  by Maria Grazia Pia <Maria Grazia Pia>,   29 May, 2006

Geant4 users in space applications may be interested to submit an abstract on their ongoing projects to the IEEE Nuclear Science Symposium (http://www.nss-mic.org/2006). IEEE NSS is a major conference in experimental technologies, including a Software & Computing session. This year it will take place in San Diego, the week before the SPENVIS-Geant4 Space User Workshop also hosted in California in 2006.

The official deadline for abstract submission is over, but direct submission to the conveners is still accepted.

If you are interested, please send your abstract to the Software and Computing convener (Maria.Grazia.Pia@cern.ch, with cc to nss2006@ornl.gov).

Looking forward to many exciting Geant4 results,

Maria Grazia Pia

NSS Software & Computing co-Convener

None regarding polarization in gps  by Manju Sudhakar <manjus@isac.gov.in>,   25 May, 2006

Hi,

This is the first time I will be simulating polarization and I'll be doing it using gps.

The command /gps/polarizarion Px Py Pz where Px, Py and Pz are the doubles that sets the polarization of the source .. Are these the Stokes parameters ?? The parameters for linear polarization, circular polarization and elliptical polarization ..So if my source photons are linearly polarized only, then the values for Px, Py and Pz would be 1,0,0 .. Is my understanding correct or is this something else ...??? These are not vectors ..but doubles .. Could you help me out ??

Regards and thanks, Manju Sudhakar

 

Question How to define the flux of particles?  Keywords: flux
by Lee Xuesheng <chaosl@pku.edu.cn>,   06 Mar, 2006

When using the geant4 in application for Space Radiation,how can I define the flux of particles(e- or gamma)? thanks

None Re: How to define the flux of particles?  by Hugh Evans <Hugh Evans>,   07 Mar, 2006
Re: Question How to define the flux of particles? (Lee Xuesheng)
Hello Lee,

I would recommend the use of the GPS module in G4 for setting up the primary
generator.

To establish a particle source you need to define:
1. the particle source surface - the surface on which the primaries are
generated
2. the angular distribution of the primaries - in space this is usually
omni-directional
3. the energy spectrum

The surface for generating the particle sources is in the most simple, a
sphere that encapsulates the model.

The angular distribution of radiation in space is generally assumed to be
isotropic or omni-directional. There are deviations from this, e.g. at low
altitudes in the South Atlantic Anomaly, during the rising phase of a solar
proton event, or astronomical events such as gamma ray bursts. For radiation
belts, solar proton event and cosmic ray analyses, though, the
omni-directional assumption is usually valid. So, the omni-directional
assumption requires a cosine-law to ensure a correct angular distribution.

If it is radiation belt or solar protons that you are trying to simulate,
then I'd recommend the SPENVIS system (http://www.spenvis.oma.be/) and it's
mulassis model implementation to generate the energy spectrum GPS macro
calls. Mulassis is a 1D general purpose Geant4 application. It will also
provide you with examples on setting up the GPS parameters.

Then there is the issue of tuning all of the above to ensure each event
produces the data you require. This can be achieved through the biasing of
the angles, energies, etc.

Regards,
Hugh

-----
H. Evans
ESA/ESTEC/TEC-EES
Postbus 299
2200 AG Noordwijk
The Netherlands

Phone:+31 (0)71 565-5109
Fax:     +31 (0)71 565-4999


Lee Xuesheng <chaosl@pku.edu.cn> wrote on 07/03/2006 08:54:08:

> *** Discussion title: Space Applications
> Email replies to space_app-g4hn@slac.stanford.edu must include:
>   In-Reply-To: <"/space_app/16"@geant4-hn.slac.stanford.edu>
>   Subject: ...change this to be about your reply.
>
> When using the geant4 in application for Space Radiation,how can I
> define the flux of particles(e- or gamma)? thanks

None Re: How to define the flux of particles?  Keywords: ISS orbit, omni-directional distribution?????
by Andres <Andres.Russu@uv.es>,   25 Mar, 2006
Re: None Re: How to define the flux of particles? (Hugh Evans)

Hello Hugh,

thank you very much for your always very useful comments!!!

As you know, I am simulating an payload (ASIM) wihich hopefully will be at a Columbus external platform one day. This instrument (MXGS) has an aperture (576 cm2) on the bottom in order to detect Transient Gamma Flashes (TGFs) from the atmosphere and also for aurora studies.

The assumption of omni-directional radiation and cosine law in a sphere is perfect for isotropic shielded structures, like ISS, but in my case I don't have the same shielding in all the instrument due to the apperture but the instrument is pointing at nadir so it has the Earth natural shielding.

Therefore, I am looking for another more accurate solution which, as my poor knowledge, will be to have a half sphere shape. What do you think? Do you know any kind of solution?

Thanks again for you and all the people who is helping me.

Regards, Andres

None Re: How to define the flux of particles?  Keywords: ISS orbit, omni-directional distribution?????
by Lee Xuesheng <chaosl@pku.edu.cn>,   05 Oct, 2006
Re: None Re: How to define the flux of particles? (Andres)

Hello Hugh, First,thanks for your answer for my question.

Now, I want to ask the question below: flux = flunce/t flunce is the sum of flux in the time t, that is flunce = flux * t How can I get the time(t) information from the Geant4? Is it the CPU time? Yes, I can get the flux from MUlassis straightly without the time. But I can not get dose rate(dose/t), if I can not get the time.

Thanks very much!

None Re: How to define the flux of particles?  by Hugh Evans <Hugh Evans>,   27 Mar, 2006
Re: None Re: How to define the flux of particles? (Andres)
Hello  Andres (and any other interested parties).

Andres <Andres.Russu@uv.es> wrote on 25/03/2006 20:38:34:

> *** Discussion title: Space Applications
> Email replies to space_app-g4hn@slac.stanford.edu must include:
>   In-Reply-To: <"/space_app/16/1/3"@geant4-hn.slac.stanford.edu>
>   Subject: ...change this to be about your reply.
>
> Hello Hugh,
>
> thank you very much for your always very useful comments!!!
>
> As you know, I am simulating an payload (ASIM) wihich hopefully will be
> at a Columbus external platform one day. This instrument (MXGS) has an
> aperture (576 cm2) on the bottom in order to detect Transient Gamma
> Flashes (TGFs) from the atmosphere and also for aurora studies.
>
> The assumption of omni-directional radiation and cosine law in a sphere
> is perfect for isotropic shielded structures, like ISS, but in my case I
> don't have the same shielding in all the instrument due to the apperture
> but the instrument is pointing at nadir so it has the Earth natural
> shielding.

To ensure there are no ambiguities, I'll break this down on particle sources:

1. For the trapped particle environment, because the trapped particles are
gyrating about the magnetic field lines and close to their mirror point there
is a strong anisotropy in the particle source distribution (see SPENVIS for a
proton model of this anisotropy). In addition, this trapping on the magnetic
field lines makes any Earth Shadow effect irrelevant. For electrons, as they
scatter well, an isotropic distribution is probably a reasonable
approximation.

2. For cosmic rays and solar protons, there is an Earth Shadow effect (which
is included in the CREME software). There is also an anisotropy from these
particles, which is due to the Earth Shadow and also a function of particle
rigidity (see Stormer theories). The Earth Shadow is a simple geometric
factor based on the solid angle subtended by the Earth. The magnetic field
induced anisotropy is a bit more complex to calculate (see ECSS-E-10-04
section 9.3.4; also available through the SPENVIS help pages). I would also
recommend L. Desorgher's MAGNETOCOSMICS package (a Geant4 application) that
can be used to model the trajectories of cosmic rays/solar protons as they
enter the magnetosphere. [if you're adventuresome, you could even try to
incorporate MAGNETOCOSMICS in your simulation - talk to Laurent about that,
though].

>
> Therefore, I am looking for another more accurate solution which, as my
> poor knowledge, will be to have a half sphere shape. What do you think?
> Do you know any kind of solution?

A very simple and basic solution for the trapped particle and solar proton
radiation would be to include an hemispherical isotropic distribution centred
over the detector opening, i.e. ignoring the effects from the back side of
the instrument - if the opening is poorly shielded compared to the backside,
then this is a valid assumption. This would give you better statistics than a
full 4PI simulation (but only by a factor of ~two) (For the cosmic rays, use
a full 4PI distribution, but scale the results by the fraction of the
detector opening that is shadowed by the Earth).

For a more accurate determination: given that there is a detector opening
angle, how well shielded is the detector outside of this acceptance angle? If
it is reasonably well shielded, then a two stage approach is probably
reasonable:

1. full isotropic (worst case scenario) distribution spectrum to confirm that
the "noise" from outside of the detector opening is acceptably negligible (if
it isn't then it gets more complicated).

2. simulate just the particle sources through the detector opening.
Initially, you could apply a simple isotropic source and see how the detector
responds. What is the solid angle opening of the detector? If it is small
enough, then you should be able to get by with ignoring the anisotropies and
just applying a global scale factor to account for them.

If the detector is exposed to space, i.e. the shielding through the detector
opening is very small in comparison to that of the remaining solid angle,
then the assumption that the energetic particle background is only from the
opening angle should be quite reasonable.

If you already have a Geant4 model of the instrument, then you could use the
Solid Sectoring Analysis Tool (SSAT) developed by QinentiQ for ESA to
evaluate the effectiveness of the shielding, i.e. the fraction of shielding
thickness as a function of solid angle, and see through which angle most of
the particles arrive.

If you don't have a model of the instrument and are only trying to get a
first approximation of the noise from radiation, then a ball park figure can
be obtained from:

      dose(*) = SUM( solid_angle(shieldthicks) * dose(shieldThicks))/(4PI)

(*) or whatever effect you've calculated as a function of 1D shielding
thickness.

Finally, despite the large shielding provided on the backside of the detector
by the space station, you can usually assume that any shielding won't really
have an effect on the cosmic ray fluxes (but you should check this anyway -
you can use a 1D simulation).

Unfortunately, putting it all together to get an anisotropic source
definition is not easily achieved. A simple software solution isn't yet
available, sorry. However, parametric runs of the configuration should give
you an estimate of whether the particle source anisotropy or shielding
anisotropy is significant in your results.

Regards
Hugh

None Re: How to define the flux of particles?  Keywords: ISS orbit, omni-directional distribution?????
by Andres <Andres.Russu@uv.es>,   25 Mar, 2006
Re: None Re: How to define the flux of particles? (Hugh Evans)

Hello Hugh,

thank you very much for your always very useful comments!!!

As you know, I am simulating an payload (ASIM) wihich hopefully will be at a Columbus external platform one day. This instrument (MXGS) has an aperture (576 cm2) on the bottom in order to detect Transient Gamma Flashes (TGFs) from the atmosphere and also for aurora studies.

The assumption of omni-directional radiation and cosine law in a sphere is perfect for isotropic shielded structures, like ISS, but in my case I don't have the same shielding in all the instrument due to the apperture but the instrument is pointing at nadir so it has the Earth natural shielding.

Therefore, I am looking for another more accurate solution which, as my poor knowledge, will be to have a half sphere shape. What do you think? Do you know any kind of solution?

Thanks again for you and all the people who is helping me.

Regards, Andres

None Untitled  by Andres <Andres.Russu@uv.es>,   06 Feb, 2006

I have two questions about spacecraft geant simulation.

1.- The number of particles. I am using gps and sphere source with spenvis histogram but I am not sure how many particles per second I have to simulate.

2.- Long term radiation. I would like to simulate long term secondary particles coming from the desactivation of the ISS materials. Do you have any idea about how to do this?

Thanks

None Re: Untitled  by Hugh Evans <Hugh Evans>,   06 Feb, 2006
Re: None Untitled (Andres)
Hello Andres,


Andres <Andres.Russu@uv.es> wrote on 06/02/2006 10:34:35:

> *** Discussion title: Space Applications
> Email replies to space_app-g4hn@slac.stanford.edu must include:
>   In-Reply-To: <"/space_app/13"@geant4-hn.slac.stanford.edu>
>   Subject: ...change this to be about your reply.
>
> I have two questions about spacecraft geant simulation.
>
> 1.- The number of particles. I am using gps and sphere source with
> spenvis histogram but I am not sure how many particles per second I have
> to simulate.

The spenvis macro files are generated for the mission fluence over the period
you specified on the orbit definition page. Thus the results from a mulassis
run on SPENVIS will give the quantities (dose, etc) for the total mission
period.

If it is the space station that you're simulating, and only looking at the
contribution from trapped protons or trapped electrons, then you should be
able to divide the results you get by the mission duration in seconds to get
the results per second as a mission average quantity. Look for the
"/analysis/normalise" macro command in the SPENVIS file, that gives the scale
factor to convert from the simulation results to environmental results.
(NOTE: spenvis/mulassis automatically accounts internally for the surface
area of the particle source sphere. If you're not running mulassis, then
you'll have to further multiply the normalisation factor by the surface area
of your sphere source).

If you're using solar protons, then it's more complicated due to the
stochastic nature of the events. You'll have to pick some events and see what
the activation is on an event by event basis (bearing in mind that you'll
have to include the geomagnetic shielding of the event).

If you mean the number of events that you need to simulate, then keep
increasing the number until your statistics are acceptable. ;-)

>
> 2.- Long term radiation. I would like to simulate long term secondary
> particles coming from the desactivation of the ISS materials. Do you
> have any idea about how to do this?

For the long term radiation, I'd investigate the time constant of the decay
that generates the long term secondary particles. If this is less than an
orbital period, then you'll need to perform your analysis for a per orbit
simulation, e.g. a trajectory crossing the SAA or the polar horns, rather
than a mission average. If, however, the decay constant is greater than a
day, then you should be able to get by using the mission average values. The
models of the trapped radiation belts are long term averages, which
simplifies the analysis - you don't have to worry about short term
(weekly/monthly) variations (if you want to, then you'll have to go to data
sets of in-situ measurements). Then, it's a matter of running your simulation
for a representative day of the mission, and calculating the activitation
from that day. From this you can then determine the long term secondary
radiation.

Regards,
Hugh

-----
H. Evans
ESA/ESTEC/TEC-EES
Postbus 299
2200 AG Noordwijk
The Netherlands

Phone:+31 (0)71 565-5109
Fax:     +31 (0)71 565-4999



None Re: Untitled  by <Andres.Russu@uv.es>,   22 Feb, 2006
Re: None Re: Untitled (Hugh Evans)
Hello Hugh,



> Hello Andres,
> 
> 
> Andres <Andres.Russu@uv.es> wrote on 06/02/2006 10:34:35:
> 
> > *** Discussion title: Space Applications
> > Email replies to space_app-g4hn@slac.stanford.edu must include:
> >   In-Reply-To: <"/space_app/13"@geant4-hn.slac.stanford.edu>
> >   Subject: ...change this to be about your reply.
> >
> > I have two questions about spacecraft geant simulation.
> >
> > 1.- The number of particles. I am using gps and sphere source with
> > spenvis histogram but I am not sure how many particles per second 
I have
> > to simulate.
> 
> The spenvis macro files are generated for the mission fluence over 
the period
> you specified on the orbit definition page. Thus the results from a 
mulassis
> run on SPENVIS will give the quantities (dose, etc) for the total 
mission
> period.
> 
> If it is the space station that you're simulating, and only looking 
at the
> contribution from trapped protons or trapped electrons, then you 
should be
> able to divide the results you get by the mission duration in 
seconds to get
> the results per second as a mission average quantity. Look for the
> "/analysis/normalise" macro command in the SPENVIS file, that gives 
the scale
> factor to convert from the simulation results to environmental 
results.
> (NOTE: spenvis/mulassis automatically accounts internally for the 
surface
> area of the particle source sphere. If you're not running mulassis, 
then
> you'll have to further multiply the normalisation factor by the 
surface area
> of your sphere source).

From the MULASSIS Software Users Manual I have learnt about the 
Normalize factor. Now, I am multiplying each bin of the histogram for 
the normalisation factor.

 
> If you're using solar protons, then it's more complicated due to the
> stochastic nature of the events. You'll have to pick some events and 
see what
> the activation is on an event by event basis (bearing in mind that 
you'll
> have to include the geomagnetic shielding of the event).

If my sphere is 5 m radius, I think I don't have to take into account 
the geomagmnetic shielding effects (they will be despreciable), Do I?
 
> If you mean the number of events that you need to simulate, then keep
> increasing the number until your statistics are acceptable. ;-)
> 

I am increasing the number until the spenvis fluence is reached. This 
point is in 7e3 particles(eg) but if I want to get the corrected 
fluence it is reached in 7e15 particles (eg) as my normalization 
factor is 1e12 (eg). So, Which is the fluence that I must reach to get 
good stadistics? 

> >
> > 2.- Long term radiation. I would like to simulate long term 
secondary
> > particles coming from the desactivation of the ISS materials. Do 
you
> > have any idea about how to do this?
> 
> For the long term radiation, I'd investigate the time constant of 
the decay
> that generates the long term secondary particles. If this is less 
than an
> orbital period, then you'll need to perform your analysis for a per 
orbit
> simulation, e.g. a trajectory crossing the SAA or the polar horns, 
rather
> than a mission average. If, however, the decay constant is greater 
than a
> day, then you should be able to get by using the mission average 
values. The
> models of the trapped radiation belts are long term averages, which
> simplifies the analysis - you don't have to worry about short term
> (weekly/monthly) variations (if you want to, then you'll have to go 
to data
> sets of in-situ measurements). Then, it's a matter of running your 
simulation
> for a representative day of the mission, and calculating the 
activitation
> from that day. From this you can then determine the long term 
secondary
> radiation.
> 
> Regards,
> Hugh
> 
> -----
> H. Evans
> ESA/ESTEC/TEC-EES
> Postbus 299
> 2200 AG Noordwijk
> The Netherlands
> 
> Phone:+31 (0)71 565-5109
> Fax:     +31 (0)71 565-4999
> 

The second point will come later.

Thank you in advance,
Andres

*****************************************
Andres Russu
Andres.Russu@uv.es
Astronomy and Space Science Group (GACE)
Ph: 0034 96 354 36 79
Fax: 0034 96 354 36 77
*****************************************




None Spacecraft activation from bkg environment  by <Andres.Russu@uv.es>,   31 Jan, 2006

Dear friends,

I am simulating a gamma instrument for the ISS. I would like to simulate how the spacecraft is getting radioactive in a long term simulation. I mean I want to take into account the particles coming form the decays of the materials around the detector plane.

Thanks in advance, Andres

Question doubt concerning the radioactive decay  Keywords: energy deposit by radioactive decay as a function of time
by Manju Sudhakar <manjus@isac.gov.in>,   02 Jan, 2006
Hi, 

   I would like to simulate the spallation reactions that occur CsI detectors 
due to cosmic ray protons.First I want to do an experiment on the energy deposited 
in CsI due to irradiation by a beam of monoenergetic protons, as a function of time.
I want to get the energy deposit spectrum in CsI after irradiation, as a function of time.

    I have looked at the examples/extended/radioactivity/exrdm, where the concept of 
theStep->GetPreStepPoint()->GetGlobalTime() is used.

Please, could someone help me in understanding this concept and how it is used in 
radioactive decay ????
How can I use this to get the energy deposit as a function of time since irradiation ???
Could someone please answer my query ...

With thanks, 
Manju Sudhakar.
Question doubts concerning grdm  Keywords: doubts about grdm commands
by Manju Sudhakar <manjus@isac.gov.in>,   29 Dec, 2005
Hi, 
  I would like to simulate the decay of induced radio-isotopes when protons 
are incident on a space-borne gamma-ray detector, cosmic-ray protons.I want 
to model the detector background.To simulate radioactive decay, we need to 
use grdm.There are some aspects about it I don't understand.

1) the command /grdm/sourceTimeProfile <filename>
    
    as given in the manual, it is for defining the "source time profile"
I donot understand ... what does this mean ?? Why is it used ?? Is it 
mandatory ?? In the extended example "radioactivity/exrdm" in geant 4.7.0, 
in one of the macros "proton-b.mac" /grdm/sourceTimeProfile beam.data is a 
command.When I opened beam.data, the file looked like this 
0  1e6
60  0

Donot understand head or tail of this ... please help me get clarified on 
this point.


2) the command /grdm/decayBiasProfile <filename>

       What about this command ?In the extended example "radioactivity/exrdm" 
in geant 4.7.0, in the macro "proton-b.mac", there is a command 
/grdm/decayBiasProfile measures.data.And, when I opened it
1000     1
1100     0
3000     1
3100     0
1e5      1
1.1e5    0

    No clue to what this file means.

Could someone please explain all this to me clearly and in detail ?

With lots and lots of advance thanks 
Manju Sudhakar.
None Re: Doubt regarding General Particle Source Module  by Hugh Evans <Hugh Evans>,   30 Aug, 2005
Hello Manju,

The energy units in the GPS histogram that you're building are in MeV. The
absolute values of the differential "flux" are not important as they are all
normalised to a probability distribution [0..1], therefore the relative
difference between bins in the histogram are of the greatest importance, not
the absolute value.

While cosmic ray fluxes aren't yet available, the trapped proton, trapped
electron and solar proton flux/fluence spectra from SPENVIS
(http://www.spenvis.oma.be/spenvis/) can be converted to GPS macros via its
Mulassis application.  By playing with this, you can easily determine how the
environmental spectra are converted to the GPS macro commands.

Alternatively, I can also send you an Excel spreadsheet within which the
CREME-86 spectra are implemented and the GPS macros are formed.

Regards,
Hugh

-----
H. Evans
ESA/ESTEC/TEC-EES
Postbus 299
2200 AG Noordwijk
The Netherlands

Phone:+31 (0)71 565-5109
Fax:     +31 (0)71 565-4999



|---------+--------------------------->
|         |           Manju Sudhakar  |
|         |           <manjus@isac.ern|
|         |           et.in>          |
|         |                           |
|         |           30/08/2005 09:25|
|         |                           |
|---------+--------------------------->
  >--------------------------------------------------------------------------------------------------------------------------|
  |                                                                                                                          |
  |        To:      space_app-g4hn@slac.stanford.edu                                                                         |
  |        cc:                                                                                                               |
  |        Subject: Doubt regarding General Particle Source Module                                                           |
  >--------------------------------------------------------------------------------------------------------------------------|




*** Discussion title: Space Applications
Email replies to space_app-g4hn@slac.stanford.edu must include:
  In-Reply-To: <"/space_app/3"@geant4-hn.slac.stanford.edu>
  Subject: ...change this to be about your reply.

Dear GEANT users ,

I have a doubt regarding one of the commands in General Particle Source
Module.

I have constructed a detector system mounted on a space platform, and I
wish to irradiate it with cosmic-ray protons, at 1AU.I have the spectrum
of the protons with me and I have to enter the energy and flux as a
user-defined arbitrary point-wise function of the form, /gps/ene/type
Arb /gps/hist/type arb /gps/hist/point ...(energy) ... (flux) .... ....
.... .... /gps/hist/inter Lin /run/beamOn 100000

        This is fine, but my doubts are as follows ,
1) What is the default value of the energy points ? Is it MeV ?
2) What is the default value of the differential flux to be entered as in the

   second column ?

Could you please answer my doubts. Regards, Manju Sudhakar.





None Re: Doubt regarding General Particle Source Module  by Andres <Andres.Russu@uv.es>,   28 Sep, 2005
Re: None Re: Doubt regarding General Particle Source Module (Hugh Evans)

Dear Hugh,

I am very interested inthe Excel Spreadsheet to convert Creme86 to GPS macros, In fact, I am using Creme96 Do you have also the spreadsheet for this version?

Best regards, Andres

Question Doubt regarding General Particle Source Module  Keywords: units of differential spectrum in gps.
by Manju Sudhakar <manjus@isac.ernet.in>,   30 Aug, 2005

Dear GEANT users ,

I have a doubt regarding one of the commands in General Particle Source Module.

I have constructed a detector system mounted on a space platform, and I wish to irradiate it with cosmic-ray protons, at 1AU.I have the spectrum of the protons with me and I have to enter the energy and flux as a user-defined arbitrary point-wise function of the form, /gps/ene/type Arb /gps/hist/type arb /gps/hist/point ...(energy) ... (flux) .... .... .... .... /gps/hist/inter Lin /run/beamOn 100000

        This is fine, but my doubts are as follows ,
1) What is the default value of the energy points ? Is it MeV ?
2) What is the default value of the differential flux to be entered as in the 
   second column ? 

Could you please answer my doubts. Regards, Manju Sudhakar.

None Geant4 Space Users Workshop  by Makoto Asai <Makoto Asai>,   02 Jun, 2005
Dear colleague,

On 3-7 October 2005, a combined workshop for the SPENVIS and 
GEANT4 Space applications communities will be held in the 
historic university town of Leuven, Belgium. The workshop venue 
is the Faculty Club of the Catholic University, situated in the 
beautifully restored Grand Beguinage, a UNESCO World Heritage 
Site.

The SPENVIS workshop will take place on 3-4 Oct, the Geant4 
workshop will follow on 6-7 Oct, with a joint session in the 
morning of 5 Oct.

The SPENVIS workshop is the second in a series that started at 
ESTEC in 2002. The emphasis during the SPENVIS workshop will 
be on user experience and requirements for future developments. 
We aim for an informal workshop with a limited number of 
presentations (and a poster session) which can serve as starting 
points for round table discussions addressing specific
developments and applications. The programme is deliberately 
kept open to allow users to actively participate in the 
organisation. If you have inputs or suggestions for the workshop 
programme, please send a note to D.Heynderickx@oma.be .

Several GEANT4 based radiation effects packages are now available 
or under development, which are or will be implemented in the 
SPENVIS user interface. In view of these developments, we opted 
for a combined workshop where users and developers of both 
communities can meet and interact.

The GEANT4 Space Users' workshop follows the events organised 
at ESTEC in 2003 and Vanderbilt University in 2004. The focus 
will be on new results for component, sensor, and shielding 
analyses, as well as on developments for Exploration and Science 
missions. Users are invited to highlight new requirements based 
on their applications.

Registration is now open at:
 http://www.spenvis.oma.be/spenvis/workshop/workshop.html 
Workshop participants are invited to submit abstracts before 
1st September, via e-mail attachment in pdf format to 
D.Heynderickx@oma.be . Please indicate for which part of the 
workshop you are submitting, and your preference for oral or 
poster presentation (subject to the final decision of the 
scientific committee). Abstracts and a more detailed programme
will be put on-line when available.

More information on SPENVIS and GEANT4 space applications can 
be found at
 http://www.spenvis.oma.be/ 
and
 http://geant4.esa.int/ 
respectively.


 Add Message Add Message
to: "Space Applications"

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 ]