Message: QGSP_BIC vs. QGSP_BERT -- which is "right" for protons in space? Not Logged In (login)
 Next-in-Thread Next-in-Thread
 Next-in-Forum Next-in-Forum

Question QGSP_BIC vs. QGSP_BERT -- which is "right" for protons in space? 

Keywords: physics lists, QGSP_BIC, QGSP_BERT, space
Forum: Hadronic Processes
Date: 16 Apr, 2008
From: Mark Looper <Mark Looper>

Hello--

I have been working on a project to model the production of secondary particles by the impact of cosmic ray and Van-Allen-Belt protons on Earth's upper atmosphere. Most recently I have been attempting to obtain cross sections for a colleague's modeling by using Geant4 to pump billions of protons into a small "box" of gas (hydrogen, helium, nitrogen, or oxygen) and tabulating all particles that come out. I had been using a physics list lifted from the radioactivedecay extended example, since it has a nice SelectPhysics command that lets me choose the specific physics list interactively. I was going to send a note to the HyperNews forum asking about some weird stuff I saw in the elastic scattering contribution under Geant4 9.0 (I showed some plots, with errors I've now fixed, to some of you at the Tokyo University meeting), but the updated examples with release 9.1 (or 9.1.p01, which I'm now running) fixed this, and I am now concerned about results for inelastic interactions. Specifically, I would like to present some plots of results from the QGSP_BERT and QGSP_BIC physics lists, and ask which list is more appropriate to use for space-physics energy ranges (here, 4 MeV to 4 GeV proton lab energies) based on the differences one sees. I don't know if the HyperNews system can handle embedded graphics, so I include links to plots hosted on one of my own websites, which you can download (they're 100 kB to 1 MB each), and to which I'll refer.

The target in my simulations is a box of gas (at STP) ten times as long as it is wide, with protons incident on the center of one face along the long axis (and dimensions chosen to minimize the energy loss of secondary particles in the target before they escape and are tabulated). The plots linked below are crossplots, for a given product species and a given case (incident proton energy, target species, and physics list), of the direction cosine vs. the kinetic energy of the product particles, in the CM frame. The direction cosine is +1 for particles coming out in the direction of the incident beam (I have attempted to exclude non-interacting primary beam particles from the proton plots), and -1 for particles moving opposite the beam direction (in the CM frame, that is; most of these particles are actually moving forward in the lab frame). The signature of elastic scattering is an unchanged energy in the CM frame, i.e., a vertical band of points in a plot like these at an energy corresponding to the initial CM energy of the projectile or of the target nucleus.

For protons incident on hydrogen, I see nothing but elastically scattered particles up to 400 MeV projectile energy (in the lab frame); inelastic scattering really turns on by 500 MeV for the QGSP_BERT physics list, and by 600 MeV for QGSP_BIC. As seen in these plots at 600 MeV lab energy for QGSP_BIC and QGSP_BERT, in both cases the inelastically-scattered protons have CM energies only up to about 66 MeV below the energy of the incident beam, and there is much more structure in the distribution for QGSP_BIC than for QGSP_BERT, with the latter being nearly isotropic in direction and flat in energy spectrum. The 66 MeV gap persists to the highest energies I simulated, and more structure does develop in the QGSP_BERT results, but as seen in these plots at the highest energy simulated (4 GeV in the lab frame) for QGSP_BIC and QGSP_BERT, the distribution remains much smoother for the latter.

For protons incident on helium, inelastically scattered particles are apparent (initially in small numbers) for all energies in the QGSP_BERT simulations, and start at 15 MeV for QGSP_BIC. By 30 MeV, something weird develops in the secondary protons under QGSP_BERT, sort of an "absorption line" about 16 MeV below the primary beam energy in the CM. That is visible in these plots at 40 MeV beam energy for QGSP_BIC and QGSP_BERT, as is the apparent lack of protons below about 0.5 MeV CM energy for QGSP_BIC and, again, a more uniform distribution of inelastically-scattered protons for QGSP_BERT. (Ignore the diagonal "tracks" of particles above the beam energy; these are protons that undergo extreme electromagnetic scattering but no nuclear interactions, so as to retain nearly full energy in the lab frame while being scattered well away from the beam direction. They are present in simulations for all targets at relatively low beam energies, and are an artifact of the finite size of the target.) As is visible in these plots at 100 MeV beam energy for QGSP_BIC and QGSP_BERT, an "edge" develops in the QGSP_BIC results at the same energy as the "line" in the QGSP_BERT results. At slightly higher energies, 150 MeV and as seen here at 200 MeV for QGSP_BIC and QGSP_BERT, some really weird structure that I don't understand at all develops at the low-energy end of the QGSP_BERT results; it becomes much less prominent by 300 MeV, though as seen here in plots at 4 GeV for QGSP_BIC and QGSP_BERT, some residue persists to the highest energies simulated.

Turning to the alphas coming out of the simulation volume for protons incident on a helium target, in these plots at 30 MeV beam energy for QGSP_BIC and QGSP_BERT, we note that again the distribution of inelastically scattered particles looks more uniform for the latter than for the former, but also that there is a cutoff in QGSP_BERT for products more than about 4.5 MeV below the CM energy of the target nuclei. Toward higher energies, as in these plots at 80 MeV primary energy for QGSP_BIC and QGSP_BERT, we see this lower-energy cutoff persisting. Around 200 MeV beam energy, a second bunch of particles appears about 30 MeV below the CM energy of the target nuclei, and the results from the two physics lists look more similar toward higher energies as here at 1 GeV for QGSP_BIC and QGSP_BERT, though the cutoff in the higher-energy bunch persists for QGSP_BERT.

The results for the nitrogen and oxygen targets display trends similar to those for the helium target, though I saw no weirdness comparable to that at 150 and 200 MeV in the plots for protons produced from protons on helium, as above. Inelastic scattering turns on by 6 MeV primary energy, with a distribution of product protons that is more uniform at low primary energies for QGSP_BERT. The additional structure that develops for QGSP_BERT at high energies, as seen in these plots for 1.5 GeV protons on nitrogen for QGSP_BIC and QGSP_BERT, takes the form of a diagonal band for the latter that's barely if at all visible for the former. In addition, the cutoff in inelastically scattered (but not disrupted) target nuclei that was apparent in the helium simulations above is present for nitrogen and oxygen targets for both physics lists, as seen here in plots for 50 MeV protons on nitrogen for QGSP_BIC and QGSP_BERT.

Well, that's a heck of a lot of plots; I hope these are of some use to the hadronic physics developers and community in evaluating the differences between these two physics lists for inelastic interactions. One further difference that we noted was that QGSP_BERT produced about 30% more neutrons at high primary energy, for all targets, than did QGSP_BIC. We also wondered about two-body interactions, for example p+4He->2H+3He, which should be constrained by kinematics to form "bands" in plots like these, but which we did not see. Are cross sections for such processes included in these physics lists? And, of course, the perennial question: for purposes of simulation of protons in energy ranges around a few MeV to a few GeV, as will be encountered in space, what physics list should I use?

Many thanks!

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

1 None: Re: QGSP_BIC vs. QGSP_BERT -- which is   (Vladimir Ivanchenko - 23 Jun, 2008)
 Add Message Add Message
to: "QGSP_BIC vs. QGSP_BERT -- which is "right" for protons in space?"

 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 ]