April 22, 2005 Meeting

Attending: Josh Klein, Tim Bolton, Matt Worcester

Agenda

Progress reports

"1. G4 is up and running. Everything looks good and I think we will may it easily, but see below. We can put it into your CVS repository whenever you wish.

"2. The way we have planned things is not going to work out exactly as planned, I think it will work out better than planned. Here's why: our idea was that I would give you a flux function containing neutrons at the vessel and veto information which you would then read in to ReactorFSim. The problem is that we expect one unvetoed neutron per day. G4 generates events at about 2 Hz and we need to sample over 20 m^2. At 450 MWE, the rate over 20 m^2 is 3.2 Hz, so to get one neutron to put into ReactorG4Sim, I need to run G4 for a day and a half.

"In addition, you do not yet have a way to propagating a 100 MeV neutron and taking into account n-p scattering at these energies. Maybe by now you do, which is fine. In any event, what we can do, which is much easier, is this: attached is a file of neutrons generated at (0,0,-3.5m) with cos_thete=(0,1). The neutron energy is flat over 0.1-1000 MeV and the polar angle distubution is flat over (0,1). I thien let G4 propogate them through the mineral oil and flag the proton scatterings. This way, you only need worry about the what happens with the proton (Birke's law) and neutron after it hits the proton. You run all these events through ReactorFSim, but be sure to keep track of the initial neutron position and energy: if you know that, then I can use G4 to give you weights for each and we're done.

"There are some points: the way G4 moves a neutron through any sort of scattering is to destroy the initial state neutron and create a final state neutron. I just write everything into the event record:

"The first 9001 energy is the initial neutron
9004 is the scattered proton
The 9001 record just before a scattered proton gives you the neutron
The 9001 record just before a scattered proton gives you the neutron
momentum vector before the scattering; you work out the neutron momentum vector after.

"This should be easy. I've attached a file of 500 events; you tell me and I can give you as many as you want."

Dave and Tim are about 90% done with the improved reconstruction based on their PMT data structure.

Matt has checked in code to do multiple compton scattering for the H and Gd capture gammas to ReactorFsim. He also defined two separate AV thicknesses for the 3-zone configuration.

New work

• The low tail in phposrec is gone. That tail was from a small amount of energy left in the scint from a compton scattered gamma from a positron produced way outside of the scint (I have rmaxgen = 350). Can this all be from attenuation in the oil? Or is there some new check in the code for particles not contained in the scint?
• phposrec has two (out of 1000) badly reconstructed energies (E > 2000 MeV). The nu vertex is out in the oil, and they have phposvis = 4 MeV which seems okay. Do you understand these events?
• Because I have rmasgen = 350 and no compton scattering on the capture gammas, I expect a large number of events with Eneu = 0, which I see in the old phneurec distribution. The new phneurec does not have these events. This also makes me think something is checking for particles near the scint.
• Finally the H capture peak is much taller, which the n capture peak is relatively the same heighth.

Dave and Tim are going to add timing for scintillation light to ReactorFsim. They might be done in about 3 weeks.

Matt is going to check in new code with some bug fixes and pass the two AV thicknesses to ReactorDetector. He is also going to start the code in ReactorFsim to handle Peter's new veto events, defined above. He is going to start looking at Cherenkov light in the mineral oil.

Discussion

Next Meeting(s)