EC To Do

From solidwiki
Revision as of 19:03, 3 December 2020 by Saw (Talk | contribs) (Created 06/14/12(Xiaochao), Last modified 06/29/12(Xiaochao))

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

SoLID EC To Do List

  1. What is causing pion rejection to level out at p>4GeV? (JP)
    • If it's hard to explain we should change the cuts and make the plots more understandable.
    • Ed is asking what algorithm we used for PID. We know it's based on two simple cuts, but can it be better? Jin's answer: we can use neural network or looking for correlation between PS and SH (difference expected between EM & hadronic shower), but we may not want to use this at the design stage.
  2. For radiaton dose plot, make sure people understand the points are "from photon entering the EC", not "photon" (since photons themselves don't do much damage). (Everyone)
  3. Also, for radiation plot, I copied from earlier slides a yellow box/arrow "Low E photon dominant". Why is low E photon dominant while the curves clearly show electrons are higher than photons?
    • Answer from Zhiwen: THe electron include both DIS electron and background low E electron. We should separate the two and see what we get.
    • Note from Jin: Zhiwen should include more realistic target material (glass, collimator, etc), and use refined physics list;
  4. If we can only reach 50:1 pion rejection we should work with other talks and make the physics requirement on PID more consistent. (Everyone)
    • Zein-Eddine's Cherenkov talk indicates 100:1. Eric claims overall we can get <1% contamination using the latest EC rejection from Jin and pi/e from Seamus ()
  5. Winstone cones don't help much (Ed). They work for parallel light and fibers have ~26 degrees outgoing light angles;
    • Zhiwen: addtional note, single clad fiber have about 26 degree light angle, double clad has more like 30 - 35 degree. So how good winestone cone is a question. I don't think we overlook it, it's just we need more information or even test about it. Hope Paul can help us and get some feedback from FNAL.
    • Jin: We should not give up Winstone cone. Losing some light might be just okay. Will simulate where is best place to put pads.
    • Xiaochao: we shouldn't rely on Fermilab for making the cones. Zhiwen will check with Simona what the cherenkov use. Update 6/29/12: Cherenkov cones are large, not sure if the company can make the small cones we need.
  6. Ed: Bending preshower fiber and let them go back for readout may suffer from too much light loss. another way to readout preshower could be use some thin scintillator pad painted with WLS material and has clear fiber embedded in it. He said Russia had use something like that before, I will ask him more info.
    • This design has the advantage of running the clear PS fibers directly backward. The clear fibers could be designed in a way to be replace-able. The cost is also lower.
    • Zhiwen mentioned we can run PS WLS fibers through hollow steel supporting rods (4 fibers per rod). But Xiaochao doesn't think this is practical, also hard to fix if a fiber is broken.
  7. Ed: Clear fibers are difficult and will be a lot of work. Can we run silicon-based sensors such as APDs just next to the module? With neutron shielding?
    • Jin: APD is silicon-based sensors. Ed mentioned they make 1-cm square APDs which will match our 100 fibers/module well. The 2nd kind of silicon-based is silicon PM but won't work for us. APD cost is higher: ~$1000 compared to PMT EA ~$400.
    • On top of that, can we use the same for forward angle ECs?
      • Zhiwen: think if no cost issue, we should use same readout for both to simply the situation. but have to think about if we want the advantage of having readout stick outside of yoke and easy access for service.
  8. Price quote must be exact: Must have quote fro IHEP for the current (segmented) design, and the latest quote on the fibers. Same for clear fiber cost. Do not say "using $1.5 instead of $1.15 for safety". Contingencies should be estimated and added separated.
    • Answer from Nilanga: "I believe the normal DOE guideline is"
      1. For off the shelf, calalogue items - 10%
      2. For things thing that need to be custom made (like GEM foil) but you have quotes etc, - 20%
      3. For things that are just at concept state, need to be designed etc, - 30%
      4. Nilanga used 30% flat on everything related to GEM.
    • Mehdi will followup on quoting the fiber cost; - update: email sent on 6/27/12 requesting a quote.
    • Zhiwen will send 2-3 options of design to IHEP and get quote for the modules.
      • Update 6/29/12: made some slides. Xiaochao commented that option 1 (routing the PS fibers in the front) might kill the fibers first. Mehdi added the bending radius condition for fibers which pose a problem for both options 1 and 2. Jin said for option #3 we can consider running the WLS fibers through the whole module, and attach the pads to the max shower position near 5 X0, instead of starting from the 1st layer. Pion rejection will need to be studied for this new design scheme.
  9. Can we make the modules in China? (Ed) - No, China is more interested in funding their own projects. They funded GEM because it's a new technology that they can learn from CERN (Haiyan).
  10. My own question (or D. Mack's): When we say the light loss of the Kuraray fiber is 13% at 0.1Mrad, is that 13% per meter or independent of the fiber length?
    • Zhiwen will check the paper.
  11. Following Lorenzo's talk:
    • We should compare our rad dose with his to see if they are the same for electron, photon, pions;
      • Jin: should compare front-face rays first.
      • Update 6/29/12: working on it!
    • We should simulate Pb shielding (Ed: 1-1.5X0 will be needed, 2mm or 1/2 X0 is too thin) and its effect on PID.