Published on January 9, 2008
NuMI Muon Monitor Absorber Studies : NuMI Muon Monitor Absorber Studies Jesse Chvojka University of Rochester October 17th, 2007 Refresher about NuMI and the Muon Monitors: Refresher about NuMI and the Muon Monitors NuMI - supplies beam to MINOS & MINERvA Protons from the main injector→graphite target to give pions and kaons pions and kaons->muons + muon neutrinos (π or K→μ+μ) One to one mapping between μ and μ Muon monitors Detect almost only muons # of μ’s indicates # of μ‘s We’re worried about delta rays (electrons scattered off rock) as a background Muon Monitor Actual person Hard hat Delta Rays, eh?: Delta Rays, eh? Mostly muons make it through the rock and beam absorber, but… eletrons can get scattered off from either the absorber or rock, or even the muon monitors! rock rock Muon Delta ray Hadron Absorber Muon Monitor We want to study effect of absorber materials on delta ray production: We want to study effect of absorber materials on delta ray production To do this, we need to: Sum up total energy for an event Figure out how much energy is from the muon and how much is from a delta ray Sort “Delta” events by where they start Internal Deltas = Deltas that originate in the “Chamber” (consider part of the muon event) External Deltas = Deltas that originate upstream of the “Chamber” (target) Problems I was having: Problems I was having I was making two BIG mistakes 1) DefaultCutValue (DCV) set way too high (it was set at DCV = 2.1 cm) This value sets the cut off for delta ray production, now use 0.001 cm 2) Binning of histograms was too large distorting mean and RMS on plots Big problem since I was reading mean and RMS from the plots Now use very small binning and calculate mean separately to verify I’m doing it correctly Also fixed some smaller issues (like how I was finding errors) Here You See the Problem: Here You See the Problem Getting inconsistent # of internal and external deltas (for DCV = 2.1 cm) Number of internal deltas should not vary Mean energy for a distribution varies erratically This number should not vary Is the problem fixed? Results from Paraffin Study, DCV = 0.001 cm: Is the problem fixed? Results from Paraffin Study, DCV = 0.001 cm A lot to digest, so please don’t strain yourself Short answer: Sort of Three Checks Before Victory: Three Checks Before Victory Look at quantities that should NOT vary for internal delta rays 1) Total number 2) The mean energy 3) The total energy for all events combined Do any of these vary in a statistically significant way when I vary the thickness of the absorber? Does the total # of δ-rays’s vary?: Does the total # of δ-rays’s vary? For this to be a flat distribution, Chi-Squared 6.7/5 dof Prob = 24% For this to be a flat distribution, Chi-Squared 28.7/5 dof Prob = 0.0027% Internal Deltas External Deltas Used √n as error Does the mean energy vary?: Does the mean energy vary? For this to be a flat distribution, Chi-Squared 4.4/5 dof Prob = 49% For this to be a flat distribution, Chi-Squared 0.6/5 dof Prob = 99% Internal Deltas Muons Does the total energy vary?: Does the total energy vary? For this to be a flat distribution, Chi-Squared 24.3/5 dof Prob = 0.02% Doubling the size of the errors gives 30% prob. For this to be a flat distribution, Chi-Squared 27.7/5 dof Prob = 0.004% Internal Deltas All Deltas OK, so these plots don’t look that nice Victory! Sort of….: Victory! Sort of…. Not perfect, but do these distributions seem acceptable? This is a huge step since chi-squared probabilities were around 10^-10 to 10^-100 before I fixed my mistakes Decided to pretend it was OK and see what a LEAD absorber looks like Pb run, DCV = 0.001 cm: Pb run, DCV = 0.001 cm Pb data seems to make more sense A note on units, I state everything in terms of cm of paraffin. I keep g/cm^2 constant 20 cm paraffin = 1.64 cm Pb, 10 cm paraffin = 0.82 cm Pb, etc. Pb absorber, # of deltas: Pb absorber, # of deltas For this to be a flat distribution, Chi-Squared 3.4/5 dof Prob = 63% For this to be a flat distribution, Chi-Squared 162.6/5 dof Prob = 10^-31% Internal Deltas External Deltas Used √n as error Pb Absorber, mean energy: Pb Absorber, mean energy For this to be a flat distribution, Chi-Squared 2.1/5 dof Prob = 83% For this to be a flat distribution, Chi-Squared 1.8/5 dof Prob = 87% Internal Deltas External Deltas Pb Absorber total energy: Pb Absorber total energy For this to be a flat distribution, Chi-Squared 14.7/5 dof Prob = 1.2% For this to be a flat distribution, Chi-Squared 11.8/5 dof Prob = 3.7% Internal Deltas External Deltas Conclusions: Conclusions Problems mostly fixed Total energy still looks funny Maybe go to a smaller DCV Paraffin seems to increase # of external deltas Pb seems to decrease # of external deltas Backup slides: Backup slides What is DefaultCutValue?Geant4 documentation Says:: What is DefaultCutValue? Geant4 documentation Says: 2.4.2. Range Cuts To avoid infrared divergence, some electromagnetic processes require a threshold below which no secondary will be generated. Because of this requirement, gammas, electrons and positrons require production thresholds which the user should define. This threshold should be defined as a distance, or range cut-off, which is internally converted to an energy for individual materials. The range threshold should be defined in the initialization phase using the SetCuts() method of G4VUserPhysicsList. Section 5.5 discusses threshold and tracking cuts in detail. http://geant4.web.cern.ch/geant4/UserDocumentation/UsersGuides/ForApplicationDeveloper/html/ch02s04.html So how do I sort events?: So how do I sort events? Consider an electron a delta ray if: Delta ray starts before 100 cm (start of chamber) Delta ray created in the chamber doesn’t have the muon as a mother. Trace back seven generations in case of deltas creating deltas in the chamber Else, count the electron as part of the muon event This seems right to me, but gives the wrong answer Delta Rays From Muons: Delta Rays From Muons Dolomite KEY RED = negative BLUE = positive GREEN = neutral Scintillator air 10 GeV mu+ Delta Ray Outdated picture. Currently, have a target, an absorber, and a complex chamber based on real muon monitor geometry.