Date post: | 14-Dec-2015 |
Category: |
Documents |
Upload: | lawson-cullin |
View: | 216 times |
Download: | 0 times |
First conclusionsThe discrepancy data/MC has a chamber “pattern”.
It is smaller for chambers placed upstream and larger for chamber installed downstream
Giacomo suggested calorimeter back splash as explanation for the dependent Z discrepancy
Before asking calo people to look at it I tried to have more clues
M1 Z
Y
X
ZA sketch of how chambersare installedOne row containschamber at Z1/Z2 or Z3/Z4A column mixes chamberupstream and downstream of thesupport wall
Z1
Z2
Z3
Z4
Hits multiplicity vs Z
The Z dependence of hit multiplicity decreased from R1 to R4R1 ~ 1.6 R2 ~1.5 R3 ~1.25 R4 ~ 1.1Is this pattern in agreement with back splash explanation?
Data
Hits multiplicity vs Z
MC
Although similar z dependence is observed the entity is much reducedR1 ~1.2 R2 ~1.1 R3 ~1.07 R4 ~1.03
Using TDC to understand the Z pattern in M1
● We do not have many handles in muon detector data. Try to look ad TDC.
● If z pattern is due to back splash naively we expect that the arrival time of those hits is smaller in Z4 chambers and larger in Z1 chambers. If back splashed particles are at small angles than Z4-Z1 ~ 35 cm so the delay in Z1 wrt Z4 is 70cm/c ~ 2ns ~ 1.5 TDC bin.
● Due to back splash component we expect that average time of hits is larger in Z1 than Z4
Time hit vs z in M1R1
All 12 chambers
TDC average vs Z
The 4 chambers in the equatorial plane
TDC distribution in 4 chambers
The chamber at Z2 has strange TDCSpectrum, maybe at differentworking condition
Conclusions part II
● The TDC distributions seems to point in the direction of back splash
● Unfortunately not proved yet it is the main effect of the discrepancy
Go back to MonteCarlo
● Assume current MC knows back splash but underestimates the rate
● Remove Calorimeter from simulation and compare with standard one. If Z dependent rate is due to (mainly) back splash it should disappear in MC w/o calorimeter
In MC there is a source of Z dependent rate apart from Calorimeter back splash
Still using MonteCarlo
● If calorimeter back splash is not the only source of Z dependence which other is present?
● Since Zdependence is higher in R1 an smaller in R4 , could it be related to the pipe?
● Remove Pipe from simulationStandardNo CaloNo Calo and no Pipe
The pipe is not responsible forresidual Z dependence
Std vs no Calo simulation
● For all regions removing calorimeter reduces a lot the effect but not cancel
What else????
● The other possibility is the material of M1 (chamber,wall, etc). Chambers downstream sees more material
Average material upstream on M1 is 40% X0
M1 material
● M1R1 chambers material non negligible, large variation in x-y plane
● For non perpendicular tracks the effective material is spreadout in x-y
● For z1 ~1.5% for z4~10%
● The chamber frame are more important in R1 that R4 since frame area wrt panel area is larger
Summary
● We see z dependent rate
● The amount is region dependent
● Some hint from TDC that back splash is a component of the z dependent rate
● From MC we see the back splash is present but also some other source of z dependent rate is in: not Pipe
● The last hipotesis is chamber material itself
M2 plug description
● Thanks to Robert we have now a realistic description of M2 plugs
● Test the effect using low threshold simulation
M2 plug descriptionTest the effect using low threshold simulation
● Compare Robert precise description with a simplified description I wrote: only X and Y outer box dimension
● The z opening implemented by Robert seems to not increase the rate. Why? Not intuitive to me
High thr. Low thr. low+my plug description
low+robert description
M2R1 11.5 16.9 17.9 17.3
M2R2 10.2 16.6 16.8 16.3
M2R3 5.2 10.6 10.6 10.3
M3R4 5.8 9.2 9.1 9.1
Material due to M1 itselself
Before first gap of chamberat Z1 ~1.5% x0 Before first gap of chamber
at Z4 ~9.0% x0
Hits multiplicity at different Z
The Z dependence of hit multiplicity decreased from R1 to R4R1 ~ 1.6 R2 ~1.5 R3 ~1.25 R4 ~ 1.1Is this pattern in agreement with bakspalsh explanation?I naively imagine that the back splash z dependence is region independentIf so backsplash seems to not be the dominant effect
Data