U.GaspariniCMS week, 11/06/021 MB3 internal alignment study using cosmic muons - 14 MB3 chambers assembled in Legnaro have been extensively tested using.

U.Gasparini CMS week, 11/06/02 1

MB3 internal alignment study

using cosmic muons

- 14 MB3 chambers assembled in Legnaro have been extensively tested using cosmic run data (some preliminary results on internal layer and SL alignment on first 3 chambers already presented at CMS week in June 2002)

- afterHV tests, pressure tests, noise/rate test, dead-channel test on each single SL (performed with scalers before chamber assembling ), assembled chambers are put on a cosmic station test ( 1/2 chamber per time is tested) and a full DAQ => local rec. chain ; ORCA, running on Obj.database; all data stored/archived there)[ Noise,rate, tmax results and wire position measurements in SL stored in ‘database’ on Web (M.Passaseo, P. Checchia, E.Torassa,A.Meneguzzo,F.Cavallo):http: //www.pd.infn.it/dbcms CMS note in preparation ]

A.Meneguzzo, M.Zanetti, U.G.

(special thanks to M.De Giorgi, F.Gonella, S.Vanini, P.Ronchese)

U.Gasparini CMS week, 11/06/02 2

Cosmic run data

Full MB3 chamber in Legnaro:

cosmic angle (rad)

x (cm)

½ chamber instrumented in this run

U.Gasparini CMS week, 11/06/02 3

Testing-SL alignment

23214 m

-0.730.13 mrad

MB3_08

x2

x1

All 4 point tracks, | < 20o

build 4 point segmentin each single SL

Extrapolate to themiddle plane of the chamber

U.Gasparini CMS week, 11/06/02 4

3 point fit residuals

.

.

.

<>=86 m

326 m

345 m

225 m

326 m

115 m

105 m

276 m

SL 2SL 1

MB3_08

U.Gasparini CMS week, 11/06/02 5

3 point fit residuals &intercept and slope difference

112 -layers (8x 14)

14 chambers

Slope difference (mrad)

intercept difference (m)

Hit residuals(m)

Ch. 3, 8, 9, 18(residual averages in all 8 layers < 75 m )

U.Gasparini CMS week, 11/06/02 6

-SL (apparent) misalignments

syst.err yextr

U.Gasparini CMS week, 11/06/02 7

Tmax in left – right half cells“Good” SL(MB3_08, SL2)

“Bad” SL(MB3_05 SL2)

left semicell

right semicell T=2.9 0.2 ns

T=-2.3 0.2 nsT=0.4 0.2 ns

T=0.0 0.2 ns

U.Gasparini CMS week, 11/06/02 8

3 point fit residuals

10 of the 14 chambers analyzed have at least 1 “good” SL:

best SL

worst SL

40 layers (=4x10)

U.Gasparini CMS week, 11/06/02 9

Average residuals w.r.t. extrapolation from other SL

“best”

“worst”

- use values marked by arrows as alignment correction of corresponding layer

- look to 3 point fit residuals in corrected SL; if all average 3 point fit residuals are < 50 m , stop here; otherwise, repeat procedure a 2nd time

U.Gasparini CMS week, 11/06/02 10

3 point fit residuals in worst SL

Before correction

After correction

(max 2 layers in a SL moved)

U.Gasparini CMS week, 11/06/02 11

-SL misalignments

4 chambers (nr. 4, 7, 11, 14) have 2 “bad” SL’s : they need a more refined layer alignment procedure

In each SL, consider the 4 (not independent) Tmax combinations:

T = TmaxL-Tmax

R

1=vdriftT234, 2=vdriftT134, ....

and the quantities:

The system:i=aijj with:

may be solved considering the misalignments

j w.r.t. one layer, under the assumption that only 2 of the other 3 have

displacement significantly different from zero w.r.the reference one

uknown misalignments

U.Gasparini CMS week, 11/06/02 12

-SL misalignments

6 possible solutions for each SL 6x6= 36 possible layer configurations in the chambers

Only fews give slope differences compatible with zero

Take the configuration with minimum (=slope difference betweenSegments in the 2 SL) as the best one and use thecorresponding x (intercept difference) as best estimate of the SL misalignment;take the maximum semi-dispersion of all possible x as estimationof systematic error on the SL misalignment

U.Gasparini CMS week, 11/06/02 13

-SL misalignments

before corrections after corrections

syst.err yextr syst.err (2 max/3.9)yextrChambers with2 “bad” SL

max.of averages of3 pont fit residuals

U.Gasparini CMS week, 11/06/02 14

single layer alignment corrections

U.Gasparini CMS week, 11/06/02 15

Conclusions

- A procedure to measure internal misalignments in the SL of the chambers using cosmic rays has been proposed and applied to 14 MB3 chambers assembled in LNL

- the SL relative position, SL , can be measured within a systematic error of less than 300 m

-12 chambers show SL values well below the 1 mm construction tolerance; 2 chambers are at the limit of this value

- 6 out of 112 layers have an alignment correction bigger than 100 m