Presentation is loading. Please wait.

Presentation is loading. Please wait.

1.Check Laser track of B=0 run and exclude some tracks in order to get precise GGV eff, which in turn is used when extract T1, T2 value by pos-B and neg-B.

Similar presentations


Presentation on theme: "1.Check Laser track of B=0 run and exclude some tracks in order to get precise GGV eff, which in turn is used when extract T1, T2 value by pos-B and neg-B."— Presentation transcript:

1 1.Check Laser track of B=0 run and exclude some tracks in order to get precise GGV eff, which in turn is used when extract T1, T2 value by pos-B and neg-B runs 2. Check Laser track of B=0 run to find mis-alignment, Gating Grid Voltage error distortion and other E field distortions. 3. Reconstruct vertex using A/C side tracks separately to measure vertex shift due to ExBTwist

2 1.Most Beams agree with Model Check the difference between ( Model dY vs lX ) and ( Real dY vs lX ) beam by beam  B=0 laser run  dY = lY (GGV=70V as reference) - lY(GGV=40V)  to calculate model dY, C 0 =1, C 1 =0, GGV eff = 0.9175*GGV set +3.91 (from Jim) Good beam Some outliers zigzag <0.01~0.02 cm

3 1. But few beams disagree with Model Few beams do not agree with model calculation. There seems to be 2 typical bad beam: systematic raise/fall at Inner/Outer Bad ?

4 1. Bad Beam List Side-Bundle-Rod-Beam, Inner/Outer, raise/fall My conclusion so far: something bad at A side rod4-beam2,4,5,6 C side rod4-beam4 C side rod5-beam2 Same conclusion if use GGV eff = 0.89627*GGV set -4.60153 (from stefan’s 2 nd order cal) I use laser track cuts: tan  <0.9&&some quality cuts (bundle 1,2,3) x (rod 0,1,2,3,4,5) x (beam 0,1,2,4,5,6) bundle1bundle2bundle3 A142,O,fA242,O,fabsent A144,O,rA244,O,rabsent A145,O,rabsentA345,O,r A146,O,rA246,O,rA346,O,r bundle1bundle2bundle3 C152,IabsentC352,I absentC244,IC344,I

5 2. Some alignment works well Check the residual between ideal lY and measured lY beam by beam  B=0 laser run, GGV=70V  plot dY(=ideal lY – measured lY) vs lX  Unfortunately, ideal lY is NOT ideal because of mirror angle tilt, but this plot must be linear if there is no distortion and mis-alignment Inter-Sector  Almost linear suggests small distortion  Alignment works well gap ~0.1 cm is corrected

6 2. Not yet align?? Some inter-sector & inner-outer mis-alignment still exist?? Rough criteria is gap above 0.1 cm

7 2. Bad Alignment?? According to 2 plots, alignment procedure make the situation worse…? 2 plots are for same chamber, only the laser beam height is different

8 2. List of “not yet align beam” &”bad align beam” “not yet align beam” list ASide-Rod1-beam1 ASide-Rod1-beam2 ASide-Rod2-beam1 (ASide-Rod0-beam6) (ASide-Rod3-beam6) CSide-Rod2-beam1 (CSide-Rod2-beam5) “bad align beam” list: ASide-Rod4-Beam1 ASide-Rod4-Beam6 (Aside-Rod5-beam5) CSide-Rod3-Beam0 Cside-Rod3-Beam1 Cside-Rod3-beam2

9 2. Bending at lX =80 ~120 cm 4 plots below are for same beam number, so the geometry is same. But inner sector of right figures are bended bending good bad

10 2. List of bended track I use laser track (bundle 2,3) x (rod 0,1,2,3,4,5) x (beam 0,1,2,4,5,6) There are 2 main reasons of track bending  GGV error distortion and E field distortion If the bending is happened both for bundle 2 and 3, then the reason may be GGV error distortion as previous graph ASide-Rod0-Beam4 (ASide-Rod0-Beam5) (ASide-Rod1-Beam2) ASide-Rod1-Beam5 ASide-Rod2-Beam1 (ASide-Rod3-Beam2,4,5) Aside-Rod5-Beam1 CSide-Rod0-Beam4 CSide-Rod0-Beam5 (CSide-Rod3-Beam1) Cside-Rod3-Beam4,5 CSide-Rod4-Beam1 CSide-Rod4-Beam4 Cside-Rod5-Beam1

11 Correlation btw chamber tilt and laser track bending

12 So far we can not find the sign of Gating Grid Voltage error with real data. Any E field distortion around readout chamber cause the drift length independent track distortion Misalignment cause 2 effects : coordinate mismatch and extra E field distortion. Chamber tilt is one of the source for E field distortion around the readout chamber. The question is Can we extract and distinguish GGV error distortion and chamber tilt distortion using B=0 Laser track?

13 Some laser track bending can be explained by chamber tilt, but the bending is very large..

14 0 1 2 3 4 5 A side Bending Laser track and Chamber tilt 04 05 15 12 21 32 34 35 55 51

15 Sector 54 rod0 Sector 55 1 2 3 4 5 04 05 31 34 35 41 44 51 C side Bending Laser track and Chamber tilt

16 Conclusion so far We can not clearly find chamber tilt effect to laser track. The laser track bending may be due to rod shift.

17 Vertex shift in real data

18 Vertex Reconstruction Code Vertex reconstruction by A/C side tracks separately Here is a part of the code in AliAnalysisTask:Exec() (please let me know if something wrong about the usage of the AliVertexerTracks class, because I don’t yet understand in detail ) TObjArray Atrk, Ctrk; UShort_t Aid[ntracks], Cid[ntracks]; // Track loop for (Int_t iTracks = 0; iTracks < ntracks; iTracks++) { AliESDtrack* track = fESD->GetTrack(iTracks); if (!track) { AliExternalTrackParam *t =(AliExternalTrackParam *)track->GetTPCInnerParam(); if(t){ Bool_t ok=kFALSE; ok = AliTracker:: PropagateTrackToBxByBz(t,2.8,track->GetMass(),1,kFALSE); if(ok){ Double_t eta = track->Eta(); Int_t id = track->GetID();

19 if(eta>0){ Int_t nn=Atrk.GetEntriesFast(); Atrk.AddLast(t); Aid[nn]=id; } if(eta<0){ Int_t nn=Ctrk.GetEntriesFast(); Ctrk.AddLast(t); Cid[nn]=id; } } //track loop AliVertexerTracks *vt = new AliVertexerTracks(5); vt->SetTPCMode(); vt->SetConstraintOff(); AliESDVertex * Apv = vt->FindPrimaryVertex(&Atrk, Aid); AliESDVertex * Cpv = vt->FindPrimaryVertex(&Ctrk, Cid); fhisAV->Fill(Apv->GetX(), Apv->GetY()); fhisCV->Fill(Cpv->GetX(), Cpv->GetY());

20 3. LHC10b pass1 pp real data Green :ESDPrimaryVertex This large shift(~3mm) may be due to the imperfection of pass1 ExBfirst. Vertex x-y position black-A side, red-C side Left: run116574 (pos-B) Center: run115884 (pos-B) Right: run115325 (neg-B)

21 Pass1 and pass2 We can get pass2 real data recently. Pass1  ExBfirst only Pass 2  ExBBShape (update version of ExBfirst) ExBTwist (Twist parameter is extracted by Global fitting by Marian)

22 Vertex Shift Simulation

23 Include distortion to simulation code AliTPC.h member  AliTPCExBBShape fexbs; AliTPCExBTwist fexbt; AliTPC.cxx constructor  fexbs = new AliTPCExBBShape(); fexbt = new AliTPCExBTwist(); AliTPC.cxx MakeSector function  at first fexbs -> SetOmegaTau(0.32, 1.0, 1.0); fexbs->SetBfield(field); fexbt -> SetOmegaTau(0.32, 1.0, 1.0); fexbt -> SetXTwist(0.001);  at primary electron loop fexbs-> GetDistortion(x, roc, dx1); fexbt-> GetDistortion(x, roc, dx2); x = x+dx1+dx2; Something strange about produced simulation data is that PrimaryVertex in AliESD.root is always (0,0) !! But, anyway, analyze this data

24 Simulation result of Vertex Shift statistics is small now, so the vertex shift due to ExBBBShape (expected O(~0.1 mm) ) can not be recognized. I produced more data now.

25 Large Vertex Shift by Twist ?? Maximum distortion is O(0.1mm), but Vertex Shift is O(1mm) !! ?? This is because The vertex is reconstructed by all tracks included large eta tracks

26 what to do next I will also produce simulation data include AliTPCBoundaryVoltError


Download ppt "1.Check Laser track of B=0 run and exclude some tracks in order to get precise GGV eff, which in turn is used when extract T1, T2 value by pos-B and neg-B."

Similar presentations


Ads by Google