1. 3/2/2000 Hobbs, Silicon Mtng 1 STT & SMT Assembly Precision Method Results –no in situ alignment –w/in situ alignment Conclusions Stress difference between assembly precision and knowledge of assembly

2. 3/2/2000 Hobbs, Silicon Mtng 2 Simulation Method Tests –Generate tracks with “true” (non ideal) geometry –Reconstruct with different(ideal) geometry Simple simulation –Correct SMT geometry (RECO) –Intersect helices and ladders STT will use alignment – , every cluster (z L =0) – r, avgs. in matrices (look up) Caveat: STT group not approved this yet...

3. 3/2/2000 Hobbs, Silicon Mtng 3 Code Sanity Checks Look at b(reco)-b(true) vs. ??? Move detector in X by 30  m Tilt detector about X by 100  rad

4. 3/2/2000 Hobbs, Silicon Mtng 4 Result I: No in situ Alignment True, non-ideal for generation Assume ideal geometry for reco This relies completely on assembly precision…* * This also limits uncertainty on alignment knowledge for positions 300  rad 250  m100  m 300  rad

5. 3/2/2000 Hobbs, Silicon Mtng 5 Result II: Perfect  True (non-ideal) for generation For reconstruction – true  from true geometry & – matrices(radii) from ideal no problems if  r < 200  m and tilt < 250  rad This relies on assembly knowledge

6. 3/2/2000 Hobbs, Silicon Mtng 6  r c = 100  m  r c = 50  m One line/layer smaller radii, larger effect  Precision (  ) Do we really know  well enough using alignment? –Alignment goal: Doesn’t effect hit resolution by more than 10%... –Look at (limited) effects   = * * (  x 2 +  r c 2 ) +  c 2 xrcxrc ( ) 2 1 (1+2 )*(1 + ) 2 ( ) xrcxrc 2 xrcxrc 2 1 r c 2 X = SMT position (local) r c = Radius at ladder center  c =  at ladder center Layer 1 Layer 8 http://sbhepnt.physics.sunysb.edu/ ~hobbs/alignment/align.pdf

7. 3/2/2000 Hobbs, Silicon Mtng 7 Conclusions Assembly requirements interplay with alignment status Constraints –Knowledge, offsets  beam 25  m –Knowledge, radial offsets: 70  m –Assembly tol., angles 200  rad Not met, require in situ alignment. –Time dependence? –Systematics in alignment (CDF)? Absolute tolerances only on angles