BTeV-doc 1599
BTeV:
What’s New and What’s Next
Rob Kutschke
Fermilab
Illinois Institute of Technology CAPP Seminar, February 27, 2003
What is BTeV?
• At the Tevatron p-pbar collider, at Fermilab:
– Forward spectrometer.
– Beauty and charm physics:
• Precision measurements.
• Exhaustive search for new physics.
• BTeV is a part of broad program to address
fundamental questions in flavor physics.
• http://www-btev.fnal.gov
– Click on “BTeV for Physicists”.
– Checkout the “Document Database”.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
2
Fundamental Questions in Flavor Physics
• Why families? Why 3?
• Quark mixing angles: Are they explained by
Standard Model (SM)? Arise from new physics?
• Mass heirarchy: Why? Related to mixing angles?
• Is CPT violated? If so, what physics is behind it?
• Matter/anti-matter asymmetry of the universe:
What interactions were involved?
• Quarks vs leptons: What are the similarities and
differences in mass heirarchies and mixing angles?
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
3
Fundamental Questions in Flavor Physics …
• These are interesting, compelling, questions which
we must answer.
• The program to answer these questions involves
present and future experiments in K, D, B, and
neutrino physics and in astrophysics.
• BTeV is a crucial part of this program.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
4
Physics Goals
• Measure: CP violation in B(uds) , Bs mixing, rare b decay
rates, CP violation and rare decays in the charm sector.
• Look for rare/forbidden decays discover new physics.
• Make an exhaustive search for physics beyond SM and
to precisely measure SM parameters.
• Test for inconsistencies in the Standard Model: If found
go beyond the SM and elucidate the new physics.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
5
A Brief History of BTeV
• January 1999: R&D program approved by lab.
• June 2000: Stage I approval from lab.
– Two arm spectrometer.
• Fall 2001: funding situation deteriorated.
– Lab asked for a proposal for a descoped detector.
– IR to reuse components from completed CDF/D0.
• May 2002: Stage I approval for descoped detector.
–
–
–
–
Instrument only one arm, at least initially.
PAC recommended lab explore other IR solutions.
Offline computed to be supplied via universities (grid).
Cost reduced about $70M to about $110M.
• In FY2002 dollars, including contingency of 33.5%.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
6
A Brief History of BTeV
• October 2002: Temple Review
– Fermilab internal review in the “Lehman” style.
– Addressed: cost, schedule, management and
technical risk.
– BTeV passed with flying colors.
• Their cost estimate about 5-10% above ours.
– ( Their report also adds overheads ).
• No new significant risks identified.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
7
Nominal Tevatron Parameters
Parameter
Center of Mass Energy
Value
2 TeV
Peak Instantaneous Luminosity
Yearly Integrated Luminosity
Bunch spacing
21032 cm-2 s-1
2 fb-1/year
132 ns ‡
Luminous region (sx,sy,sz)
Interactions/Beam Crossing
(0.003, 0.003, 30.) cm
2
sbb_ /stot
1/1000
scc_ /stot
1/100
‡
Will address 396 ns later in this talk.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
8
Why a Hadron Machine?
• Large samples of b quarks are available:
– ~4x1011 b hadrons per 107 sec at nominal Tevatron
parameters.
• e+e- machines at the Y(4S) at L = 1034 cm-2 s-1
– Only ~2x108 B’s per 107 s.
• Bs , Bc & b-baryons are produced at hadron
machines but not at Y(4S) e+e- machines.
• Lots of charm also produced.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
9
Why a Forward Detector?
• The higher momenta b’s
are produced at large
pseudo-rapidity ().
bg
• Higher momentum:
– Longer decay lengths.
– Less multiple scattering.
• b and c hadrons have
lifetimes of  0.1 to
1.5 fs ( ct of 30 to 450 mm).
Beauty02, February 27, 2003

• Background interactions
have decay lengths of zero.
Rob Kutschke, Fermilab
10
Why a Forward Detector?
• b’s are produced in b
b-bar pairs.
• The b and b-bar are
closely correlated in .
• Important for flavor
tagging …
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
11
Flavor Tagging
• Bd and Bs undergo flavor oscillations:
– A particle born as a Bd can decay as an Bd–bar, + vice versa.
– To observe mixing we need to determine the flavor at birth
( b or b-bar )
• Determined by looking at other properties of the event, often the
properties of the recoil b(bar).
– Also need flavor tagging to observe CP violation in final
states which are CP eigenstates.
• e  efficiency
• D  Dilution  (Nright-Nwrong)/(Nright+Nwrong)
• Effective tagging efficiency  eD2
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
12
The Trigger Concept
• Many forward produced b and c hadrons have lab
frame decay lengths in the range of a few hundred
mm to a few cm.
• Most of the background processes have decay
lengths of zero or >> a few cm.
• Trigger on detached vertices at the lowest level.
– Previous experiments have used detachment triggers
but not at the lowest level.
– Trigger is wide open, just like ARGUS, CLEO, BaBar,
Belle: will be efficient for many decay modes whose
signficance is not yet appreciated!
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
13
p
Beauty02, February 27, 2003
p
Rob Kutschke, Fermilab
14
The “One Arm” Configuration
• The full vertex detector:
– Covers the full length of IR.
– Level 1 trigger: finds tracks in both z hemispheres to
ensure the best primary vertex.
• 1 each: fwd tracking, RICH, EMCal, m systems.
• The steel for the muon toroid on the
un-instrumented side:
– Shielding; support the compensating dipole; keep floor
loading constant in case we instrument the other arm.
• We retain the full trigger and DA bandwidth.
– Original estimate was conservative.
– See discussion of offline computing …
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
15
Toroids
• Production: bb pairs.
• Both b hadrons
go into one arm.
p
p
• Retain 100% of
trigger/DAQ bandwidth
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
16
Key Design Features of BTeV
• Magnet on the IR
• Strong Particle ID
– Allows momentum
measurement in the trigger.
• Precision vertex detector
– Planar pixel arrays.
• Vertex trigger at Level 1.
– Can trigger on final states
with only hadrons.
• PbWO4 Calorimeter
– g and p0 reconstruction.
– Ring Imaging Cerenkov
(RICH) detector.
– Hadron and lepton ID!
– Background rejection.
– Flavor tagging.
• Excellent muon ID system
– Redundant triggering of
final states with muons.
• Fast, high capacity DAQ.
– Can record a significant
fraction of all B decays.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
17
Pixel Vertex Detector
• Superior signal to noise.
• Excellent spatial resolution.
• 5 to 10 microns depending
on angle of incidence.
• Very low occupancy.
• Very fast.
• Radiation hard.
• 3 bit FADC on every channel
• Used directly in Level 1 Trigger.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
Elevation View
10 of 31 Doublet Stations
18
Pixel Substrates (Mechanical and Thermal)
• Existing baseline: fuzzy carbon with embedded glassy
Carbon tubes carrying water/glycol as coolant.
– Sole source proprietary process.
– Pin hole and mechanical stability issues.
• Backups: Be, pocofoam, CFRP, and TPG
– Be – CTE mismatch with Si, material budget, joint reliability
– Pocofoam – joint reliability, tube insertion, still needs a lot of R&D
effort
• All but TPG have have coolant joint reliability concerns.
• Thermal Pyrolytic Graphite: TPG “cold-finger”
– No cooling joints. Needs design and test to prove the concept.
– By far the simplest solution but a small hit to the materials budget.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
19
Material Properties of TPG
• Thermal Pyrolytic Graphite (TPG)
– http://www.advceramics.com/acc/products/tc1050/
– Excellent in-plane thermal conductivity (1700W/mK at RT), increasing to >
4000 at cryogenic temp.
– Density 2.26 g/cm2
– CTE: -1.0ppm/C (in-plane) and 25ppm/C (out of plane)
– Tensile strength: 1000 Ksi (in-plane) and <1 Ksi (out of plane)
– X0: 198.5 mm
– Electrically conducting
– Used in HERA-B Silicon, will be used by ATLAS, LHCb
• Pyrolytic Graphite Sheet (PGS)
–
–
–
–
http://maco.panasonic.co.jp/ad/fd/a1e.html
Thermal conductivity 600-800 W/mK
Flexible and withstand repeated bending
Light weight : specific density of 1
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
20
Schematic of a Straw Station
Schematic of Straw Station
300 mradian tracking region
• 7 Stations
– 3 views per station.
– 3 planes per view.
• Forward tracking system,
straws+silicon:
– Key to excellent momentum
resolution.
– Precision measurements of
impact points on RICH and
ECAL.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
21
Glass beads
Frames
(10cm width)
Straw Station #1
Layout for U&V View
Silicon Strip detectors
cover the deadened
area.
Deadened Area (27cm)
Deadened Area (27cm)
(Station #2 has similar
geometry)
Straw Length =54 cm
384 straws/view
1152 straws/station
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
22
Baseline Silicon Tracker Design
• 7 stations
• Coverage from beam pipe
to 13.5cm from the beam
• Each station has 3 planes
of 300 mm thick SMD
with 100 mm pitch
• Each detector is 7x7 cm2
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
23
Particle ID
• Two critical roles:
– To reduce combinatoric background signal channels.
• Eg distinguish B0 p+p- from the more copious B0 Kp.
– Flavor tagging.
• Charged particle ID ( e,m,p,K,p):
– RICH, EMCal, and Muon systems.
• EMCal used to identify g’s, p0’s and electrons.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
24
Ring Imaging Cherenkov Counter (RICH)
• Two radiators: to cover desired momentum range.
– Original design: gas C4F10, aerogel.
– One array of Hybrid PhotoDiodes (HPDs) for both.
• Aerogel has proven inadequate:
– Large, diffuse rings with too few photons, lost in gas rings.
– Thicker aerogel limited by scattering in bubbles.
• New solution:
– Liquid C5F12: very large angle rings.
– Rings do not overlap with gas rings. Separate readouts.
– Readout by PMTs on the sides of the gas box.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
25
Before and After: K/p Separation
Aerogel
C5F12
c2(K) - c2(p)
c2(K) - c2(p)
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
26
New Layout of the BTeV RICH
• PMTs do not need
quartz windows.
Borosilicate glass OK.
Tracks from IR
•PMT costs partly
offset by reduced
size HPD arrays.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
27
HPD Photos
Photocathode
Readout
Pins for
Multianode
Pixel readout
• 61 channel HPD shown (we have in and will use 163 channel HPD).
• Pins through the glass envelope connect pixels to the electronics.
• Backup solution: Hamamatsu R7600 M16 Multianode PMT.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
28
Cherenkov Angles in the BTeV RICH
• If only the gas radiator,
then for p < 9 GeV/c:
• No K/p separation
– Hurts kaon tagging.
• K/p sep. is threshold only
– Every badly reconstructed
track is a kaon!
– Hurts kaon tagging.
Gas
• For 9 < p < 18 GeV/c, K/p
separation is threshold
only.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
29
Geant 3 Simulation of RICH Response
K/p Separation in Gas
K/p Separation in Liquid
MisID (B0 Kp)
p < 9 GeV/c
Efficiency(B p+p-)
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
30
Using the RICH for Lepton ID
• Many leptons which
miss the Ecal and
Muon systems are
accepted by the
RICH.
• These leptons are at
low momentum
where RICH has
lepton ID power.
• Perfect match!
Leptons from B decay
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
31
RICH has e-p and m-p Discrimination!
C4F10
• Wide angle particles are mostly low p. Perfect Match!
• Efficiency: ×2.4 (×3.9) for single (double) lepton ID.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
32
Electromagnetic Calorimeter
• Projective geometry crystal calorimeter.
• PbWO4: for s(E) and for radiation hardness (like CMS).
Block from Shanghai Institute of Ceramics
5X5 stack of blocks from Bogoroditsk.
• Also evaluating crystals from Beijing and Apatity(Russia).
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
33
Electromagnetic Calorimeter
• BTeV can use PMTs for readout.
– CMS uses avalanche photodiodes(barrel) and triodes (endcap)
because their readout is inside B field.
• Mock up the mechanical design exists:
– To help estimate cost and schedule.
– Basic design: Al lattice
– Cooling design also underway.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
34
EMCal IHEP Test Beam Program
• IHEP Protvino: Dec/00, Mar-Apr/01, Nov-Dec/01, MarApr/02, Nov-Dec/02.
– Electron and pion beams (momentum measured).
•
Study s(E) vs incident energy, position and angle:
– Time, temperature and rate stability.
– Radiation hardness and recovery.
• Realistic dose rates and artificially high dose rates.
•
Used 10 stage tubes in tests, will use 6 stage in BTeV.
– 6 stage + QIE chip almost identical to KTeV.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
35
s(E) Measured in IHEP Test Beam
• Stochastic term 1.8% is good enough.
– Original estimate of 1.6% was for different size crystals.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
36
Geant 3 Simulations of By, By/
gg
s = 5 MeV/c2
M(gg)
Beauty02, February 27, 2003
p+p-g
s= 6 MeV/c2
M(p+p-g)
Rob Kutschke, Fermilab
(p+p-)gg
s = 5 MeV/c2
M(p+p-gg)
(GeV/c2)
37
G3 Simulation of B0K*0g, K*0 K+p*
CLEO
barrel
e=89%
Radius (cm)
• + minbias events: Possion distributed mean = 2/beam crossing.
• Plots are for events in which the charged tracks pass all cuts.
• CLEO/BaBar/Belle-like performance in a hadronic environment.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
38
The BTeV Muon System
• Two main functions:
– Muon ID both for signal selection and for tagging.
– Triggering:
• Redundant trigger to test the detached vertex trigger.
• More efficient for some states of particular interest such as
BJ/y , BJ/y /, B K*m+m- …
• Toroidal B field for momentum measurement in
trigger.
• 3/8 inch dia. stainless steel proportional tubes.
• 36864 channels per arm.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
39
Muon Vocabulary: Review
Three (3) stations of detectors
Toroids
IP
Compensating Dipole
• Each station is made up of
overlapping octants.
• 4 views/octant: r,u,v, r.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
40
…Muon Vocabulary: Review
• Two octants form a quad
• Quads will be built at institutions and delivered to FNAL

Each quad will
individually
supported: it
will be possible
to install/remove
a quad during
run.

Scale mock up
built at Univ. of
Illinois.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
41
Level 1 Trigger
• Uses only information from the pixel detector.
• Inspects every beam crossing.
– Heavily pipelined.
• Finds tracks, and looks for a primary vertex with
evidence for a nearby secondary vertex.
• Momentum measurement in trigger allows
identification of low momentum tracks which can
scatter a lot and fake tracks from a secondary
vertex.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
42
Level 1 Trigger Performance
• Nominal requirement: at least 2 tracks detached from
primary vertex by more than 4s.
• Rejects 99% of background beam crossings
• Achieves the following efficiencies:
State
B  p+ pBs  DsK
B-  DoKB-  Ksp-
efficiency(%)
55
70
60
40
state
efficiency(%)
Bo  K + p54
Bo  J/y Ks
50
Bs  J/yK*
69
Bo  K*g
40
• Efficiency is quoted relative to a events which pass all
analysis cuts.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
43
Simulation of Level 1 Di-Muon Trigger
Minimum bias rejection
Beauty02, February 27, 2003
• Geant 3 simulation with 2
interactions per crossing.
• Signal events: J/y m+mfrom generic B decays.
• Find tracks standalone
within each octant.
• Require two oppositely
charged tracks in different
octants + indicated cuts.
• J/y efficiency denominator
has good reconstructable
tracks, with p>5 GeV.
Rob Kutschke, Fermilab
44
Simulation of DiMuon Trigger …
• In order to meet the DAQ bandwidth restrictions,
we need a rejection of minimum bias events of
around 1 to a few hundred.
• There is a wide range of possible operating points.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
45
L2/L3 Trigger Plans
Trigger Level
Level 2
refined tracking,
vertex cut
Level 3
uses full event data
Trigger Parameters
Input Rate
Output Rate
Reduction
Factor
Processing
Time
75KHz
100KB/event
8 GB/s
7.5KHz
100KB/event
800 MB/s
10
13ms
7.5KHz
100KB/event
800 MB/s
4KHz
50KB/s
200MB/s
2
130ms
• L2 and L3 triggers are implemented in the same hardware, a PC Farm
• L2 Uses tracking information to look for detached vertices and detached tracks
• Prototype code is working.
• L3 does the full reconstruction and writes DSTs (similar to traditional offline)
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
46
Flavor Tagging
• Methods in order of decreasing dilution:
–
–
–
–
Away side Kaon (Reasonably high efficiency).
Away side muon (Low efficiency).
Same side tag K (for Bs) and p ( for B0).
Jet charge (Large overlap with ASK and ASM).
• Still to come: away side electron.
• To remove overlaps, use the simplest method:
– Poll methods in order of decreasing dilution.
– Stop when a method gives an answer.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
47
Summary of Flavor Tagging
Tag Type
Bs
Bd
eD2
Independent
eD2
Correlated
eD2
Independent
eD2
Correlated
Away Side K
5.8%
5.8
6.0%
6.0%
Away Side m
1.3%
1.3
1.2%
0.8%
Same Side K(p)
5.7%
4.5
4.8%
1.4%
Jet Charge
5.4%
0.4
1.8%
1.0%
Total
12.1
9.2%
Nominal
13%
10%
Extra 1% • Include electrons
• Allow for optimized use of all info.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
48
CP Violation, CKM Physics and
all That
• Physics reach estimates are quoted for one Snowmass
year (107 s) of running at these parameters.
– Exception: a using B0 rp  p+p-p0 is 1.4 years.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
49
Formulation of CKM Matrix
d
s

1 2
u  1- l
l
2


1
c
V= 
-l
1 - l 2 - i A2 l 4
2

t  Al 3 (1 - r - i )
- Al 2


b

 1 2 
Al  r - i 1 - l   
 2 


2
2
Al (1 + il )


1



3
• Good l3 in real part & l5 in imaginary part.
• We know l=0.22, A~0.8; constraints on r & .
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
50
The 6 CKM Unitarity Triangles
 VtbVtd* 

b = arg * 
 Vcb Vcd 
 Vcs* Vcb 

c = arg - *
 Vts Vtb 
a
=
p-(b+g)
*
 VudVus 
c  = arg - * 
 VcdVcs 
 Vub* Vud 
g = arg - * 
 VcbVcd 
a, b & g probably large, c~0.03, c smaller
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
51
Measuring b
• sin(2b) has already been shown to be non-zero.
– First hints CDF. Best measurements BaBar/Belle.
• We presume that sin(2b) will be well measured
before BTeV starts running.
• BTeV should still be able to improve the
measurement.
• Sensitivity sin(2b) using BoJ/y Ks only in one
year of running : ± 0.017.
– Determined using Geant 3 based simulation.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
52
Removing Ambiguities from b
Beauty02, February 27, 2003
For sin(2b)=0.7
Decay Rate
• Bo  yKo, Kop+l-n
• Exploits K0S/K0L interference.
– Equal amplitudes to p+l-n.
• Low yield:
– 1/100 of Ksp+p-.
– 1700/year (untagged)
• Can determine sign of b with
O(100) low background,
tagged events.
• Sensitivity improves for
smalle sin(2b).
Rob Kutschke, Fermilab
tK integrated over tB
53
Measuring a
• B0 p+p-:
– Sensitivity to ACP in
one year: ±0.030
– But penguin pollution!
– Need p-p0 and p0p0 to
unpollute. Tough to do!
• B0 rp  p+p-p0
– Dalitz plot analysis
(Snyder and Quinn ).
• (next page)
Beauty02, February 27, 2003
• Sensitive to both
sin(2a) and cos(2a).
Rob Kutschke, Fermilab
54
Mini MC Study of B0 rp  p+p-p0
• Dalitz plot density:
– Synder Quinn matrix element.
– Incoherent BG: S:BG = 4:1.
• Non-resonant (flat).
• Resonant in r.
• Acceptance and smearing
parameterized from Geant
based study.
• Results for trials of 1000
signal events + BG.
• Sensitivity (1.4 years) < 4o.
Beauty02, February 27, 2003
a (gen)
Rres
Rnon
a (recon)
da
77.3o
0.2
0.2
77.2o
1.6o
77.3o
0.4
0
77.1o
1.8o
93.0o
0.2
0.2
93.3o
1.9o
93.0o
0.4
0
93.3o
2.1o
111.0o
0.2
0.2
111.7o
3.9o
111.0o
0.4
0.2
110.4o
4.3o
Rob Kutschke, Fermilab
55
Four Ways of Measuring g
Model
independent
• Time dependent flavor tagged analysis of BsDsK-.
• Sensitivity in one year: ±8o
• Rate difference between B-DoK- & B+DoK+.
• Sensitivity in one year: ±13o
• Rate measurements in Kop and Kp (Fleisher-Mannel) or
rates in Kop & asymmetry in Kpo (Neubert et al) .
• Sensitivity in one year: ±4o + Theory uncertainties.
• Use U spin symmetry ds: measure time dependent
asymmetries in both Bop+p-& BsK+K- (Fleischer)
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
56
Measuring c
• BsJ/y  and BsJ/y .
– y  l+l– We now use both electrons and muons.
• This measurement is possible because of the
excellent photon and p0 detection provided by the
PbWO4 calorimeter.
• Excellent S/B: 15:1 for  and 30:1 for /.
• Sensitivity for one years running: ±0.024.
• Will take several years to resolve expected: c0.03.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
57
• If xs is less than
about 70, BTeV
will be able to
measure it in about
1 year.
• If it is less than
about 80, BTeV
will be able to
measure it in about
3.2 years.
Beauty02, February 27, 2003
Time Required for 5s significance (years)
xs Reach
6
5
4
3
2
1
Rob Kutschke, Fermilab
58
Physics Beyond the Standard Model
• New Physics (NP) effects can be subtle:
– More than just: a + b + g  180o.
• Suppose there is NP in B0 mixing:
– If we measure b and a via mixing mediated modes,
J/yK0S and p+p-, we may measure:
• 2b = 2b + q
• 2a = 2a - q
– And measure g via a non mixing method.
– a + b + g = a + b + g = 180o
– The triangle closure test misses this sort of NP.
• We need to be careful when we do this!
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
59
Generic Tests for New Physics
• Specific decays, non-specific models
– BKl+l- & BK*l+l-: can observe NP in distribution of
M(l+l-) and Dalitz plot is sensitive to subtle interference
effects. See hep-ph/9408382.
Reaction
B(10-6)
Yield/year
S/B
BK*m+m-
1.5
2530
11
BKm+m-
0.4
1470
3.2
bs m+m-
5.7
4140
0.13
• Check for inconsistencies in SM without reference
to a particular model.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
60
Critical Checks using c
• Silva & Wolfenstein (hep-ph/9610208), (Aleksan, Kayser &
London):
– Measure c using BsJ/y() , gg, rg.
• Can also use J/yf, but need a complicated angular analysis.
– The critical check is:
2 sin b sin g
sin χ = l
sin( b + g )
– Very sensitive since l =0.2205±0.0018
– Since c ~ 0.03, need lots of data.
– Sensitivity to sin(2c) for one year of running: ±0.024.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
61
Survey of New Physics Sensitivities
• See the BTeV “Proposal Update” for a discussion
of how many NP ideas can be tested in B decay.
–
–
–
–
–
–
–
–
MSSM and othe SUSY variants
Left-Right Symmetric models
2 Higgs doublet models
Extra d type single quarks.
FCNC couplings of the Z.
Non-commutative geometries
Mixing with a 4th generation.
Extra dimensions
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
62
Tests for New Physics (Nir, hep-ph/9911321)
• Suppose NP in Bo mixing, qD , Bo decay, qA, Do
mixing, fKp.
Model
dN/10-25
qD
qA
SM
10-6
0
0
0
asyDKp
Approximate
Universality
10-2
(0.2)
(1)
0
Alignment
10-3
(0.2)
(1)
(1)
Heavy squarks
~10-1
(1)
(1)
(10-2)
Approx. CP
~10-1
-b
0
(10-3)
• Specific pattern in each model ways of distinguishing models.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
63
Summary of New Physics
• Using b and c decays mediated by loop diagrams BTeV is sensitive
to mass scales of up to few TeV.
• The New Physics effects in these loops may be the only way to
distinguish among models.
• Masiero & Vives: “the relevance of SUSY searches in rare processes is not
confined to the usually quoted possibility that indirect searches can arrive ‘first’
in signaling the presence of SUSY. Even after the possible direct observation of
SUSY particles, the importance of FCNC & CPV in testing SUSY remains of
utmost relevance. They are & will be complementary to the Tevatron & LHC
establishing low energy supersymmetry as the response to the electroweak
breaking puzzle” (hep-ph/0104027)
• We agree, except we would replace “SUSY” with “New Physics”
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
64
Comparison with LHC-b
• Advantages of LHCb
– 5 higher b cross-section; 1.6 higher sb/stot.
• Advantages of BTeV
– Detached vertex trigger at level 1.
• Enabling technologies:
– Magnet on the IR: we can reject low p tracks at level 1.
– Pixels: very low occupancy, only 6mm from beam ( cf 1 cm ).
• Allows us to trigger on very general properties of b’s.
– PbWO4 Ecal with CLEO/BaBar/Belle-like performance.
– Trigger/DA design lets us read out 5 as many b’s/second.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
65
Comments on the e+e- Super B-Factories
• At the Y(4S), it would take a 1036/cm2/s e+e- collider
to match the performance of BTeV for Bo & B±.
• There would be no competition on Bs, Lb, …
• KEK is only proposing 1035/cm2/s.
• For Super-BaBar there are serious technical
problems for both the machine & the detector.
• We believe the cost will far exceed that of BTeV.
Recent HEPAP subpanel mentions $500M.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
66
A New Wild Card: 396 ns
• Original plan for Run II bunch spacing:
– 396 ns, for the first few years ( Run IIa ).
– 132 ns, afterwards ( Run IIb ), to increase lumi.
• Recent change: stay at 396 ns.
• BTeV originally designed to operate at 132 ns and
a peak luminosity of 21032 cm-2 s-1 ( 2 fb-1/year ).
– 2/interactions/beam crossing at peak lumi.
• For the same luminosity and 396 ns:
– 6 interactions/beam crossing at peak lumi.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
67
A Partial Solution: Luminosity Leveling
• Squeeze the beams more tightly as a Tevatron
store progresses:
– Luminosity drops more slowly.
– Can achieve 2 fb-1/year with a peak luminosity of
1.31032 cm-2 s-1.
– Only 4 interactions/beam crossing at peak luminosity
and 396 ns bunch spacing.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
68
Impact of 396 ns on BTeV
• Issues: Radiation damage, occupancy, false triggers.
• Pixels, Forward Silicon, Muon:
– Probably OK; may need to revisit shielding.
• Straws, ECAL:
– Need to carefully study occupancy. Work underway.
• Trigger:
– Two nearby background collisions can look like a b.
– Need to retune cuts and algorithms. Work underway.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
69
Peering into the Future
• Particle Physics Project Prioritization Panel (P5)
– A HEPAP subpanel created at request of DOE/NSF.
– Charge to prioritize experiments in light of the wider
and long term US HEP program.
– First charge is to review:
• BTeV
• CKM
• CDF and D0 Run IIb detector upgrades.
– Will meet at Fermilab, March 26-27, 2003.
• We will understand the implications of 396 ns by this time.
– Response due by June 2003.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
70
Peering into the Future …
• Summer 2003, test beam at Fermilab for most components.
• Presuming success in P5, there will be a Lehman review
(Fall 2003?).
• FY 2004-2008: Construction funding.
– Lab has had BTeV in long range plan for several years.
– Only small amounts of money presumed for FY2004
• For a few items with long lead times.
– Staged installation to get some components in early and to allow
early commissioning.
• 2008: Start of physics running.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
71
Summary and Conclusions
• The Fermilab director has given Stage I approval
to a revised proposal to run BTeV with only one
arm fully instrumented.
• Full DAQ and Trigger bandwidth retained.
• Aerogel RICH radiator replaced with liquid C5F12.
• We have learned use the RICH for lepton ID:
– Single(double) lepton ID efficiency up ×2.4 (×3.9).
• The reduced scope BTeV remains an excellent
detector and will be a leader in b and c physics in
the last half of this decade.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
72
Backup Slides
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
73
HPD Schematic
HPD Tube
HPD Pixel array
HPD Pinout
PC
ped
1 pe
2 pe
20 kV
3 pe
Si pixel array
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
Pulse Height from
163 pixel prototype
HPD, using old
CLEO electronics.
Note pedestal,
1, 2, 3 pe peaks.
74
Miscellaneous RICH Issues
• PMTs and HPDs will be in a region of low but
non-zero B field.
– Now testing shielding options.
• Mirror options:
– Glass
– Glass + composite
– Composite Replica
• Test pieces for first two are in hand.
• Mechanical, cooling, electronics underway.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
75
Summary of CKM Physics Reach (107 s)
Reaction
B (B)(x10-6)
# of Events
S/B
Parameter
Error or (Value)
4.5
14,600
3
Asymmetry
0.030
Bop+pBs Ds K-
300
7500
7
g
8o
BoJ/y KS , J/y l+ l -
445
168,000
10
sin(2b)
0.017
3000
59,000
3
xs
(75)
170
1
g
13o
Bs Ds pB-Do (K+p-) K-
0.17
B-Do (K+K-) K-
1.1
1,000
>10
B-KS p-
12.1
4,600
1
BoK+p-
18.8
62,100
20
Bor+p-
28
5,400
4.1
Boropo
5
780
0.3
330
2,800
15
670
9,800
30
BsJ/y ,
BsJ/y 
J/y
l+l-
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
<4o +
g
theory errors
a
~4o
c
0.024
76
Specific Comparisons with LHC-b
BR
Mode
BTeV
Yield
Bs Ds K-
3.0x10-4
Bor+p-
2.8x10-5
5400
Boropo
0.5x10-5
776
Beauty02, February 27, 2003
7530
Rob Kutschke, Fermilab
LHC-b
S/B
Yield
S/B
7660
7
4.1 2140
0.8
7
0.3
880
not
known
77
Comparisons With Current e+e- B factories
• Number of flavor tagged Bop+ p - (B=0.45x10-5)
e+ eBTeV
L (cm-2s-1) s
#Bo/107s e
eD2 #tagged
1034 1.1 nb 1.1x108 0.45 0.26
56
2x1032 100mb 1.5x1011 0.021 0.1
1426
• Number of B-Do K - (Full product B=1.7x10-7)
e + eBTeV
L(cm-2s-1) s
#Bo/107s e
1034 1.1 nb 1.1x108 0.4
2x1032 100mb 1.5x1011 0.007
#
5
176
• Bs , Bc and Lb not done at Y(4S) e+e- machines
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
78
Reconstructed Events in New Physics Modes:
Comparison of BTeV with B-factories
BTeV (107s)
Mode
Yield
BsJ/y()
B-fKBofKs
Tagged
12650 1645
11000 11000
S/B
B-fact (500 fb-1)
Yield
>15
>10
700
700
4
75
~50
4
3
Bom+m-
2000
2530
6
1
200
2530
0.7
0.1
5.2
11
>15
>10
250
~50
0
0
D*+p+Do, DoKp+
~108
~108
large
8x105
BoK*m+mBs m+m-
Beauty02, February 27, 2003
Tagged S/B
Rob Kutschke, Fermilab
8x105 large
79
Summary of Required Measurements for CKM Tests
Physics
Quantity
sin(2a)
sin(2a)
cos(2a)
sign(sin(2a))
sin(g)
sin(g)
sin(g)
sin(2c)
sin(2b)
cos(2b)
xs
 for Bs
Decay Mode
Borpp+p-po
Bop+p-  BsK+KBorpp+p-po
Borp & Bop+pBsDs KBoDo KBK p
BsJ/y, J/y
BoJ/yKs
BoJ/yK* & BsJ/yf
BsDspBsJ/y, K+K-, Dsp-
Beauty02, February 27, 2003
Vertex
Trigger







Rob Kutschke, Fermilab


K/p
sep











g det Decay
time s











80
Offline Computing Model
• Reuse Level 2/3 trigger farm.
– 2500-4000 Linux processors
– Sized to deal with peak lumi.
– About 2/3 of cycles are available for other uses:
• Lower lumi late in run.
• Machine filling, downtime etc
• Use of large computing clusters at universities.
– Grid aware but not grid dependent.
• See talk by Joel Butler Tuesday afternoon.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
81
Changes in Efficiencies wrt Proposal
• We lost one arm: Factor = 0.5
• We gained on leptons:
– We now use RICH to improve lepton ID:
• Larger solid angle; larger momentum range.
– In proposal we used only m+m-, now we include e+e-
– Factor = 2.4 (or 3.9), depending on whether analysis
required one or two leptons to be ID’ed.
• DA bandwidth constant for one arm: Factor = 1.15
• For Bs only: improved eD2: Factor = 1.3
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
82
Summary of Efficiency Changes
Mode
[Quantity]
Yield in
Proposal
BoJ/y Ks
[sin(2b)]
80,500
BsJ/y()
[sin(2c)]
Bs Ds K
[g]
Yield
Factors
New
Yield
New
eD2
Sensitivity
Proposal One Arm
0.5*3.9* 168,000
1.15=2.24
0.10
0.025
0.017
9,940
0.5*2.4*
1.15=1.38
12,600
0.13
0.033
0.024
13,100
0.5*1.15
=0.58
7,500
0.13
6o
8o
• In the proposal all eD2 were 0.10.
Beauty02, February 27, 2003
Rob Kutschke, Fermilab
83