1 David Hitlin Frascati SuperB Workshop March 17, 2006.

David Hitlin Frascati SuperB Workshop March 17, 2006 1

David Hitlin

Frascati SuperB WorkshopMarch 17, 2006


Reusing existing detectors at SuperB With luck, we will soon have to face the question of whether

SuperB should have a new detector built from scratch, or whether an existing detector (BABAR, Belle, or even CLEO-II) could be upgraded to do the job

In order to meaningfully discuss this question in detail, it is necessary to nail down a few important parameters Do we collide every bunch or collide trains kicked out of rings? The energy asymmetry (7x4, 8x3.5, 9x3.1) Two beams vs four beams

A four-beam machine can have boosted decays in both the “forward” and “backward” directions. This is a major perturbation on the design Requires a detector that has essentially the forward

section of BABAR or Belle in both the forward an backward directions A much longer solenoid and flux return is needed


By the end of the year, we may need to provide preliminary details on schedules, costs and R&D needs for both the collider and detector

In order to confront the detector side, it is helpful to have a specific configuration in mind An upgrade of either BABAR or Belle would, at first glance,

seem to be a perfectly adequate detector for 1036 at SuperB A first look at modifying BABAR, however, turns up issues

Using BABAR as the foundation seems feasible, however Using Belle would likely generate essentially a similar end

result (cf Tim Gershon’s talk) A first look turns up a variety constraints, many of which

are generic

We need to grapple with this question now


Collider scheme has a direct effect A linear collider-type machine and a storage ring-type

machine have very different time structures and currents These differences directly effect detector design Sensitivities

Beam pipe diameter and thickness (cooling?) Trigger/DAQ system - fundamental Response time of detector subsystems

Tracker Electromagnetic calorimeter

Radiation hardness


Linear collider vs storage ring Linear collider

Small diameter, uncooled beampipe “SLD” like DAQ CsI(Tl) OK, at least for the barrel Drift chamber tracker

Storage ring with LC final focus Cooled beampipe thicker material (two walls + water)

likely larger diameter SuperBABAR type DAQ Fast, radiation hard EMC (LSO/LYSO, pureCsI)

Barrel/endcap radiation sensitivity depends on luminosity term

Fast decay time is an advantage A drift chamber is marginal

A silicon tracker presents a multiple scattering problem


Backgrounds Naïve scaling by current indicates that the occupancy

and radiation damage issues at a conventional 1036 machine are much reduced at a linear collider type machine and are slightly better than SuperPEP-II or SuperKEKB in the new storage ring scheme, due to reduced circulating currents (scaling as the current, if the PEP-II luminosity term can be controlled)

Background sources Current-related Luminosity-related Beam-beam Touschek (intrabunch scattering) Synchrotron radiation

We must move beyond naïve scaling to a Decay Turtle type calculation of lost particle backgrounds


Detector elements – BABAR foundation

More conservative Less conservative

Retain Rebuild/Add Retain Rebuild/AddSolenoid SVT(Strips) Solenoid SVT(Pixels+strips)

Flux return Drift Chamber Flux return Drift Chamber

Barrel EMC Endcap EMC(s) Barrel EMC

LST’s DIRC SOB Endcap EMC(s)

Trigger/DAQ Barrel PID

Forward PID

IFR (especially EC)

Trigger/DAQ

A constant of the motion: remove the Support Tube


SuperB BABAR strawman – I

M A G N E T

C A B L E S

CL

C

Q 5Q 2

C Y L I N D R I C A L R E S I S T I V EP L A T E C H A M B E R

D .C .E L E C T .

Q 4Q 2

Q 5

1 1 4 9

D R I F TC H A M B E R

C A B L E S

5 0

F L O O R

1 6 5 5

3 7 0

E N D D O O R F L U X R E T U R NP L A T E S E G M E N T AT I O N

( 9 ) 2 0 M M T H K P L A T E S

( 4 ) 3 0 M M T H K P L A T E S

( 4 ) 5 0 M M T H K P L A T E S

( 1 ) 1 0 0 M M T H K P L A T E

+

-

R E S I S T I V E P L A T E C H A M B E R S I N G A P S( 1 9 ) P L A N E S I N B A R R E L

( 1 8 ) P L A N E S I N E N D D O O R S

C U R TA I N W A L L

S U P E R C O N D U C T I N G S O L E N O I D

F L U X

R E T U R N

Q 4

F O R W A R D E N D P L U G

C O U N T E R W E I G H T1 5 5 5

3 0 0 0

-

+

B A R R E L F L U X R E T U R N

P L A T E S E G M E N T AT I O N

( 9 ) 2 0 M M T H K P L A T E S( 4 ) 3 0 M M T H K P L A T E S

( 3 ) 5 0 M M T H K P L A T E S ( 2 ) 1 0 0 M M T H K P L AT E S

3 2 0 9

3 5 0 0

D IR CE L E C T .

1 5 0

2 3 9 5

1 7 .2 0

2 2 9 5

4 5 6 6

3 2 0 9

e

e

e

e

4 5 0 0

3 1 5 0

1 1 4 9

D E W A R

S V T

B A R R E L C A L O R IM E T E R

5 5 0 0

H O R S E C O L L A RA S S E M B LY

1 3 8 4

F

F

C

C

A

A

L

L

F

B


Detector protractor – 9 on 3.1 GeV ’s

lab9 G eV on 3.109 G eV

60 o60 o

80 o80 o

70 o70 o

50 o50 o

40 o40 o

30 o30 o

20 o20 o

10 o10 o

0 o0 o

90 o

c o s c m

.7.5

.5

.2

.2

0

.4

.4

.1.1

.6

.6

.3

.3

.7.8

.8

.9

.85

.92.94

.96

.98

.99 .9.85

.92.94 .96.98

(4 ) =0.56S

BACKWARDPOLAR ANGLES

FORWARDPOLAR ANGLES


Detector protractor – 8 on 3.5 GeV ’s


Detector protractor – 7 on 4 GeV ’s



M A G N E T

C A B L E S

CL

C

Q 5Q 2

C Y L I N D R I C A L R E S I S T I V EP L A T E C H A M B E R

D .C .E L E C T .

Q 4Q 2

Q 5

1 1 4 9

D R I F TC H A M B E R

C A B L E S

5 0

F L O O R

1 6 5 5

3 7 0

E N D D O O R F L U X R E T U R NP L A T E S E G M E N T AT I O N

( 9 ) 2 0 M M T H K P L A T E S

( 4 ) 3 0 M M T H K P L A T E S

( 4 ) 5 0 M M T H K P L A T E S

( 1 ) 1 0 0 M M T H K P L A T E

+

-

R E S I S T I V E P L A T E C H A M B E R S I N G A P S( 1 9 ) P L A N E S I N B A R R E L

( 1 8 ) P L A N E S I N E N D D O O R S

C U R TA I N W A L L

S U P E R C O N D U C T I N G S O L E N O I D

F L U X

R E T U R N

Q 4

F O R W A R D E N D P L U G

C O U N T E R W E I G H T1 5 5 5

3 0 0 0

-

+

B A R R E L F L U X R E T U R N

P L A T E S E G M E N T AT I O N

( 9 ) 2 0 M M T H K P L A T E S( 4 ) 3 0 M M T H K P L A T E S

( 3 ) 5 0 M M T H K P L A T E S ( 2 ) 1 0 0 M M T H K P L AT E S

3 2 0 9

3 5 0 0

D IR CE L E C T .

1 5 0

2 3 9 5

1 7 .2 0

2 2 9 5

4 5 6 6

3 2 0 9

e

e

e

e

4 5 0 0

3 1 5 0

1 1 4 9

D E W A R

S V T

B A R R E L C A L O R IM E T E R

5 5 0 0

H O R S E C O L L A RA S S E M B LY

1 3 8 4

F

F

C

C

A

A

L

L

F

B


60 o60 o

80 o80 o

70 o70 o

50 o50 o

40 o40 o

30 o30 o

20 o20 o

10 o10 o

0 o0 o

90 o

c o s c m

.7.5

.5

.2

.2

0

.4

.4

.1.1

.6

.6

.3

.3

.7.8

.8

.9

.85

.92.94

.96

.98

.99 .9.85

.92.94 .96.98

(4 ) =0.56S




60 o60 o

80 o80 o

70 o70 o

50 o50 o

40 o40 o

30 o30 o

20 o20 o

10 o10 o

0 o0 o

90 o

c o s c m

.7.5

.5

.2

.2

0

.4

.4

.1.1

.6

.6

.3

.3

.7.8

.8

.9

.85

.92.94

.96

.98

.99 .9.85

.92.94 .96.98

(4 ) =0.56S


Support tube

The support tube has some construction/alignment advantages, but the resolution penalty due to multiple scattering is severe It should be done away with in an upgrade Mount pixels/SVT on the beam pipe or on main tracker


Vertexing The SVT we have been discussing, originally based on the

SuperBABAR concept, involves two initial pixel layers followed by a ~5 layer SVT, starting at a very small radius (≤1 cm) This has sufficiently good primary vertex resolution to allow an

energy asymmetry as small as 7x4 GeV if the beam pipe radius is ≤10mm

In the storage ring-based design, things may be different With currents of 1-2 A, water cooling required A cooled beam pipe would have more material and would

require a larger radius for the first tracking layer What is the minimum practical beam pipe radius?

Needs a preliminary IP design and simulation


Main Tracking In a linear collider, a drift chamber (even a jet chamber)

would be the clear choice In the storage ring design, the viability of a gas-based solution

remains to be demonstrated Drift Chamber

Carbon fiber mechanics would be advantageous for the endplates, especially if we want to deploy endcap PID

Where do we mount the on-chamber electronics? In BABAR, the drift chamber electronics is entirely in the

backward direction Can tracking solid angle be extended below ~300 mrad?

Silicon Tracker Requires a large area of silicon (CMS=200 m2) Must perforce be double-sided, with Si thickess < 200m Even so, mass resolution is worse than for a gas-based tracker

Choice requires a detailed understanding of backgrounds


EMC Should the barrel CsI(Tl) calorimeter be retained?

There is already some radiation damage observed (barrel & EC) The mechanical structure of the existing EMC barrel is a major

constraint Projective towers, displacement of collision point

Calorimeter would have to be completely disassembled for shipping

Several of the important new physics objectives (most are related to recoil-related studies) make hermeticity increasingly important

This motivates the addition of a backward EMC endcap Both a forward and backward EMC endcap should be fast and

radiation-hard (LSO/LYSO), as should the barrel, if it is replaced

Smaller Molière radius and radiation length and fast decay time are a significant advantage


EMC Projectivity & mechanics EMC crystals are projective in , very nearly in The projective geometry itself is independent of boost The offset of the IP, meant to optimize solid angle

coverage for a given boost, is a non-negliglible constraint on other boosts

If the EMC mechanics were to be rebuilt: By removing crystals from the forward barrel and adding new

crystals at the rear, one could optimize for a lower boost By stiffening the carbon fiber egg crate structures with an

inner carbon fiber wall, one could reduce the dead material between crystals, improving the energy resolution

If the barrel were taken apart, the shaping time constants could be re-optimized

Were both the barrel and forward endcap to be rebuilt, a geometry with no real barrel/endcap break (à la H1) could be built without precluding access to the tracker


Particle ID The DIRC water standoff box is a source of background from

beam-related particles In a detector for the storage ring design it is desirable to remove

the SOB Doing so depends on development of compact DIRC readout

It is difficult, but perhaps not impossible, for a barrel DIRC and a rear EMC endcap to coexist Even if a readout that works in a magnetic field is developed,

there would be a substantial amount of material in the barrel/rear endcap corner

Endcap PID A proximity-focussed Cherenkov ring imaging device with

aerogel radiator(s) appears to be a good choice Presents 10-20% X0

Requires ~ 30 cm of space


PID geometry interacts strongly with the EMC

The current BABAR DIRC readout precludes a useful backward EMC endcap

An evolved design, with a quartz standoff and pixel pmt readout in a magnetic field poses its own severe limitations on an upgraded EMC If the readout is brought out beyond the barrel EMC, there

is perforce an awkward break between the barrel and BEC If the readout is inside the EMC, the barrel/EMC break can

be much more graceful, but there will be a rather large concentration of high Z material in the EMC corner region


SVT/Main tracking upgrade options

SVT Option Pros Cons/QuestionsVery small radius uncooled beampipeTwo layers of pixelsFive layer SVT

Most robust pattern recognition Best vertex resolution Allows lowest energy asymmetry

Most expensive Are pixels really needed?

Small radius cooled beampipeFive layer SVT

Less uncertain (at least until IP is designed and backgrounds calculated)

Vertex resolution is worse Energy asymmetry greater

In all options: remove support tube

Tracker Option Pros Cons/QuestionsDrift chamber Best momentum resolution Most expensive

Are pixels really needed?

Silicon strips Best rate capability Better solid angle coverage Less material in front of EC PID/EMC

Much more expensive Worse momentum resolution


EMC upgrade options

Option Pros ConsRetain Barrel & Forward Endcap (FEC)

[Add BEC]

Cheapest Radiation damage impairs performance Must disassemble to move Restricted solid angle [Awkward barrel/BEC interface]

Retain Barrel, new FEC

[Add BEC]

Second cheapest Improved FEC segmentation, resolution, speed

Radiation damage impairs barrel performance Must disassemble barrel to move [Awkward barrel/BEC interface]

Retain Barrel crystals, rebuild mechanics +New FEC and BEC

Allows better BEC geometry Can improve material, electronics Improved solid angle Improved FEC, BEC segmentation, resolution, speed

Radiation damage impairs barrel performance

New LSO EMC Better resolution, speed, segmentation More flexible geometry

Most expensive


PID/IFR upgrade options

PID Option Pros ConsBarrel - DIRC Proven concept

Good momentum range

Material in front of EMC Difficult to have backward EMC endcap

Endcap(s)

Proximity focussed Cherenkov

TOF

Extends PID solid angle

Extends PID solid angle

Material in front of EMC endcap(s) Compatible momentum resolution?

Material in front of EMC endcap(s)Extremely good timing in magnetic field required Compatible momentum resolution?

IFR Option Pros ConsBarrel LST Likely has adequate rate

capabilty and radiation hardness

Endcap LST May have adequate rate capabilty and radiation hardness

May require substantial additional shielding in upstream direction


Trigger/DAQ upgrade options

See talk by Gregory Dubois-Felsmann


Comparison – BABAR and Belle for SuperB

From Yamauchi’sHawaii 2005 talk


Comparison - II

From Yamauchi’sHawaii 2005 talk


Moving forward We need an R&D plan

Formulate R&D objectives Develop an R&D schedule Formulate a budget See David Leith’s talk

We need a reliable cost estimate of a BABAR (or Belle)-based upgrade


Conclusions The details of a SuperB detector depend intimately on

whether SuperB is a linear collider or a (new type of) storage ring

Parenthetic remark In 1977 I ran a group at Aspen that was charged with

designing (in the Summer Study sense) a detector for what eventually became the TevatronAt that time, it was undecided whether the machine would be based on collisions between the existing 200 GeV machine and a new 1 TeV superconducting ring, or whether the collisions would take place solely in the 1 TeV ring.In other words, would the CM be stationary in the lab, or moving in the lab (perhaps that’s where Oddone got the idea)The SuperB detector issue is trivial by comparison

BABAR, with very substantial upgrades, provides a suitable platform for a SuperB detector

1 David Hitlin Frascati SuperB Workshop March 17, 2006.

Documents

existing detector b

b ar type daq

forward section of b

new detector

linear collidertype

linear collider type

beam machine

adequate detector