Minutes of first MDI/Integration WG meeting

Date and time : 4th October, 2007, 2:00pm-3:40pm (Japan time)
phone meeting with the ILC Webex system
Content : Overview of LDC and GLD and discussion for joint study
Participants : K.Buesser, A.Vogel? at DESY, D.Karlen in Canada, Y.Sugimoto, A.Miyamoto, T.Omori, T.Sanuki, Y.Iwashita, K.Tsuchiya and T. Tauchi at KEK, H.Yamamoto at Tohoku university
Agenda : http://ilcagenda.linearcollider.org/conferenceDisplay.py?confId=2286

Summary

We had two presentations of LDC and GLD IR Overviews by K.Buesser and T.Tauchi, respectively. The detailed explanation with discussion can be found in following sections.

Y.Sugimoto pointed out that the charge of this working group should be clearly defined. H. Yamamoto will write the draft charge to be distributed in the working group.

Also, a mailing list of the working group was suggested for good communication. K.Buesser will prepare the mailing list.

As the first step, we agreed to list up issues and parameters which are relevant for designing a joint IR region. Especially, we should concentrate on studies of differences between the two IR design for resolution. we will discuss on these issues including the integration at ALCPG07 workshop, FNAL, 22-26 October.

Next meeting of the integration will be held in a week for 15th through 19th October.

The Interaction Region Design of LDC, Karsten Busser

LDC has size of 12m x 12m x 12m. L* and crossing angle are 4.05m and 14mrad, respectively, while 2 and 20mrad exist as alternatively. Minimum angular coverage of EM calorimeter is 5mrad with BeamCal. Calorimeters are centered on outgoing beam. Material in front of LumCal is minimized for luminosity measurement with high resolution. LHCAL is installed in a free space behind of Lumcal as hadron calorimeter, but there is no detailed study for performance.
Q : Where are physical centers in the calorimeters ?
A : They are aligned around z-axis which is center of LDC detector. The Lumcal and Beamcal are segmented around the extraction line for better performance which has been studied by FCAL collaboration.
The recent modification of LDC V5 includes LumiCal simplification for detector opening procedure, ECAL for hermeticity and addition of silicon pixel device in front of LumCal to measure positions of incident particles. Since it is preliminary, a detailed study is necessary.

The endcaps can be moved out in 2.5m longitudinal direction for maintenance of vertex detector in the garage position without breaking the vacuum.

Q : Is vacuum kept during the push-pull operation, too ?
A : Yes, the vacuum can be kept with gate valves between QD0 and QF1.
Forward region has been designed with respect to pair background; trapping back-scattered particles in the area between LHCAL and BeamCal surrounded with tungsten shield and Low-Z absorber is set in front of BeamCal.

Background hit rates are shown as a function of the VTX layer in three machine parameter sets (LowP, nominal and TESLA at Ecm=500GeV).

C : We would like to see the hit rate/cm2 in order to be compared with tolerance.
C : Thickness of low-Z absorber looks to be 20mm if it is compared with one in page 15 which shows that 50mm thick absorber can reduce hit rate by half.
Also, neutron background has been estimated in the VTX for nominal runtime with 500fb-1 of ILC-NOM-500 and ILC-LOWP-500. Results show that neutrons can be tolerable for CCD-VTX even with the estimated errors of more than 100%.
Q : What is source of neutrons ?
A : Neutrons were generated within ±10m from IP by pairs. As shown in figure, the dominant source point is BeamCal.
C : There was a presentation on neutrons at IRENG07, which showed that neutrons from the beam dump may contribute 100 time more background in VTX than those generated by pairs.
A : Yes, it has to be studied. I understand that this issue comes with large crossing angle since there is no such problem in the headon collision at TESLA. It is actually difficult to estimate neutron background whose uncertainty would be a factor of 10.
Q : What is the horizontal scale in energy spectrum of neutrons ?
A : It is logarithmic scale, so 3 and 6 means keV and MeV, respectively.
TPC background hits have been simulated by Mokka. The radial distribution and the time structure were shown.
Q : Can you explain more the time structure?
A : It shows overlay events of 100 BX starting time from BX. So, prompt particles create hits at t=0 and backscattered particles ( photons) create hits t= about 28ns from BeamCal.
Q : Only pairs were simulated ?
A : Yes. As you may see, the simulation includes radiative excitation of nuclei.
Q : How much is the total drift time in TPC?
A : It is 40usec corresponding to 150BX.
Performance of BeamCal has been studied in order to estimate the detection efficiency of 2-photon events with pair background by FCAL collaboration. Also, the beam parameters have been reconstructed by using BeamCal as beam diagnostics instrument.
Q : How much is statistics in the error estimation of reconstructed parameters ?
A : It corresponds to a single bunch crossing (BX).
In summary, LDC IR design is optimized with respect to background suppression and low angle instrumentation (LumiCal of precise luminosity measurement and BeamCal of hermeticity to 5mrad). Detailed design depends on full detector simulations which are very time consuming. Engineering solutions exist on conceptual level.
Q : Is there any background problem in ECAL and HCAL ?
A : There is no major problem. Effect on electronics of HCAL might be evaluated.
Q : did you estimate the hit rate in them ?
A : We have estimated radiation dose and energy deposits. The number can be available if necessary.
Q : How much is the outer radius of the support tube ?
A : It is 29cm.
C : The GLD support tube has 35cm radius, which has been designed for a large bore SC final quadrupole magnet at small angle crossing. Some space may be needed for fine adjustment system for the standard QD0 in the support tube.
C : There is small difference in location of LumiCal (LDC) and FCAL(GLD). Front face of LumiCal is the same as ECAL, while FCAL is slightly in front .
A : Shower leakage in ECAL from LumiCal has been estimated, which prefer such geometry.
C : BeamCal is located as far as we could from IP in LDC for lower angle measurement.

GLD IR Overview, Toshiaki Tauchi

GLD has size of 14.4m x 14.4m x 15m. Compact GLD ( GLDc ) has size of 13.8m x 13.8m x 14m, which has been proposed as working assumption towards a joint detector as well as for the push-pull scheme.

First, background tolerances were shown for VTX, TPC and CAL for designing IR region. The sources are pairs, neutrons and muons. Tolerable hit rates are 1 x 104 hits/cm2/train and 4.92 x 105 hits/50usec in VTX and TPC, respectively. This VTX tolerance has "Hit finding efficiency" of more than 93% at 0.5GeV momentum in VTX with 5um x 5um pixels (FPCCD).

GLD IR design has been optimized on beam pipe, VTX innermost layer with respect to pair-background and synchrotron radiation profile, FCAL inner radius for background hits in TPC and BCAL hole radius for out-going particles etc. .

The beam pipe has been designed along with envelop of core of pair-background in the detector solenoid field B=3T . Also, the VTX is required to cover angular region of |cos|<0.95, where the sensitive silicon wafer length is longer by 2mm than the angular coverage and the ladder length is the wafer length + 15mm . The (inner) radius of beam pipe must be larger by 2mm from the core along the beam line, while this offset must be larger than 5mm at FCAL. A straight beam pipe with the minimum radius (RBe) is made of Berylium. The innermost VTX radius ( RVTX ) must be larger by 2mm than the beam pipe .

The core envelops have been estimated with the nominal and high luminosity beam parameters at Ecm=500GeV, B=3,4 and 5 T and with high luminosity beam parameters at Ecm=1000GeV, B=3 T. RVTX has a weak B dependence as B-1/2, while it has a strong dependence on the beam parameters. So, we have designed the beam pipe and RVTX with most conservative parameter set of "High luminosity Option 1" at Ecm=1000TeV, i.e. RBe=1.5cm and RVTX =2.0cm as the nominal GLD geometry.

Background hit rates have been simulated in VTX and TPC with the nominal, LowP beam parameter sets and the nominal with anti-DID (Ecm=500GeV) by Jupiter . All the resultant hit rates were well smaller than the tolerance of 10,000 hits/cm2/train. Also, the TPC hit rates were smaller than the tolerance.

Q : How many BXs the right figure has for the TPC hit distribution ?
A : Not a single BX, but I do not know how many.
FCAL inner radius (RFCAL) has been determined to be 8cm in order to shadow the back-scattered photons from BCAL as shown in figure, page 13. The front face of FCAL is located at 2.3m from IP, while the BCAL is 4.3m from IP. The inner radius of TPC is 0.45m. The number of particles entering the TPC was calculated as a function of RFCAL by Jupiter. There was no dependence on RFCAL in the range from 7cm to 10cm .
Q : What is the energy threshold for photons in simulation ? p.14
A : I do not know exact value. The photon energy has a dominant peak at 0.5MeV due to positron annihilation.
C : It might be interesting to see the TPC time structure as a function of LumiCal radius in LDC, if possible.
Synchrotron radiation profile can be controlled by collimators (SP2, SP4 and SPEX) and masks (MSK1, MSK2) at the collimator section and the final focus section, respectively. The apertures have been optimized by A.Drozhdin for higher B field ( BDIR05) . We have simulated the profile by LCBDS based on GEANT4. LCBDS is very similar to BDSIM. The figure shows the profiles at IP, page 16, where yellow and red parts show the profiles of beam-core and halo with the collimators, while blue part shows those without MSK1 and MSK2. Vertical scale in projected figures is an arbitrary unit. The nominal parameter set has been used for this study at Ecm=500GeV, L*=3.5m and 20mrad crossing angle. As we can see the red-profile, the synchrotron photons are "collimated" in R< 1cm at IP . So, they should be no problem in GLD for RBe=1.5cm .
C : The collimator/mask apertures depend on L*. In general, longer L* requires smaller apertures.
Q : Do major SR photons come from final doublet ?
A : Yes. But, you can see contribution from SR photons from upstream magnets in blue part.
One reason of longer L* in GLD is smaller backscattered charged particles as shown in left figure, page 17. The charged particles are produced at BCAL, i.e. outside of holes, typically R>1.8 cm . Since BCAL is set just in front of QD0 ( final focus Q-magnet ) and the detector solenoid magnetic fields are bent outward in R, the longer location of BCAL can help the charged particles are trapped inside of the beam pipe.

Finally, opening procedure of GLDc in the garage position was shown. First, the endcaps are moved out by 1.6m and they are split for making a space enough for TPC extraction. Also, the VTX can be accessed for maintenance. The detailed procedure will be discussed at the next meeting.

In summary, GLD will evolve to GLDc for the push-pull scheme, while we need detailed evaluation for optimization with full simulation. GLD IR region has been optimized with respect to backgrounds ( pairs, synchrotron photons, muons ..) at VTX, TPC and minimum veto angle for 2 photon process. Relevant parameters for IR optimization are listed; which are machine parameter set, L*, B, RBe, RVTX, VTX angular acceptance, RFCAL, RBCAL and support tube for QDO, FCAL, BCAL .

C : GLDc has been proposed to be an average of GLD and LDC as working assumption to study a joint detector (ILD) .