Minutes of ILD MDI/Integration WG meeting, 12 August, 2008

1. Progress report of ILD2

Saclay meeting by K. Buesser

We had a meeting with CMS-solenoid magnet designers at Saclay. We brought many questions to them for the ILD yoke design at DESY.

The minutes was already distributed.

We agreed parameters of coil as a working assumption for designing iron structure and anti-DID by DESY and B.Parker (BNL), respectively, which are called "ILD L3 Saclay"; rin = 3.4 m, rout = 3.75 m, length = 7.35 m and B = 4T.

The another interesting issue was the cold coil movement in the push-pull. F. Kircher ( CMS magnet designer) denied a possibility of the cold movement, if the ILD coil follows the CMS design. So, the push-pull scheme would not work since it takes a month to warm up and cool down the coil.

According to him, the supports between the CMS coil and the cryostat may be sensitive to vibrations when cold, especially titanium sheet which easily develops cracks under pressure even on a platform. We may need specific R&D effort for a push-pull coil.

He will come to the Cambridge meeting, although it depends on the CMS schedule. So, we can discuss this issue in more details.

DESY has started to simulate the fields with three programs of MAFIA, OPERA and ANSYS. The endcap must be thicker for smaller movement of 2mm compared with cm-order at CMS. For the tail catcher, iron structure should have 4 and 5 layers of 10cm thick iron plate in barrel and endcap, respectively, at least.

Q : There are frequently earthquakes in Japan. Vibrations could be larger than those on a platform. If so, the CMS coil would be broken in Japan. However, there are many coils in Japan. We do not understand such fragility.
A : Detailed description of the CMS coil can be found in the CMS technical design report.
C : LC-TPC does not require a field uniformity of "2mm" , thought earlier (e.g. for the Aleph coil), as was shown in the LC-DET-2008-002 at http://www-flc.desy.de/lcnotes.
A : We understand from that LC Note that precise field measurement is essential. Saclay studies show that only field correction coils can not make such uniformity, but shaping plates of 20cm thickness can make 3mm of uniformity.
However, as just said, 2 or 3mm is not needed, and Francois Kircher says that around 6-8mm will be possible without the thick plate, which is therefore not needed since 6-8mm is fine for the LCTPC.
Q : How significant is stray field or how to manage it?
A : In order to reduce it, gaps are the most important which has been studied by Y.Sugimoto and confirmed by F. Kircher. The magnitude must be less than 200Gauss outside of the detector. However, we may ask the tolerance in longitudinal direction, i.e. along the beam line, if 3m thick iron would be needed in endcap.

Tail catcher by F.Sefkow

Tail catcher issue was introduced by referring my (Felix) presentation at Sendai meeting and Zutshi's (SiD, CALISE) slides. ( 16-sectors of HCAL have less leakage than 8-sectors, which will be another argument in " 8 v.s. 12 shape", where leakage is an escaped energy of more than 5% of incident energy. )

The ILD (SiD) HCAL depth ( 5&lambda) is lower than 10&lambda in the 4th concept, which has been criticized at every opportunity. The leakage has been measured in 10&lambda of "HCAL+TCMT" by the test beams. Also, it has been simulated in order to see improvement of energy resolution with tail catcher by V.Zutshi (NIU). In the simulation, the tail catcher consists of 0.5cm - 4.0cm thick scintillators interleaved with 10cm thick absorber plates. Even with coil behind HCAL, the improvements could be seen. So, it seems that the tail catcher could improve the HCAL performance.

In conclusion, it is too early to discard the option. It would need to be addressed in more serious studies ( PANDORA ).

C : It is good starting point to assume 10cm thick iron plates.
Q : We need a contact person, since the tail catcher significantly impacts on design of iron yoke such as informations of cables, gaps in addition to thickness of absorbers.
Q : There is already muon system in ILD. What does it differ?
A : In the (GLD) muon system, the thickness of iron plates is 25cm at the first layer, while the tail catcher may require 10cm thickness.
C : The 10cm thickness is big challenge for strength against huge magnetic force, especially in the endcap.
A : It could be done. So, DESY will design the 10cm case.
C : Since the magnetic force is large, thicker is more preferable. It is 60cm in CMS. In the endcap, there is no coil, then HCAL can be thicker with thicker iron plates.

2. Progress report of ILD1

8 and 12 shape in the barrel calorimeters by T. Takeshita

The differences were summarized in terms of (1)overlap in ECAL, (2) readout space between ECAL and HCAL, (3) "dead" space between TPC and ECAL, and (4) azimuthal asymmetry of depth in ECAL ( or path length of a muon in ECAL) . The readout card assumes to have 5cm width.

In conclusion, those differences listed especially (1) and (4) should be tested by simulations for the physics point of view for their effect.

C : Karsten ( who is not MDI person) has pointed out that HCAL should have the same symmetry as the ECAL.
C : If we require the same minimum thickness in the calorimeters, the 8 shape has larger structure. So, the 12 shape is more cost-effective design.

"cylindrical" support tube by H. Yamaoka

Overall sizes of the support tube are 6m length, 75cm and 65cm outer and inner diameter, respectively, i.e. made of 5cm thick cylindrical tube. We studies choice of material (SUS, Aluminum), structure models of full, half cylinders and half one with ribs, and static and modal/dynamic analyses for deformation and vibration properties, respectively. In the analyses, the support tube is assumed to be a cantilever with a fixed point, which is placed outside of the detector. Loads have been provided by Matthieu Jore, which are 700Kg of QD0, 100Kg of BeamCAL, 3,000Kg of LHCAL, 250Kg of LumiCAL and 420Kg of ECAL-ring in addition to the self-weight. In the dynamical analysis, the support tube is vibrated at the fixed point by measured ground motion (GM), where typical GM acceleration is 2 x 10-7m/s2 at frequency of greater than 0.3Hz.

Results are summarized in a table. Deformations of SUS half and full cylinders are 19.7(46.9) and 3.2(6.7)mm, respectively, where values in parentheses are those of Al. The deformations can be reduced by about half with ribs. Vibration amplitudes of SUS(Al) half and full cylinders are 7.8(15.3) and 2.7(3.0) nm, respectively, which can be reduced by about half with ribs, too. The first natural frequencies of SUS(Al) half and full cylinders are 3.7(2.3) and 9.5(6.3) Hz.

In conclusion, the half cylinders are too weak in terms of deformation which should be less than a few mm. However, the support tube must be split into two halves for installation of QD0-system. So, SUS seems to be the best for realistic structure, i.e. bolting of 2 half cylinders with ribs .

Q : Can you also calculate the square support tube?
A : Yes in principle. However, that can be estimated by Matthieu Jore .
Q : Can we install QD0 in the full cylinder ?
A : It may be possible if there is enough space along the beam line. Since QD0 must be connected to beam pipe and cryo-lines, it must be difficult solution.
Q : Did you take account vibrations due to He and power lines ?
A : No, the ground motion is only input to the dynamical analysis.
C : Major conclusion of this work is the possibility of cantilever fixed outside of the detector.
C : Yes, it should to be necessary, since the iron endcap must slide out without deforming the support tube.

3. Discussion

As already distributed in the ILD MDI/I mailing list, the coil parameters were sent to B. Parker to design an anti-DID in ILD.

We will discuss issues of coils, iron structures in more details at the Cambridge meeting, since F. Kircher, Alain Herve will come as the CMS technical experts.

Q : When will Brett produce the design ?
A : I(Ron) will ask him. We expect that it will be available by the Cambridge meeting .

Next meeting will be 2 September, 2008.