Cooling Tests

1
Cooling Tests

• Hubert Becker, Philippe Bouvier,Allan Clark, Hans
  Dietl, Greg Hallewell, Tapio Niinikoski, Eric
  Perrin, Vic Vacek, András Zsenei
• Presented by
• András Zsenei

2
The setup of the first run
3
The split circumferential cooling block

• Sensor and hybrid heat paths separated
• C-C block connected with thermal grease to the
  pipe (Araldite on the edges)
• Al block glued with Araldite to the pipe
• 70/30 CuNi pipe with Dout3.7 mm, 50 µm thickness

4
Details of the split cooling block
5
The simple model
Qh
Qs
Rhs
Th
Ts
Rh
Rs
Tcoolant
6
A more complete model
Qh
Qs
Rfan
Rsub
Rspi
Rhs
Th
Ts
Rh
Rs
Tcoolant
7
Thermal resistances of Run 1
8
Comments

• For the C-C block the values were deteriorating
  over time so we think that maybe the glue did not
  support well enough the thermal cycling, so the
  grease contact became worse. The values in the
  table are the values measured at the beginning of
  the experiment.

9
Layout of blocks for Run 2

• All cooling blocks are C-C
• All glued to the CuNi pipe with silver-loaded
  epoxy with a thermal conductivity of 1.7 W/Km
• Thermal conductivity of blocks 400 W/mK and
  75W/mK
• Thermal grease contact with the modules

PT100
10
The setup of the first part of the second run (2
half-modules)
11
Calibration plot for one of the PT100s on the pipe
12
The calibration procedure
13
Heat transfer coefficient
14
Results of the first part of Run 2
15
Comments

• The HTC is lower than in the simulati-ons
  (J.Blocki took 4000 for the hybrid part of the
  block and 3000 for the other parts)
• Rhs500 is rather optimistic not taking into
  account conduction, convection, and shorts by the
  thermal grease
• Alignment problems can deteriorate the resistance
  values

16
Thermal module

• TPG spine 20mm wide
• Araldite 2011 glue
• AlN spacers, except in the fan-in area where they
  are made of Al2O3
• Fan-in material glass
• W31 and W32 dummy detectors from CSEM
• Sensor heaters MINCO 9931-9812/HK5438R140L12B

17
The setup with the full thermal module on the
middle block
18
Thermal module on middle block 1-point cooling 1W
on sensors, high ambient temperature (-10.3)
32
-9.3
9.6
-9.1
16.4
-10
-11
-9.8
Coolant _at_ -26.4

• Uniformity 1.7 ºC

19
Thermal module on middle block 1-point cooling 2W
on sensors, high ambient temperature (-10)
32.6
-3.8
11.7
-4
17
-3.6
-5.7
-3.9
Coolant _at_ -26.4

• Uniformity 1.9 ºC

20
Thermal module on middle block 1-point cooling 1W
on sensors, low ambient temperature (-23.2)
28.8
-14.4
3.9
-14.5
14.7
-15.7
-15.6
-16.2
Coolant _at_ -26.5

• Uniformity 1.8 ºC

21
Thermal module on middle block 1-point cooling 2W
on sensors, low ambient temperature (-23.8)
29.3
-9.4
5.9
-9.8
14.7
-9.9
-10.5
-10.9
Coolant _at_ -26.5

• Uniformity 1.5 ºC

22
Thermal module on left block 1-point cooling 2W
on sensors, high ambient temperature (-10.6)
28.2
2.2
8.5
0.7
4.7
0.7
1.2
-0.9
Coolant _at_ -26.6

• Uniformity 1.5 ºC

23
Comments

• Effects of the ambient temperature are very
  significant 5-6 ºC change on sensors for 14 ºC
  change in ambient
• Good contact is crucial compare the 4.7 ºC in
  the middle of the hybrid when it was on the left
  block with the 17 ºC in the same place under the
  same conditions when it was on the middle block
  (good contact lowers T by 12 ºC )

24
Results from the second part of Run2 module on
middle block
25
The setup with the full thermal module on the
left block
26
Results from the second part of Run2 module on
left block
27
Comments

• Do not take into account Rs for the thermal
  module on the left block. It looks very nice but
  it is an artifact of the bad contact between the
  block and the module (see slide 29). One has to
  check always whether the module is sufficiently
  cooled. All the other resistor values has been
  checked and look OK.

28
Thermal resistances for the half-modules (more
complete model)
29
Difficulties

• Sensor path did not have good contact
• Geometry problems of the setup
• Planarity of the cooling block and the module

30
Achievable temperatures
Rh 2.1
Rs 2.2
-5.3
Rhs 31
-9.4
Rsub 1.7
Rfan 125
Rspi 2.0

• Supposing good machining, correct mounting,
  reasonably good contacts and adequate alignment

31
Contact between the cooling block and the pipe

• Glue can be used, however jigs have to be
  developed (effects of ageing, radiation and
  thermal shocks have to be understood)
• Ideally soldering should be used (both on Al and
  C-C blocks) see presentation of Serge Mathot in
  the engineering meeting

32
Conclusions

• Split circumferential cooling block can work
  satisfactorily
• Precise machining and mounting are necessary to
  achieve good results
• Good uniformity across the sensors
• Convection has a big impact and has to be taken
  into account

33
Comments

• One has to look at how much power the thermal
  enclosure can absorb
• Cooling the discs is probably necessary
• Sensitivity of contact due to the small contact
  area is worrying