Study on DIRC barrel polishing in PANDA - PowerPoint

							Study on DIRC barrel polishing in PANDA
Annalisa Cecchi INFN Ferrara & GSI for Ferrara group PID Session 27th March 2007

PANDA Collaboration Meeting, Genova, 26-29 March 2007

Outline
• Overview on the requirements for a DIRC detector • Reflection coefficient and number of reflections • Calorimetric test in Ferrara • Conclusions

Annalisa Cecchi, 26th Mar. 2006

2

DIRC detector
Quartz barboxes (Cherenkov radiators) must transmit the light over long optical pathlengths with a large number of internal reflections.

This imposes a number of stringent requirements on the bars properties.
Annalisa Cecchi, 26th Mar. 2006
3

Which requirements?
• The internal reflection coefficient must be close to one (R > 0.999)
• The high surface reflectivity must be mantained for the timelife of the experiment

• The bars must have sharp edges (less than 6 mm2 damaged surface per bar)

Quartz bars need to be very polished, parallel,

and rectangular with sharp edges.
Annalisa Cecchi, 26th Mar. 2006
4

How to polish?
• Ferrara is doing a polishing feasibility study on natural quartz samples.

Annalisa Cecchi, 26th Mar. 2006

5

Reflection coefficient
The fraction of the light transmitted may be written as a combination of reflection losses and bulk attenuation:

These two parameters are very important!!!!
Annalisa Cecchi, 26th Mar. 2006

L = length of fused silica bar b = width of the bar Λ = bulk attenuation length R = internal reflection coefficient N = number of bounces of the photons

6

Study on the number of reflections
MC simulation of 50 charged particles passing through the bar: 0.7<β<0.99; 250nm <  < 650nm

1) Random simulation over long the bar from every direction:

Mean = 49 L= 1.9 m

Mean = 125 L= 4.9 m

NRefPANDA < NRefBABAR
Annalisa Cecchi, 26th Mar. 2006
7

Study on number of reflections
2) Random simulation from the interaction point ~ (0,0,0):

Mean = 83 L = 1.9 m

Mean = 170 L = 4.9 m

This study confirms that reflectivity requirements on PANDA DIRC can be less stringent than BABAR DIRC, because the loss of Cerenkov light is totally less.

Annalisa Cecchi, 26th Mar. 2006

8

Calorimetric test
In BaBar, people measured the reflection coefficient at five different laser wavelengths. They compared the data with the scalar scattering theory (see plot) Study done for BaBar DIRC

The internal reflection coefficient was consistent with a surface finish between 5 and 8 Å (RMS) in the UV-visible light region.
Annalisa Cecchi, 26th Mar. 2006
9

Calorimetric method in Ferrara
Setup to measure the internal reflection coefficient in studies of wavelength dependence

Test of surface contamination

• Comparing the obtained values with the “Scalar theory”, we could have an idea of which surface contamination is allowed. • PANDA people from Ferrara (Vito Carassiti, Annalisa Cecchi, Diego Bettoni) plan to do two kind of calorimetric test, in collaboration with other experts (Vincenzo Guidi).

Annalisa Cecchi, 26th Mar. 2006

10

Calorimetric test [1]
• Quartz samples already existing (10 cm long).

• We will set up an optical system, to test the power of Ferrara polishing machine.
• We will do a test with the “calorimetric method”, to measure the reflectivity. • We will compare the values obtained with the scalar theory (explained before).

Annalisa Cecchi, 26th Mar. 2006

11

Calorimetric test [2]
• We will purchase two polished synthetic quartz samples (30 cm long) to test, from DISCO company – Munich. • We will do the calorimetric test, with both samples glued. • In this way we can also test the efficiency lost due to the glue.

• Both tests will be done in a optical lab in Ferrara
Annalisa Cecchi, 26th Mar. 2006
12

Conclusions
 With MC simulation we confirm that polishing requirements for Panda DIRC are less hard than BaBar.  Ferrara experts plan to set up an optical system to measure the reflectivity, with the calorimetric method  All these test will give us more informations on the feasibility to polish quartz bars.

Annalisa Cecchi, 26th Mar. 2006

13