Transcript magic ii - CERN Indico

The Reflective Surface
of the MAGIC Telescope
Michele Doro
on behalf of the MAGIC Collaboration
University of Padova & INFN
[email protected]
6th RICH - Trieste, Italy
15-20 October, 2007
Overview

Part I: The MAGIC detector


Part II: The Reflective Surface






Overview on the IACT technique
Demands
Mirror
Tests and measurement
MAGIC I problem
MAGIC II upgrade
Overview on future
15-20 Oct 2007
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
2
PART 1
The MAGIC Telescope
The MAGIC Telescope




QuickTi me™ and a
TIFF ( Uncompressed) decompressor
are needed to see thi s pi ctur e.
Collaboration of 22 institutes (mostly European) ~150
physicists
Installation completed 2003,
fully-operating since fall 2004
~50 publications on journals
Currently on III-year cycle of
scientific-observations
15-20 Oct 2007
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
4
The IACT technique
QuickTi me™ and a
TIFF ( Uncompressed) decompressor
are needed to see thi s pi ctur e.
Physics of the atmospheric showers:

Cosmic rays (protons, heavier Z, electrons,
photons) hit the upper atmosphere

Interactions create cascade of billions of
particles:



Charged particles in turn emit Cherenkov
light:




Electromagnetic shower (e+,e-,)
Hadronic shower (, , e+,e-,)
Blueish flash
~2ns duration
~1º aperture
Cherenkov cone reaches the ground


Circle of ~120m radius
Effective telescope area ~ 104-5 m2
15-20 Oct 2007
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
5
Imaging Technique

Light is reflected on a multi-pixel
camera (576 total)



396 central PMT 1 inch
180 outer PMT 1.5 inch
Image is ellipsoid



QuickTi me™ and a
TIFF ( Uncompressed) decompressor
are needed to see thi s pi ctur e.
Pointing to the centre for gammas
Randomly distributed for hadrons
Study of the image gives information
on primary particle
The reflective surface must ensure
a PSF possibly smaller than the
pixel size
15-20 Oct 2007
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
6
Physics program
Pulsars
QuickTi me™ and a
TIFF ( Uncompressed) decompressor
are needed to see thi s pi ctur e.
AGNs
GRBs
SNRs
?
Origin of
Cosmic Rays
Cold Dark
Matter
15-20 Oct 2007
Quantum Gravity effects
cosmological
-Ray Horizon
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
7
MAGIC II
QuickTi me™ and a
TIFF ( Uncompressed) decompressor
are needed to see thi s pi ctur e.

Currently a second telescope is
being built




Stereoscopic MAGIC I + II will
have increased performance:




15-20 Oct 2007
Structurally a clone of MAGIC I
Each system adopted new enhanced
solutions
Better telescope than MAGIC I
Increased angular resolution
Increased energy resolution
Increased flux sensitivity
Inauguration 21/09/2008
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
8
Many challenging solutions

Reflector and mirrors:





Drive



Lacquer-coated enhances photon
conversion of PMT
Operation with moonlight
Signal transmission


15-20 Oct 2007
Faster repositioning ever achieved
Camera


World largest dish diameter 17m
Light undercarriage made of CFRP
All aluminium mirrors with
sandwich structure and diamondmilled surface
Active mirror control
Ultra-fast acquisition (2GhZ)
Optical transmission instead of
coaxial
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
9
PART 2
MAGIC reflective surface
Brief History
QuickTi me™ and a
TIFF ( Uncompressed) decompressor
are needed to see thi s pi ctur e.
2001-2004 MAGIC I
mirrors are designed,
tested and installed




732 INFN mirrors (76%)
224 MPI mirrors (24%)
2005-06 MAGIC I
Upgrade of the design
Substitution of damaged
mirrors

MAGIC I
MAGIC II

2006- MAGIC II mirrors
1m2 Aluminium mirrors
(INFN)
2
1m Glass mirrors (INAF)

15-20 Oct 2007
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
11
Mirror Technical Demands
Lightweight

Telescope must rotate fast and then mirrors
need to be as light as possible
Mirror Shape

Mirrors profile is spherical
Each mirror has different radius of curvature
because reflector profile is parabolic (f=17m)

Rigidity


Insulation



Mounting


Optical
quality
15-20 Oct 2007


QuickTi me™ and a
TIFF ( Uncompressed) decompressor
are needed to see thi s pi ctur e.
Avoid oscillations due to wind
Avoid bending during tracking
Sometimes strong rains and snows
Also high humidity
Strong UV light
Coupling with actuators of Active Mirror Control
Easy mounting and substitution
Maximize reflectivity
Minimize reflected spot size
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
12
The Reflector

Parabolic profile




preserve temporal structure of the shower
slightly increased aberrations
Huge dimension demands tessellation of
the surface
Radius of curvature changes according to
position


So-called average radius used (mean of
paraboloid principal radii )
34 to 36.5 meters radius range
QuickTime™ and a
TIFF (LZW) decompressor
are needed to see this picture.
15-20 Oct 2007
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
13
Shape
QuickTi me™ and a
TIFF ( Uncompressed) decompressor
are needed to see thi s pi ctur e.

QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
Large reflector area (~234m2)
requires to tessellate the
surface

Geometry of the mirror tile



Size



15-20 Oct 2007
Past used solutions: round,
hexagonal
Solution: MAGIC has square
mirrors to minimize empty regions
and easier production
Construction reasons
Aberrations
Solution: MAGIC I has 0.5m side,
MAGIC II has 1m side (2x) mirrors
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
14
Materials
QuickTi me™ and a
TIFF ( Uncompressed) decompressor
are needed to see thi s pi ctur e.
Established experience with glass
mirrors (astronomy) for other IACTs


AlMgSi0.5
plate


1-2cm thick glass layer
Aluminized for reflectivity
Protection
some drawbacks

Hexcell

Usual large weight
Difficulty of producing different focal
lengths
Al-box

Mounting
and laser
15-20 Oct 2007
Idea of the full-aluminium sandwich
Al-alloy surface
Al-box
Hexcell honeycomb structure
Diamond-milling of the surface
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
15
Assembly

1

QuickTi me™ and a
TIFF ( Uncompressed) decompressor
are needed to see thi s pi ctur e.
The sandwich is assembled with the use of the
aeronautic glue 3M™ AF163-2K
The sandwich in then inserted between two verystiff aluminum-moulds



2
Plane for MAGIC I
Already curved for MAGIC II
and everything is put into a plastic vacuum-bag
Autoclave curing


5 bar pressure
120º temperature
MAGIC II
Result is
the
3
35m
QuickTi me™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
raw-blank
15-20 Oct 2007
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
16
Diamond-milling

3

QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
A diamond mills the surface
To give spherical shape, rotation on two
axis



Mirror rotates around optical axis
Machine axis rotates tilted and diamong at
distance d
Adjustable R curvature

In the MAGIC reflector around 20 different bins
of radius of curvatures are needed
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
15-20 Oct 2007
QuickTi me™ and a
TIFF ( Uncompressed) decompressor
are needed to see thi s pi ctur e.
The mirror gets the
reflective properties
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
17
Coating
QuickTi me™ and a
TIFF ( Uncompressed) decompressor
are needed to see thi s pi ctur e.


The aluminium must be protected against
environment
Solutions

QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.




QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
Solution adopted: quartz because of
costs and transparency in 300-700nm
The width must be optimized for positive
interference in the wavelength where
Cherenkov light is peaked (blue)


15-20 Oct 2007
Diamond chemical vacuum deposition
(CVD)
Al2O3 anodization
SiO2 vacuum deposition
Width ~ 100 nm
Measurement of the roughness gives 4nm
on average
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
18
Testing the quality - Reflectivity

The reflectivity was measured in the
Cherenkov range (200-800nm) using:





Reflectivity is correctly peaked at 400nm (close
to peak of the Cherenkov spectrum)
Average reflectivity around 80-85%
MAGIC II:

15-20 Oct 2007
Perkin-Elmer device (mirror must be cut and
put inside the machine)
Spectro-photometer (local measure on
3x3cm2 of surface)
MAGIC I:
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.

QuickTi me™ and a
TIFF ( Uncompressed) decompressor
are needed to see thi s pi ctur e.
Mantained the same qualities
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
19
Testing the quality - Spot Size

QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.

Mirror is put at twice the focal length
(~35m) and illuminated with point-like
bright source
Analysis of the CCD image:



LED
MAGIC I mirror

d90 ~ 10mm at the
camera distance
So-called “d90”= diameter containing
90% of the focused light
d90 = spot size
It is not a measure of reflectivity due to
difficulties in estimating scattered light
LED
MAGIC II mirrors



15-20 Oct 2007
QuickTi me™ and a
TIFF ( Uncompressed) decompressor
are needed to see thi s pi ctur e.
d90 ~ 5mm
Better of factor 2
BEST RESULT!
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
20
QuickTi me™ and a
TIFF ( Uncompressed) decompressor
are needed to see thi s pi ctur e.
Panel & AMC

Major difference between MI and MII

mirrors are grouped into panels of 4 (3 in some
cases)





Panel is also Al-sandwich (20kg)
Inter-alignment and fixing
Single mirror host AMC for MII
The back of the panel hosts the actuators for the
Active Mirror Control
AMC moves panel to
re-adjust the
focussing to correct
small bending during
the tracking

Use of laser
15-20 Oct 2007
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
21
MAGIC I experience

Main problem with mirrors installed in
MAGIC I after two years from
installation





Humidity was entering from edge of
top plate of the mirrors
Condensation into water
Ice formation and bubbles
Due to strong rigidity, deformation is
local and mirror maintains reflective
area
Substituted around 100 mirrors
(~10%)



15-20 Oct 2007
QuickTi me™ and a
TIFF ( Uncompressed) decompressor
are needed to see thi s pi ctur e.
Re-designed mirrors for insulation
Test for large mirrors
Now problem seems solved
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
22
Reflector performance monitor

Reflector performance can be monitored



QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
QuickTi me™ and a
TIFF ( Uncompressed) decompressor
are needed to see thi s pi ctur e.
Reflector PSF, now stable <11mm
Single mirror abs.reflectivity
SBIG CCD at the centre of the reflector, observe a
star and the camera at the same time


All mirror focussed:PSF
One mirror focussed and others defocussed: single
mirror reflectivity
M. Garczarczyk
Phd Thesis 2007
~11mm
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
SPOT
Inner PMT
15-20 Oct 2007
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
23
Upgrade on MAGIC I

Improved design



15-20 Oct 2007
QuickTi me™ and a
TIFF ( Uncompressed) decompressor
are needed to see thi s pi ctur e.
Thinner skin and pre-shaped box: the
mould is not spherical and the rawblank comes out already with ~35m
curvature
Larger top-plate and gluing of the edge
with 3M™ DP190
External heater to avoid coupling
between plastic and aluminium
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
24
MAGIC II Al-mirrors

Basically an extension of MAGIC
I-upgraded mirror:





QuickTi me™ and a
TIFF ( Uncompressed) decompressor
are needed to see thi s pi ctur e.
Increased honeycomb width
resulted in increased rigidity:
Larger upper plate with respect
to box
Use of aeronautic glue DP190
for insulation
No internal heater
Curvable box

Best spot size due to more
accurate diamond milling of the
surface
MAGIC I
MAGIC II
15-20 Oct 2007
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
25
MAGIC II

QuickTi me™ and a
TIFF ( Uncompressed) decompressor
are needed to see thi s pi ctur e.
MAGIC II will have



Upgrade to 1m2 facet





144 m2 of INFN-Padova mirrors
104 m2 of INAF-Brera mirrors
Technologically achievable
Less number of items
Decreased weight and direct
coupling to active mirror control
No-need of inter-alignment
Drawback:

Less approximating the parabola: increased
aberrations, nevertheless the coma aberration
dominates for tilted incidence
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
15-20 Oct 2007
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
26
MAGIC II Glass mirrors
 A thin glass sheet (1-2 mm) is elastically
deformed so to retain the shape imparted by a
mould having convex profile. If the radius of
curvature is large, the sheet can be pressed
against the mould using the vacuum suction.
 On the deformed glass sheet (under vacuum
force) is glued an honeycomb structure that
provide the structural rigidity.
 Then a second glass sheet is glued on the top to
create a sandwich.
 After releasing the vacuum, on the concave side
is deposited a reflecting design for their mirrors
coating (Aluminum) and a thin protective coating
(Quartz)
15-20 Oct 2007
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
27
Summary and outlook





15-20 Oct 2007
MAGIC II mirrors production is
already on the production-line
Technique gave excellent results
in term of light concentration
Insulating problems seem solved
Price is decreased wrt to MAGIC
I, nevertheless is still main
drawback: 2.8k€/m2 can be a
problem for third generation
IACTs
Scale production can decrease
costs or find other techniques
(glass)
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
28
Back-up slides
Results
QuickTi me™ and a
TIFF ( Uncompressed) decompressor
are needed to see thi s pi ctur e.
http://tevcat.uchicago.edu/


Around 50 publications on journals

~21 VHE source observed (6 MAGIC discoveries!)

7 new analysis techniques

23 technical papers
Observed sources:

12 = extragalactic

9 = galactic
15-20 Oct 2007
M.Doro - The Reflective Surface of the MAGIC Telescope - RICH 2007
30