Report on WP2/C1:
Report on the thermal gradient measurements with the payload
, M. Ciaccafava
, E. Majorana
, V. Moscatelli
, M. Perciballi
, P. Puppo
, F Ricci
Department of Physics, University of La Sapienza, Rome I
INFN Sezione di Roma, Rome, Italy
Coordinators: F. Ricci (Department of Physics, University of La Sapienza and INFN Sezione
di Roma, Rome, Italy)
The payload has been cooled using our cryogenic facility, the Vibration Free Cryostat (VFC)
d in a previous report of the ILIAS STREGA project. We recall here just its main
The cryostat is equipped with a Pulse Tube refrigerator SR052 made by the Sumitomo
cooler cold head is clamped to a platform con
nected to the cryostat by a
special silicon bellow designed to mechanically decouple it from the cryostat flange. The platform sits
on three vertical piezo
actuators located at 120
where signals can be applied. They allow actuating the
vertical position o
f the platform, holding the cold head, along frictionless guides. Springs, with cut
frequency around 5 Hz, soften the load on the piezo
A thermal shield made of
gold plated copper is hung to the cryostat flange by means of the steel 0.6
wires at 120
. It is thermally connected to the first stage
of the cryo
cooler, at ~ 40 K,
by means of
soft heat links. Then, an aluminium vacuum chamber is similarly hung in cascade to the copper shield
flange. The vacuum chamber is equipped with a 2 mm
indium gasket ensuring the vacuum tightness; it
can host the small scale cryogenic payload. Its aluminium top flange is thermally connected to the
e of the cryo
cooler by similar
heat links all made of ETP copper strips arranged on a ring.
strips are thermally treated before the assembly to reduce the stiffness and provide the main
mechanical decoupling between the inner vacuum chamber and the vibrating cold head.
The vibration attenuation is obtained by an active control system based on pie
zoelectric stacks mod.
90 by Physik Instrumente. We sense the displacement of the PT cold finger with respect to
stat flange by the Philtec fibre
is used for servicing to a
PC computer the access to ADC an
d DAC boards, located in the VME crate.
The computer runs applications that we developed specifically for data acquisition and control in
LabView environment. We are able to attenuate the mechanical excitation of the cryo
cooler up a
factor 220 on the fr
equency peak of the fundamental harmonics of the PT (~1 Hz).
First all we cooled the cryostat without the payload in it. In Figure 2 we show on the left the
temperature behaviour of the empty cryostat during the cooling phase, while on the right we show a
zoom of the final cooling phase for the innermost part of the cryostat. We notice that there is a
significant thermal input in the vacuum chamber so that the SR052A cryo
cooler is not able to bring
below 20 K the temperature of the aluminium chamber in a
time lower that 2.4 10
s. Moreover we
have to notice also that we are using the Sumitomo refrigerator in a configuration different far from the
optimized one. This implies that the cryo
cooler is less providing a lower refrigeration power.
Vibration Free cryostat cooling without payload.
After this first cooling the
payload was located in the inner vacuum chamber of the cryostat, tightened
by an indium gasket. In order to speed up the cooling phase, we kept in the chamber a residual pr
of helium gas exchange of 298 mbar at 300K. At low temperature we pumped out the residual gas and
we achieved a residual pressure of 10
mbar in the inner vacuum chamber.
Picture of the payload ready to be inserted in the cryostat.
Temperature vs. time of the various parts of the cryostat during the cool down.
In figure 3 we plot the temperatures of the cryostat elements during the cooling process. In particular
the temperature change of the first and sec
ond stage of the pulse tube,
the temperature evolution of the top and bottom of the intermediate thermal shield anchored
to the first stage of the refrigerator,
the temperature of the top and bottom of the inner vacuum chamber.
In figure 4 we show the temp
erature versus the time of the various elements of the payload. We notice
that the temperature gradient between the payload top and the aluminium flange of the inner vacuum
chamber is negligible.
Temperature vs. time of the various parts o
f the cryostat during the cool down.
Although the payload assembly implies additional sources of thermal input (as the for example the
thick copper wires of the actuator coils), thanks to a more accurate cryostat assembly we were able to
reduce the total
thermal input on the inner vacuum chamber. In a quasi
stationary conditions with a
residual pressure of 10
mbar in the inner vacuum chamber, we achieved a temperature value of the
top flange below 20 K.
In the figure 5 we show the temperature value of t
he payload top nearly coincident with both the mirror
temperature curve and that of the aluminium flange. We conclude that we have been able to optimize
the path for the transmission of the refrigeration power from the inner vacuum chamber to the mirror.
Temperature vs. time of the various parts of the payload in a quasi stationary condition at
The pressure in the inner vacuum chamber is 10
We notice also that still the temperature of reference mass is converging slowly toward the
value of the mirror. We measured a cooling time constant of 5
which it results to be compatible
with the thermal capacity of the mass and the thermal conductivity of the coil wires.
fact, in absence
of the helium gas exchange in the vacu
um chamber, it is easy to realize that the heat extraction from
the mass is trough the coil wires.
An other important aspect of the cooling process is the temperature fluctuation of the
mechanical elements cooled at low temperature. In particular we noti
ce in the plot of figure 5 that the
temperature of the second stage of the cold heat is fluctuation significantly. In order to clarify this
crucial aspect, we sampled the thermometer outputs at
ms, a time interval higher than the typical
e of our Si
Temperature vs. time of the second stage cold head of the pulse tube and of the mirror
As it can be seen in figure 6 on the secon
d stage cold head we measured
e of the order of 0.7 K with the typical periodicity of 1 Hz related to the main harmonic of the
gas standing wave of pulse tube refrigerator. Concerning the temperature fluctuation of the reference
mass we can set just an upper limit of 0.01 K. Similar re
sults is obtained for the mirror temperature
Finally we conclude by reporting the change in the dynamical behaviour of the payload at low
temperature. In figure 7 we show the mirror displacement spectra measured at room and low
th the fiber bundle. We noticed that some of the oscillation modes of the payload were
shifted in agreement with the change of the elastic properties of the suspension material.
Displacement spectrum of the mirror at hig
h and low temperature.
Room Temperature (T=300K, P = 1000 mbar)
Low Temperature (T=6K, P = 10
Mirror displacement spectrum (Fiber bundle readout V