Cryogenic plans for ET-pathfinder - PowerPoint Presentation

Slide1

Cryogenic plans for ET-pathfinder

H.J. Bulten, for the ET-pathfinder collaboration andin collaboration with the EMS-group, TU Twente

1

GWADW 2019, Elba (May 22)

ET pathfinder

ETpathfinder facility (Limburg) to study cryogenic optics and operations, under design

2

Cryogenic option for ET: silicon mirrors at 1.550 micrometerLots to be designed/developed (coatings, monolithic suspensions, mirrors, coolers)We want help develop techniques for ET and test cryogenic operation in an interferometerTest facility in Limburg, under design

Si mirrors (later monolithic suspension)We want to be able to measure at room temperature, 123 K, ~ 10 KWe aim for ASD(x) < 10-18 m/sqrt(Hz) from 10 Hz upwards.

Currently design is ongoing. ~15 m long sections can contain 2 FB-cavities with small optics or 1 cavity with large (ET-size) optics. Roof cleanroom is 8m high. At the start, we will employ cooling in 1 arm.

M.

Doets

37[m]

25[m]

Cleanroom 8.2[m] high

Towers max 6.5[m] high

Class:10 000 ?

Beam splitter tower

Suspended bench

End Mirror tower

Mirror tower

End Mirror tower

Mirror tower

crane with double hook

2.6 m

End mirrors.

Cryogenic (10K)

ET (pathfinder) cryogenic challenges

Arms may possibly be cooled to 80K with liquid N. Not needed with small optics.

Cool end mirrors in the arms to about 10 K

Without introducing vibrations (< 10

-18

m/sqrt(Hz))

Without freezing water on the monolithic suspension and mirror surface

With thermal shields with openings for laser beam and optical levers

Without use of MLI superinsulation in the primary vacuum (CH partial pressure should be minimal, pumping/baking should be possible)

Maintain good vacuum in arm and around the end mirrors

Be able to operate at room temperature and at cryogenic temperature (displacement control)

Control temperature-dependent changes in optical path length – control mirror temperature.

Beam splitter tower, warm

Injection tower, warm

End mirror towers

Start configuration: 2 FP cavities in 1 arm.

Primary vacuum: <10

-7 mbar required (residual gas noise). No MLI allowed.

Thermal shields

Double-walled with holes for pumping and viewports.

Should not vibrate too much (NN, scattered light)

Mass cryogenic payload ~ 70 kg (

Etpathfinder

), ~1000 kg (ET).

Mass inner cryogenic shield ~200 kg

Mass liquid Nitrogen shields ~ 300 kg.

Area shields ~ 5-10 m

2

so incident radiation is at kW level. Liquid N cooling required to reduce load on inner shield.

Liquid N cooling (~200W). Inside primary vacuum. Here: vessel connected via 192

Litze wires (60mm long, 8mm diameter; DT = 12 K

Vibration-free cooler needed for mirror and suspension

ET cooling (conceptual design 2011)

Vibration-free cooling links needed!5

End mirrors need to be cooled to 10 K (mirror coating thermal noise peak at 18 K)

Separate cooling for shields/beam pipe and marionetteMarionette cold link is suspended.Mirror cooled to 10 K via conduction (monolithic silicon wire suspension).Mirror heat input load ~ <100 mW (thermal radiation and laser power)

Thermal load cold shield (4K) ~1 WIntermediate shield ~ 50WCold shield (blue) needs to be vacuum-tight and need conductance to pump before cooling.Initial cooling via contact gas.

Vibration isolation of thermal link to marionette.

Heat link, complicated

50-75 MLI layers required!

Pulse-tube cooler (

Kagra)Gas injection at JT restriction of a pulse-tube cooler lead to vibrations

6

From

Ushiba(KAGRA), GW workshop Taiwan(2015)

Kagra

, CQG 31, 224001 (Chen et al.)

Projected vibrations due to

pulsetube

cooler KAGRA (red: off, black: on, blue: ground)

Pulse-tube cryocooler: typically pressure spikes (few bar ripple on high pressure before JT restriction, several N force, repetition rate ~1.5Hz)

Liquid He: complicated and vibrations from boiling

Preferred option: Sorption coolingCollaboration with EMS-Twente (group

ter Brake) to develop a sorption-cooling strategy for ET7

Compressor: (reversible) physisorption of gas on Carbon. Pumped by heating. No moving parts, no degradations.Sorption cooling: pressure ripple several orders of magnitude below that of a pulse-tube cooler. Accelerations due to gas in feedline are at least 10,000 times smaller than in commercial pulse tube cooler.

Technique developed in Twente and demonstrated in the METIS instrument of the ELT telescope.

Wu et al., Physics Procedia 67 ( 2015 ) 411 – 416, Cryogenics 84 (2017)

Tzabar

&

ter

Brake, Adsorption 24 (2018) p325-332

Presented by

ter

Brake, ET-meeting Hannover, Oct 23, 2013

compressor

Sorption coolers

Technique, demonstrated in satellite missions (DARWIN and ATHENA) and ground-based( METIS/ELT)8

Ter Brake, TU Twente

METIS instrument: 3 heat inputs: LM-band 40K, 1.4W, N-band detector 25K, 1.1W, N-band imager 8K, 0.4W.4 sorption compressors used (1 Neon, 2 hydrogen, 1 He) for optimal performance.Heat exchangers at different temperatures used.The He sorption cooler efficiency is strongly

dependend on the sink temperature: for ET-pathfinder we consider using a pulse-tube cooler to precool the gas for the He sorption cooler.

cold tip temperature stabilized

by PID controller and heater

DARWIN cooler: extreme temperature stability required

Sorption coolers

9

Vibration measurements for the DARWIN cooler, slide from ter Brake, ET-meeting Hannover, Oct. 23, 2013Exerted forces by the gas are negligible.

Cryogenic shields around mirror

(Preliminary) design cryogenic shields (M. Doets)

10

300-K shield, stabilize filter temperature

Floating reflective shields to reduce radiation load on first shield.

Inner thermal shield, < 40 K.

Jellyfish wires (ultra-pure aluminum) to cool marionette and limit vibrations

Outer (80 K) shields (liquid nitrogen)

Shields should reduce thermal radiation, but have holes to be pumped out. Holes through all 6 shields for optical levers and for the laser beam (shielded with pipes). Scattered light and thermal radiation must be absorbed somewhere.

Vibration-free cold finger for cooling mirror

(JT restriction sorption cooler)

Pipes with baffles to limit view factor of mirror- tower

Shield modeling, vacuum

Raytracing code to calculate vacuum performance, scattered-light absorption, and temperature gradients11

For the design of the sorption cooler we need to be able to calculate the conductive and radiative heat loads in the system. Vacuum conductance and scattered-light contribution of shields is modeled too.Water sticking time to SS/Al halves for every ~5.5 deg. C heating.In order to avoid freezing water on the mirror/monolithic suspension, we need baking OR cooling down the liquid Nitrogen shield while heating the mirror and inner thermal shield.

Water monolayer, baking required.

Shield modeling, thermal equilibrium

Shields all low-emissivity (0.1) except for baffles. Assumed conical (45 deg) baffles with 15 mm diameter at top opening.Detailed studies (diffuse/specular scattering, size holes, baffles, etc) starting now.12

292K

300 K

285 K

85 K, 180W

117 K, 12.6W

120 K, 21+14 W

132 K, 7 W

116 K (inside)

30.8K, 0.2W

30K, 3.1 W

30.6K, 0.24W

30.6K, 1.0W

Coldhead

10K, 50

mW

13K, 10mW

269K (inner)

Liquid Nitrogen coolingLiquid Nitrogen cooling

13

Shields

Total volume 40 liter

Outlet bellows D65[mm]

Inlet bellow D10[mm]

Slotted inlet pipes to avoid bubbles in inlet.

Wide vessel to reduce vibrations from boiling.

Vessel decoupled from 80-K shield with 192

litze

braids (8mm diameter, 60mm length – may be replaced).

Outer thermal shield supports the inner thermal shield. Flexible joints to accommodate shrinkage when cooling down.

SS and Kapton flexures to accommodate shrinkage when cooling

Link between cryogenic coldhead and mirror suspension

We intend to produce a jellyfish connection and measure the transfer function.14Jellyfish connections: introduce minimal vibrations

Ultrapure Al : up to 15000 W/m/K at 10 K (bulk)Monolithic silicon suspension: flexible, >2 kW/m/K

Purple wires to marionette, blue to reaction chain

To

coldhead

. Slot for suspension wire

Indium foils

SummaryDesign studies for ET-pathfinder are in progress

15Baseline design for mechanics, vacuum, cooling is well underwayWe aim for liquid-N cooling to reduce thermal radiation and to define a point where the water vapor might be frozen – Else we need to bake for a week after every vacuum breach.

We aim for sorption cooling for the cryogenic partWell-tested technique, both in space and ground-based astrophysical detectorsCoolants and design sorption cooler depend on requirements, which depend on the thermal shields (study ongoing)

We will fix the thermal shields to the ground. Baffles on the reaction chain of the mirror should define the beam profile (shields should have zero scattered light contribution)We intend to both build 1 sorption cooler and test vibrations of the cold mass and build a jellyfish transition to measure mechanical transfer of vibrations; this depends on available time and extra funding.

22-5-2019M Doets/ Nikhef

1610k shields

80k shields

floating shields

Support for upper vessel

Option: Shields extended

80k shield cooled with Ln2

Option: Shields extended

80k shield cooled with Ln2

Optical baffles

300k shield

For ET: In the arms:

cryolinks

for better vacuum (80K shields) and to limit thermal radiation view factor.

We left space to put that in for ET-pathfinder, if needed (for the large optics)

FEM modeling needed to design shields, optimize hole sizes and

emissivities

, calculate scattered light and configure the sorption cooler scheme

Einstein telescope CDR 2011 - noise limits

Thermal noise limits the sensitivity in mid-frequency range.

Low frequencies:Quantum noise: heavy mirrors, 18 kW beam power, squeezing. Seismic noise: underground. Thermal noise: cryogenic.

High-frequencies: shot noise limited. Room temperature, 3 MW beam power in arms.

Shot noise

Radiation pressure

17

Thermal noise dominant 10-200 Hz.