Cryogenics in space: a review of the missions and of the technologies

Cryogenics 40 (2000) 797±819

www.elsevier.com/locate/cryogenics

Cryogenics in space: a review of the missions and of the technologies B. Collaudin a,1, N. Rando b,* b

a Mechanical System Department, European Space Agency, Noordwijk, Netherlands Astrophysics Division, Space Science Department, European Space Agency, ESTEC, Keplerlaan 1, 2200 Noordwijk, Netherlands

Received 13 November 2000; accepted 12 February 2001

Abstract Cryogenics has made a remarkable amount of progress over the last 15 years. The increased reliability and simplicity of operations of cryogenic equipment have allowed to install and to successfully operate them onboard spacecrafts. At the same time, the improved performance of cryogenic devices, such as sensors and cold electronics, has drastically enlarged their utilisation, creating new perspectives for space-based applications. In this paper we provide an up-to-date review of the non-military space missions making use of cryogenic instrumentation and a summary of the present and envisaged applications of cryogenic equipment in space. The impact of cryogenics on the spacecraft system design and the main technical solutions presently adopted are also discussed. Finally, this paper provides an outlook on near- and mid-term future activities involving cryogenics in space and on the related technology development, with particular emphasis on the work carried out by the European Space Agency. Ó 2001 Elsevier Science Ltd. All rights reserved. Keywords: Cryogenics; Detectors; Spacecraft engineering

1. Introduction Since the ®rst liquefaction of 4He and the discovery of superconductivity by H. Kamerlingh-Onnes (1908 and 1911), cryogenics and its applications have gone a long way. The continuous improvement of cryogenic equipment [1] has made it easier and easier to achieve temperatures well below the liquefaction point of nitrogen (77 K) either by means of cryogens (liquid gases such as Xe, H2 ; O2 ; N2 ; 4 He and 3 He) or by means of mechanical cooler (Fig. 1). Cryogenic devices, such as sensors and cold electronics, have taken advantage of the progress made in materials science, o€ering a reliable and e€ective solution to otherwise unsolvable problems. In the last 15 years, several spacecrafts were involved with cryogenic equipment, mostly in relation to astrophysics missions, targetting the electromagnetic radiation emitted by celestial objects over a wavelength range otherwise dicult to work with from ground *

Corresponding author. Tel.: +31-71-5653638; fax: +31-715654690. E-mail address: [email protected] (N. Rando). 1 Present address: Alcatel Space Industries, BP 99, 06156 Cannes la Bocca cedex, France.

(Fig. 2). Among such missions we should mention IRAS (Infrared Astronomical Satellite, launched in 1983), COBE (Cosmic Background Explorer, launched in 1989) and ISO (Infrared Space Observatory, launched in 1995) [2]. The Japanese satellite Astro-E (X-ray Observatory, launched in 2000) was designed to carry a spectrometer (XRS) built by NASA and operating at a temperature of 65 mK [4]. At this moment, several space missions involving cryogenic applications are in advanced development in Europe as well as in Japan and US. The ESA space science missions Planck (dedicated to the mapping of the cosmic background radiation) and Herschel (formerly known as FIRST ± Far Infrared and Submillimetre Telescope) are clear examples [3]. Both satellites will carry scienti®c payloads working at temperatures of 0.1 and 0.3 K, respectively. SIRTF (Space Infrared Telescope Facility, launch in May 2002), one of the four large NASA's observatories, is based on a 4He cryostat and it will perform photometry, spectroscopy and imaging in the 3±180 lm spectral range [5]. On a longer time-scale, other space missions with cryogenic equipment are planned. On the ESA side we have XEUS (a post-XMM mission, planning to use cryogenic detectors at temperatures of about 100 mK)

0011-2275/01/$ - see front matter Ó 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 0 1 1 - 2 2 7 5 ( 0 1 ) 0 0 0 3 5 - 2

798

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819

Fig. 1. Historical development of refrigeration techniques. The rapid progress of the last 30 years is clearly visible (reproduced from [1] with permission from the editor).

[6] and DARWIN, a mid-infrared (k ˆ 5 to 30 lm) space observatory dedicated to the search of planets. On the US side, we should mention Constellation-X (the high throughput X-ray spectroscopy mission) also envisaging micro-calorimeters working at T < 100 mK [7]. In the higher temperature range, between 100 and 10 K, many missions are already operational or in development. They include military reconnaissance satellites (such as Helios), earth observation satellites (SPOT) and meteorological spacecrafts (MSG, Meteosat second generation), with IR detectors operating at about 85 K [8].

Considerable progress has been achieved in the area of low temperature superconductors (LTSs), with the fabrication of thin ®lm-based micro-devices (Nb, NbN, Ta), showing highly uniform properties and excellent resistance to thermal cycles. Such devices include single Josephson junctions (or superconducting tunnel junctions, STJs), superconducting quantum interference devices (SQUIDs) and simpler components, such as resonating cavities and low loss RF ®lters. Even if at a lower pace than that initially predicted, also high temperature superconductors have moved from a pioneering phase to more mature technological applications. In particular, the advent of radiofrequency (RF) devices based on HTS has opened up new perspectives for spaceborne cryogenic instrumentation, with the potential for more energy ecient and better performing telecommunication platforms. Smaller-scale developments are now being considered for the International Space Station (ISS), in the form of technology demonstrators for telecommunication applications [9]. It cannot be denied that developing, installing and operating cryogenic instrumentation in space adds a degree of complexity, risks and associated costs. Any application must then be justi®ed on the basis of its speci®c return. In the case of scienti®c missions, the cryogenic detectors and related payloads are the only candidates for the accomplishment of the task proposed, o€ering unmatched performance and unique advantages. In the case of other applications, such as telecommunications, the advantages o€ered by superconducting devices must be evaluated against their development and operating costs and compared with alternative technologies.

Fig. 2. Atmospheric transmission as a function of photon wavelength at sea level. At several wavelengths the atmosphere attenuates strongly the propagation of electromagnetic waves (e.g. in the range 30±500 lm). At high altitudes (e.g. onboard aircraft and balloons) the situation improves, but a residual absorption of 1% is responsible for the emission of a background which is intense enough to saturate sensitive low background IR cameras.

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819

2. Applications of cryogenics in space In this section we provide an overview of the applications involving cryogenics in space, including scienti®c instrumentation, telecommunications and large-scale superconductivity. 2.1. Cryogenic detectors for space science Cryogenic photon detectors o€er two main advantages over conventional sensors: 1. The much higher sensitivity (expressed by the noise equivalent power, i.e. the amount of incident power required to achieve a signal-to-noise ratio equal to unity). 2. The better energy resolution (expressed in terms of resolving power, i.e. the ratio E=DE ˆ k=Dk, with DE representing the full width at half maximum of the detector response to a monochromatic excitation of energy E). Cryogenic detectors have driven the utilisation of cryogenics in space, determining the requirements in terms of operating temperature, temperature stability and architecture of the payload system. This trend is now well established across the electromagnetic spectrum; Fig. 3 provides a summary of the di€erent detectors, including operating photon energy and temperature range. Table 1 provides an overview of other characteristics of the detectors, including typical power dissipations, array size and operating temperature. Applications involving the lower energy end of the electromagnetic spectrum (i.e. sub-mm wave and IR) are the ones which bene®t most from the utilisation of cryogenic detectors. Semiconductor bolometers have provided the ®rst answer to the needs of astronomers in this wavelength range, with operating temperatures ranging from just above 50 (NIR) to 0.1 K (sub-mm) and with NEP reaching values below 10 17 W/Hz1=2 .

799

A bolometer is a detector that works by recording the temperature rise due to the absorption of radiation via a resistance thermometer. Dramatic developments have recently taken place in the infrared detector technology, mostly driven by the vast investments made by the US Department of Defence throughout the 1980s. Such developments embrace a very large spectrum range, from the near-infrared (NIR) (k ˆ 1 lm) to the far-infrared (FIR) (k ˆ 200 lm), and have focused on lowbackground, high sensitivity and large format arrays. IRAS used a total of 62 detector elements, while since 1995 large format arrays for IR astronomy are available with a total number of pixels in excess of 106 [10]. Astronomical observations in the FIR investigate objects that are colder than those observed in the visible or in the NIR as blackbody radiation in the 30±300 lm wavelength range (emitted by bodies at temperatures ranging from 100 to 10 K). An example of such cold objects is the interstellar dust in our galaxy (at 20±30 K), as detected by IRAS in 1983, which con®rmed the existence of interstellar dust and detected its thermal emission. FIR radiation is totally blocked by the atmosphere, thus requiring observations from stratospheric balloons (40±50 km altitude) or, better, from space. Photo-conductors represent the main detection technique use throughout the IR range. At low temperatures and low photon ¯uxes, the conductivity of these semiconducting materials is in¯uenced by the absorbed IR photons, which can ionise impurities and free charge carriers. The ionisation energy of the impurities sets the cut-o€ wavelength of these detectors, ranging from 200 lm (stressed lattice Ge:Ga) down to 40 lm (blocked impurity band Si:Sb or Ge:Ga). Such photo-conductors are typically operated at T < 3 K. In the case of ISOPHOT, a broad band photometer ¯own onboard ISO, Ge:Ga detectors were combined with low noise CMOS integrating pre-ampli®ers and multiplexers operating at 2 K to achieve an NEP of order 1=2 10 18 W=Hz [11].

Fig. 3. Overview of photon detectors and related operating temperatures. It is worth mentioning the extended sensitivity region of the cryogenic detectors such as STJ and bolometers.

800

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819

Table 1 Main characteristics of photon detectors and SQUIDsa Detector type

Ge crystal CCD STJs l-Calorimeters TESs Photo-conductors ± NIR Photo-conductors ± MIR Photo-conductors ± FIR Sub-mm bolometers SQUIDs (LTS) a

Temperature range (K)

Dissipation range (W)

Detector size (pixel and array)

Utilisation

Min

Max

Min

Max

Pixel (lm)

Array …n  n†

Wavelength

50 150±200 0.01 0.05 0.05 30 2 1 0.1 1

100 300 1 0.3 0.3 100 20 2 0.3 4

0 0.1 10 9 10 12 10 11 0.01 0.01 0.001 10 9 10 12

0 20 10 6 10 11 10 9 0.02 0.02 0.003 10 8 10 11

10000 10±30 20±50 100 100 30±50 50±100 50±100 100±500 na

<10 106 <103 <100 <100 106 <104 <103 <102 na

Gamma X-ray/Vis. X-ray±UV±Vis.±NIR X-ray X-ray±UV±Vis.±NIR NIR MIR FIR Sub-mm Read-out/accelerometer

na: Not approved.

Bolometers have also been used to detect sub-mm photons. Neutron-transmutation-doped (NTD) Ge detectors are well established and operate at temperatures between 300 and 100 mK, with NEP of order 1=2 10 17 W=Hz . Such devices will be used onboard the ESA mission Planck (see Section 2.2). In the sub-mm wavelength range heterodyne receivers provide very high sensitivities up to frequencies as high as 500 GHz. Several laboratories have shown that receivers based on superconductor±isolator±superconductor (SIS) devices (such as STJs) o€er better performance than the conventional Schottky diodebased systems [12]. Nb-based junctions have been used for such applications in the 300±500 GHz range, showing noise temperatures ®ve times lower than the corresponding values of Schottky devices. Operating temperatures are of the order 2 K. At frequencies m > 500 GHz the so-called hot-electron bolometers (HEBs) compete with SIS and Schottky diodes for the next generation of heterodyne receivers (e.g. onboard the ESA missions Planck and Herschel). In such devices, the incoming radiation excites the electron population (either in a semiconductor or in a superconductor absorber) without heating up the corresponding lattice. The excited population of charge carriers determines changes in the resistance of the device, according to a non-linear behaviour, which can be exploited for mixing the signal voltage with the local oscillator voltage. Operating temperatures range from 70 (2DEG, InSb HEBs) to 0.3 K (NIS HEBs). In the NIR (at wavelengths between 1 and 5 lm) other photo-conductors are used, mainly PtSi, HgCdTe and InSb. Over the last decade, the introduction of two-dimensional InSb arrays has drastically changed the ®eld of IR astronomy, with 1k1k pixel array, based on the hybrid technology (e.g. InSb detector array bumped to a silicon source follower read-out in NMOS, PMOS or CMOS technology). These detectors

have operating temperatures ranging between 77 and 35 K and have already been used onboard satellites such as the Hubble Space Telescope (NICMOS camera) [13]. A new generation of photon detectors is represented by STJs [14] and transition edge detectors (TESs) [15], both photon counting in the visible and NIR, with intrinsic spectroscopic capability. STJs have operating temperatures ranging between 0.5 and 0.1 K depending on the superconductors used (typically Nb, Al, Ta), responsivities of order 104 e-/eV, resolving power of order 10 at k ˆ 500 nm and maximum count rate of order 104 event/s. TESs operate at about 0.1 K, have also very conspicuous responsivities, comparable energy resolution and a max count rate of order 103 event/s. Such a combination of performance is particularly attractive to modern astronomy, opening the way to new research activities [16]. Both STJs and TESs can operate over a large photon energy range with very interesting performance in the UV and X-ray. Both technologies o€er distinct advantages over the traditional UV detectors, based on a photo-cathode, micro-channel plates and related read-out system (such as multi-anode microchannel arrays and intensi®ed CCDs). The key bene®ts are the much higher detection eciency (close to 100%), the photon counting and intrinsic spectroscopic capability and the good imaging resolution (with individual pixels of order 20 lm). In the case of STJs an energy resolution of 15 eV at 6 keV has been demonstrated, while TESs have achieved even better performance (a few eV at 6 keV) [17]. On this basis, future space observatories are likely to make use of cryogenic detectors in narrow ®eld instruments dedicated to nondispersive spectroscopy. Fundamental physics and planetary sciences can also bene®t from the utilisation of cryogenic detectors. An example is represented by SQUID-based gravity gradiometers to be used for low altitude earth and planetary

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819

missions. In addition to the mapping of the intensity of the gravitational forces, these sensors can be used to verify the so-called equivalence principle, which postulates the coincidence of gravitational and inertial mass. Such an issue is being addressed in the feasibility study of di€erent space missions such as STEP (ESA) and LISA (NASA). SQUIDs-based accelerometers are the only ones capable of achieving the required accuracy; to date SQUID devices based on LTSs are favoured with operating temperature around 4 K. SQUIDs based on high temperature superconductors (HTSs) are also investigated in view of their capability to operate at about 77 K [18]. 2.2. Scienti®c missions: a review Scienti®c missions dominate the present scenario of cryogenics applications in space due to the advantages o€ered by cryogenic detectors over conventional sensors. In this paragraph we provide a review of the scienti®c missions. The review is organised in chronological order, starting from IRAS, the ®rst `cryogenic mission', which ¯ew in 1983. Mission in operations (or post-operations), missions presently under development and missions under study are grouped in di€erent sections. Tables 2 and 3 provide a summary of all nonmilitary space missions that involve cryogenics. 2.2.1. Missions in operations/post-operations Infrared Astronomy Satellite (IRAS) is the ®rst scienti®c satellite based on cryogenic instrumentation. It was launched in January 1983 as a joint project sponsored by the US, the UK and the Netherlands. Its mission was to map the entire sky at IR wavelengths from 8 to 120 lm. The satellite was equipped with a 0.6 m telescope cooled with liquid He to about 4 K. The focal plane assembly was located at the Cassegrain focus at about 3 K. It contained the survey detectors (based on 62 photo-conductive elements made from four di€erent materials), a low resolution spectrometer and a chopped photometric channel [19]. German Infrared Laboratory (GIRL) was a project for atmospheric and astronomical observations promoted by a consortium of German industries and universities aiming at the construction of a 40 cm liquid He-cooled IR telescope for repeated ¯ights on Spacelab. The study started in 1978, but it was cancelled in 1985 due to the high costs of the Spacelab ¯ights. The experience gained with GIRL proved very useful to the ISO mission. Cosmic Background Explorer (COBE) was developed by NASA's Goddard Space Flight Centre to measure the cosmic background radiation. The satellite was launched in November 1989 and operated for about 10 months in survey mode. It carried three instruments: an FIR absolute spectrometer (FIRAS), a di€erential microwave radiometer (DMR) and the di€use IR

801

background experiment (DIRBE) at wavelengths between 1.25 and 240 lm. FIRAS and DIRBE operated at 1.6 K cooled by a 650 l super¯uid helium cryostat [20]. Infrared Space Observatory (ISO) was developed by ESA and operated at wavelengths from 2.5 to 240 lm between November 1995 and May 1998 in a highly elliptical orbit. The satellite (Fig. 4) is based on a cryostat containing about 2200 l of super¯uid helium and on a 0.6 m diameter telescope, feeding four instruments (an infrared camera, a photometer and two spectrometers working in di€erent wavelength ranges). The instruments made use of di€erent photo-conductors based on InSb, Si and Ge and operating between 1.8 and 10 K [21]. Midcourse Space Experiment (MSX) is a Ballistic Missile Defence Organisation (BMDO) project also open to scienti®c application in the form of a co-operation agreement. MSX has been launched in 1996, with an IR telescope onboard, cooled by a solid hydrogen cryostat to about 8±9 K for a period of about 20 months. Near-Infrared Camera and Multi-Object Spectrometer (NICMOS) is a Hubble Space Telescope (HST) instrument based on three cameras designed for simultaneous operations and operating between 0.8 and 2:5 lm and using HgCdTe photo-conductive detectors. NICMOS has been installed on the HST during the second servicing mission in 1997. The cameras are cooled down to 50±60 K via 120 kg of solid nitrogen. Due to structural deformation of the cryostat the actual system performance and lifetime did not meet the original expectations and the instrument is not active since January 1999. It is planned to overcome such problems by replacing the instrument Dewar with a turbo-Brayton cooler, which will ensure a temperature of about 77 K [13]. A mission study involving cryogenic equipment not selected for ¯ight is represented by ESA's satellite test of the equivalence principle (STEP). Such a mission would have performed experiments on the equivalence of gravitational and inertial mass to a precision of one part in 1017 . Such a level of precision would have been guaranteed by superconducting accelerometers cooled at 1.8 K by a bath of super¯uid He [22]. Among the post-operation missions, we should mention Wide Field Infrared Explorer (WIRE), one of the NASA small explorers, launched in February 1999 and scheduled to remain operational for a period of 4 months. Due to technical problems, the satellite was lost during its commissioning phase. It was supposed to survey the sky at mid-infrared wavelengths between 12 and 25 lm with a sensitivity 1000 times better than IRAS. A two-stage, solid hydrogen cryostat maintained the optics colder than 19 K and the 128  128 Si:Ga detector array below 7.5 K [23].

802

Table 2 Summary of cryogenic space programmes (space science)a Mission

Type/class

Launch year

Cryogenic system

In-¯ight T (K)

Lifetime

Orbit

Status

3

290 dd

Near-polar

Post-ops.

2±4 1.4±1.6 1.8 0.3

± 305 dd 840 dd 30 dd

LEO Near-Earth HEO LEO

Not approved Post-ops. Post-ops. Post-ops.

<8 60 0.065

600 dd 700 dd 1 dd

LEO LEO Suborbital ¯ight

Post-ops. Post-ops. Post-ops.

<7.5 1.8 0.065

120 dd 180 dd 730 dd

LEO LEO LEO

Post-ops./loss Not approved Loss

85 1.4 <7.5 0.1±0.3

2±5 yr 2.5 yr 400 tbd

HEO Earth trailing Polar LEO

Development Development Not approved Study

0.05±0.3

>10 yr

LEO

Study

0.3 & 1.7

4.5 yr

Sun±Earth L2

Development

0.1 & 20 4±40

460 5±10 yr

Sun±Earth L2 Sun±Earth L2

Development Study

0.05 20 4 30

3±5 yr tbd tbd 5 yr

Sun±Earth L2 HEO L2/Earth trailing L2/Earth trailing

Study Study Study Study

80 1.4±1.6

10 yr ?

Heliocentric ±

Development Not approved

Science/IR

Satellite (surveyor)

1983

4

Science/IR Science/IR Science/IR Science/IR

Spacelab payload Satellite (surveyor) Satellite (observat.) Instrument (IRST)

± 1989 1995 1995

4

MP/UV to FIR Science/NIR Science/X

Satellite (observat.) Nicmos, instrument Sounding rocket

1996 1997 1996±1999

Science/IR Science/FP Science/X

Satellite (surveyor) Satellite Satellite (observat.)

1999 ± 2000

INTEGRAL (ESA) SIRTF (NASA) NGSS (NASA) Submillimetron (ASC)

Science/gamma Science/IR Science/IR Science/sub-mm

Instrument (observ.) Satellite (observat.) Satellite (surveyor) ISS telescope

2001 2002 ± >2004

XEUS (ESA)

Science/X

Instrument (observ.)

2005

Herschel (ESA)

Science/IR

Satellite (observat.)

2007

Planck (ESA) NGST (NASA)

Science/FIR Science/NIR

Satellite (surveyor) Satellite (observat.)

2007 2008

Constellation-X (NASA) ARISE (NASA) DARWIN (ESA) TPF (NASA)

Science/X Science/radio Science/IR Science/IR

Satellite Satellite Satellite Satellite

2008±2010 2008 >2009 2010

Rosetta (ESA) LEDA/MORO (ESA)

Science/Comet Science/Moon

Instrument (probe) Satellite (surveyor)

IRAS (NASA, NIVR, SERC) GIRL (FMST, D) COBE (NASA) ISO (ESA) SFU (ISAS/NASDA/ MITI) MSX (BMDO,US) HST (NASA) XQC (NASA-Wisc. Univ.) WIRE (NASA) STEP (ESA) Astro-E (ISAS, NASA)

(observat.) (VLBI) (VLBI) (VLBI)

FP: fundamental physics; MP: multipurpose mission (defence + science).

2003 ±

He (k) cryostat

He (k) cryostat He (k) cryostat 4 He (k) cryostat 4 He (k) cryostat ‡3 He SC sH2 cryostat sN2 cryostat 4 He (k) cryostat ‡ ADR Dual, sH2 cryostat 4 He (k) cryostat sNe ‡4 He cryostat ‡ ADR Stirling cooler 4 He (k) cryostat Dual, sH2 cryostat 4 He (k) cryostat ‡3 He SC Stirling cooler ‡ ADR 4 He (k) cryostat ‡3 He SC H2 & 4 He JT ‡ DR Passive radiation ‡ cooler Astro-E like/coolers Cryo-cooler ‡ H2 JT Cryo-cooler ‡ H2 JT Passive radiation ‡ cooler Stirling cooler Cryostat ‡ cooler 4

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819

a

Application

Table 3 Summary of cryogenic space programmes (applications/technology)a Application

Type/class

Launch year

Cryogenic system

In-¯ight T (K)

Lifetime

Orbit

Status

Meteosat 1±7 (ESA/EUM) ERS-1 (ESA) ERS-2 (ESA) CRISTA (DARA, D) MSG-1 (ESA/EUMETSAT) ENVISAT 1 (ESA) ENVISAT 1 (ESA) Metop (ESA/EUM/NOAA) MSG-2 (ESA/EUMETSAT)

Meteostat Earth observat. Earth observat. Earth observat. Meteosat Earth observat. Earth observat. Meteosat Meteosat

P/L P/L P/L P/L P/L P/L P/L P/L P/L

1977±1997 1991 1995 1994/1997 2000 2001 2001 2001 2002

Passive radiator Stirling cooler Stirling cooler 4 He …>k† cryostat Passive radiator Stirling cooler Stirling cooler Passive radiator Passive radiator

90 80 80 4 75±85 80 80 100 75±85

2 yr 2 yr 10 dd 7 yr 5 yr 5 yr 5 yr 7 yr

GEO LEO ± near-polar LEO ± near-polar LEO GEO LEO ± polar LEO ± polar Sun-synchr. polar GEO

Post-ops./ops. Operations Operations Post-ops. Development Development Development Development Development

USMP/LPE (NASA) SHOOT (NASA) HTSSE I±II (NRL/USAF) STRV-1B (DRA) IN-STEP/CSE (NASA)

Technology Technology Technology Technology Technology

P/L (STS-52) P/L (STS-57) P/L (ARGOS) Mini-satellite P/L (STS-63)

1992 1993 1993±1999 1994 1995

4

2.2 <2.2 70±80 80 65

>6 dd >6 dd 3 yr 3 yr 8 dd

LEO LEO Sun-synchr. polar GTO LEO

Post-ops. Post-ops. Loss/ops. Post-ops. Post-ops.

BETSCE (NASA)

Technology

P/L (STS-77)

1996

10

<1 dd

LEO

Post-ops.

MIDAS (NASA) CheX (NASA) ISS/Bosch (ESA)

P/L (STS79/MIR) P/L (STS-87) P/L (ISS)

1996 1997 >2005

80 1.6 77

>8 dd >6 dd >1 yr

LEO LEO LEO

Post-ops. Development

LTMPF (NASA) FACET (NASA/JPL)

Technology Technology/MS Technology/ TLC Technology/MS Technology

He …> k† cryostat He …k† cryostat Stirling cooler Mechanical cooler Mechanical cooler ‡ heat pipe H2 Stirling ‡ JT + sorpt. Mechanical cooler 4 He …k† cryostat Mechanical cooler

P/L (ISS) P/L (STS)

2003 <2003

4 He …k† cryostat sCO2 ‡ sNe cryostat

1.6 1.9

180 dd >6 dd

LEO LEO

Development Development

UARS (NASA) Landsat 7 (NASA) Terra (NASA) Aqua (NASA) Aura (NASA)

Atmosphere Earth observat. Earth observat. Earth observat. Earth observat.

P/L P/L P/L P/L P/L

1991 1999 1999 2000 2002

sNe ‡ 4 He (k) cryostat Passive radiator Stirling cooler Stirling-PTR ‡ passive Stirling-PTR ‡ passive

16 90 80 60±85 65

1.5 yr 5 yr 5 yr 6 yr 5 yr

Near-circular Sun-synchr. polar Sun-synchr. polar Sun-synchr. polar Sun-synchr. polar

Post-ops. Operations Operations Development Development

(ATSR) (ATSR) (STS-66/85) (Seviri) (MIPAS) (AATSR) (IASI) (Seviri)

(CLAES) (MISR) (MODIS) (CERES) (HIRDLS)

4

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819

a

Mission

P/L: payload/instrument onboard a satellite/STS/ISS; STS: Space Shuttle ¯ight; MS: materials science; TLC: telecommunications.

803

804

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819

Fig. 4. ISO, the Infrared Space Observatory of the European Space Agency, fully integrated and ready for transport to the launch facilities. The solar panels shield the satellite from the direct sun illumination. The cryostat is ®xed to the service module via struts visible in the lower part of the picture.

Astro-E is a satellite for X-ray astronomy launched in 2000 [4]. The satellite is not operational due to a launcher problem. Astro-E is a combined NASA and ISAS (Japan's Institute for Space and Astronautical Science) e€ort and was designed to provide X-ray images along with high resolution spectra from 0.4 to 700 keV. One of the instruments onboard, the high resolution spectrometer (XRS), was based on an array of 2  6 micro-calorimeters operating at 65 mK. Such a temperature is maintained by an adiabatic demagnetisation refrigerator (ADR) hosted in a liquid helium cryostat and thermally shielded by solid-neon cooled outer jacket. The cryogens should have ensured an in-orbit lifetime of about 2 years. It is also worth mentioning the rocket program X-ray quantum calorimeter (XQC), which has ¯own Nike± Black Brant rockets provided by NASA and carrying a small array of X-ray micro-calorimeters cooled at 65 mK by means of an ADR [24]. 2.2.2. Missions under development Several spacecrafts presently undergoing development will make use of cryogenic instrumentation. An overview is provided below (see also Tables 2 and 3).

Planck is the third medium size mission (M3) in ESA's scienti®c plan Horizon 2000 [25]. Its main objective is to map the temperature anisotropies of the Cosmic Microwave Background (CMB) over the whole sky with a sensitivity of DT =T ˆ 2  10 6 and an angular resolution of 10 arc-min. Such goals require bolometers operating at 0.1 K, HEMT at 20 K and a low emissivity, cooled telescope (60 K). The cryogenic system proposed for Planck is based on pre-cooling to 60 K by passive radiators, cooling to 20 K with a H2 Joule±Thomson cooler (adsorption compressors), cooling to 4 K with a He Joule±Thomson cooler (mechanical compressors) and ®nal cooling to 0.1 K with an open-loop dilution refrigerator. Nominal mission lifetime is 15 months. Herschel (formerly known as FIRST) is the fourth cornerstone mission of the ESA's scienti®c plan [26]. It is dedicated to astronomical observations in the FIR and sub-mm wavelength range from 85 to 600 lm. Herschel is a multi-user observatory, based on a super¯uid helium Dewar at 1.65 K and on a 3 He sorption cooler delivering a base temperature of 0.3 K. The scienti®c goals will be achieved with three instruments operating, respectively, at 2 K (HIFI, heterodyne receiver based on SIS mixers), 1.7 K (PACS, spectro-photometer partly based on

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819

photo-conductors) and 0.3 K (again PACS and SPIRE, another spectro-photometer using bolometers). Herschel is presently scheduled for launch in 2007 and its He Dewar is designed for a mission lifetime of 4.5 year. Due to the commonality in technologies, science objectives and ®nal orbit (around the second Lagrangian point of the Sun±Earth system), ESA has decided to develop Herschel and Planck together, and to launch them with a single Ariane-5 ¯ight. Detail engineering assessments are ongoing. Space Infrared Telescope Facility (SIRTF) is the fourth member in NASA's family of `Great Observatories' [27]. It is designed to perform imaging and spectroscopy in a large wavelength range from 3 (NIR) to 180 (FIR) lm via a 0.85 m diameter helium cooled telescope. The detectors' temperature is 1.4 K, while the cryogenic system is optimised (passive radiation and ecient use of helium gas enthalpy) to make use of only 360 l of super¯uid He for a minimum lifetime of 2.5 years. SIRTF is now in the development phase and is presently scheduled for launch in May 2002. Thanks to a number of trade-o€s, it has been possible to drastically reduce the mission costs by selecting a solar obit and limiting the satellite mass to about 900 kg. Finally, among the missions under development, we would like to mention International Gamma-Ray Astrophysics Laboratory (INTEGRAL), as an example of utilisation of space quali®ed Stirling cryo-coolers [28]. INTEGRAL is a medium size ESA science mission dedicated to spectroscopy and imaging between 15 keV and 10 MeV. The spectrometer on the spacecraft is based on about 30 kg of germanium detectors maintained at a temperature of 85 K. The satellite is scheduled for launch in 2001. 2.2.3. Missions under study Next Generation Space Telescope (NGST) is considered to be the successor of the HST [29]. The programme calls for a 6±8 m diameter passively cooled telescope to minimise thermal self-emission and enabling observations in the NIR and medium-IR (MIR) from 1 to 30 lm. The science objectives of the NGST are the study of galaxies, stars and planets formation and the study of the chemical and geometrical evolution of the universe. The so-called `NGST Yardstick Mission' (a mission design developed by NASA, academia and industry starting from 1996) baselines a deployable three-mirror telescope passively cooled below 50 K. The science instruments are an NIR camera, an NIR low resolution spectrograph and an MIR camera±spectrograph combination. The ®rst two instruments operate at 30 K (passive radiator cooling), the third one makes use of either a turbo-Brayton mechanical cooler or a combination of H2 and He sorption coolers to achieve a base temperature of about 8 K. ESA is also involved in the

805

NGST project with a ®nancial participation of order 15%. The telescope should be launched in 2008. Next Generation Sky Survey (NGSS) is another NASA mission proposed as candidate for NASA's medium-class Explorer (MIDEX) programme, but ®nally not selected (October 1999). NGSS should have provided a sky survey from 3.5 to 25 lm with a sensitivity 1000 times better than IRAS. The satellite was based on a design similar to WIRE with a cryostat based on solid hydrogen. With a total mass of about 890 kg, NGSS would have included a 0.5 m telescope and a four-channel imager based on HgCdTe and Si:As photo-conductive 1024  1024 element arrays. When selected, NGSS should have been launched either in 2003 or in 2004. X-ray Evolving Universe Spectroscopy Mission (XEUS) is the potential follow-on mission to the ESA XMM cornerstone (launched at the end of 1999). The mission aims to place a permanent X-ray telescope in orbit by exploiting the facilities available on the ISS and by ensuring a signi®cant growth and evolution potential [6]. Main features of the proposed observatory are the very large telescope aperture and the utilisation of cryogenic detectors in two narrow ®eld imaging spectrometers (respectively, TESs and STJs). The cryogenic design would be based on Stirling mechanical coolers combined with ADR systems in order to extend the mission lifetime beyond which is achievable with consumable cryogens. Another science study involving cryogenics in space is Submillimetron, a sub-millimetre wave cryogenic telescope proposed for the Russian segment of the ISS [30]. The project is a co-operative e€ort of Chalmers University (Sweden), Astro Space Centre and IREE (Russia) and JPL (US). Submillimetron would make use of a cryostat ®lled with super¯uid helium and of a 3 He sorption cooler to achieve a base temperature of 300 mK. The telescope would be cooled to about 5 K in order to minimise the role of any residual thermal background. Constellation-X is a next generation X-ray observatory presently under study by NASA [7]. Such a mission is based on 2±4 identical satellites orbiting at the second Lagrangian point of the Earth±Sun system (L2), thus achieving a signi®cantly larger photon collection area in a wide energy range between 0.25 and 40 keV. At the focal plane of the telescopes, micro-calorimeter arrays will ensure high resolution imaging spectroscopy. The cryogenic system needs to allow operating at 50 mK; di€erent solutions are presently being considered, including a design similar to Astro-E (outer solid neon, intermediate 4 He and inner ADR stage) and a revised version, replacing the solid neon with mechanical coolers. Cryogens should ensure a mission lifetime between 3 and 5 years. Launch would take place in 2008±2010. Advanced Radio Interferometry between Space and Earth (ARISE) is another NASA mission presently

806

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819

under study, consisting of a 25-m radio telescope in highly elliptical Earth orbit, observing in conjunction with a large number of radio telescopes on the ground and using Very Long Baseline Interferometry [31]. This would ensure high resolution images (10 larcsec) of the most energetic astronomical phenomena, such as active galactic nuclei (AGN). The mission requires low noise ampli®ers and receivers operating at 20 K; the present baseline consists of using a mechanical pre-cooler combined with hydrogen sorption coolers, similar to those used on Planck. DARWIN (Infrared Space Interferometry Mission) is a cornerstone candidate in the ESA `Horizon 2000+' science plan. Its goal is to detect terrestrial planets in orbit around other stars and to allow high resolution imaging in the medium infrared between 5 and 30 lm. Interferometry would be carried out over a 50±500 m baseline including six free ¯ying 1.5 m telescopes. Both the telescopes and the focal plane detectors would be cooled to about 20±30 K [32]. Terrestrial Planet Finder (TPF) is a similar mission under study at NASA, based on ®ve spacecrafts ¯ying in formation at about 1 AU from the Sun and focusing on the identi®cation of terrestrial planets outside our solar system. As such, TPF is facing technological issues very similar to DARWIN including cold optics and IR detectors. The present system baseline relies on passive cooling to about 40 K and a Brayton cryo-cooler to cool the IR detectors down to 5 K [33]. 2.3. Technology validation missions A number of cryogenic space missions have been dedicated to the validation of speci®c technological issues. It is the case of payloads onboard research satellites, ¯own onto the Space Shuttle or to be ¯own on the ISS. High Temperature Superconductivity Space Experiment (HTSSE) I and II are payload units for validating HTS components in space-based systems [34]. They were developed by the US Naval Research Laboratory (NRL) and launched in 1993 and 1999 onboard the US Air Force satellite ARGOS (Advanced Research and Global Observation Satellite). HTSSE II is based on a Stirling cooler with an operating temperature between 70 and 80 K. A number of HTS-based devices have been tested onboard this facility, including ®lter banks, patch antenna arrays, delay lines, ADCs and multiplexers. STRV-1A and 1B were developed by the Defence Evaluation Research Agency (DERA) of the UK Ministry of Defence as two small space technology test-beds and launched together in the geostationary transfer orbit (GTO) by an Ariane-4 in June 1994 [35]. The unit 1B carried a mechanical cooler capable of a base temperature of 80 K.

NASA has ¯own a number of cryogenic payloads on several Space Shuttle ¯ights. The lambda point experiment (LPE) is one of them, a small payload with a cryostat containing super¯uid helium and dedicated to the study of its behaviour in the absence of gravity [36]. LPE ¯ew on STS-52 as part of the US Microgravity Payload (USMP) ± 1 programme in 1992. A ¯ight demonstration has been conducted by NASA with super¯uid helium on-orbit transfer (SHOOT), a facility dedicated to experiments on the transfer of super¯uid helium between two vessels in low gravity conditions. The facility successfully ¯ew on STS-57 in 1993 and demonstrated the possibility to exploit the fountain e€ect as the basis of a thermal±mechanical pump [37]. Another payload ¯ew in 1995 as part of the NASA InSpace Technology Experiments Program (IN-STEP), the cryo-system experiment (CSE), dedicated to the inorbit characterisation of a hybrid cryogenic cooling system incorporating a 65 K Stirling cryo-cooler and an experimental diode oxygen heat pipe from Hughes [38]. Such a heat pipe allows to increase the distance between the cooler and its load and to thermally disconnect them when the cooler is o€ (reducing the reverse heat ¯ow). Brilliant Eye-Ten Kelvin Sorption Cooler Experiment (BETSCE) ¯ew onboard STS-77 (1996) and demonstrated the capability to cool down IR detectors at about 10 K by using a Joule±Thompson hydrogen cooler, based on a metal-hydride sorption compressor/pump system and pre-cooled by a Stirling cooler (65 K) [39]. The base temperature could be maintained for over 20 min with a heat load of 100 mW. Materials in Devices as Superconductors (MIDAS) is a NASA cryogenic facility for the characterisation of high temperature superconductors during extended ¯ights. It is based on a Stirling cooler capable of 1 W cooling at 80 K, with a maximum power consumption of 60 W [40]. Total weight of the facility corresponds to only 30 kg. MIDAS ¯ew onboard STS-79 (1996) to be transferred onboard the Russian space station MIR, where it remained for about 3 months before returning to the ground with STS-81. Con®ned Helium Experiment (CHEX) is another test facility from NASA that ¯ew on STS-87 (1997) as part of USMP-4. The experiment hardware is based on the LPE equipment (see above), with a 4 He ®lled cryostat at 1.6 K, con®ned in 57 lm wide gaps between 400 thin silicon wafers. The experiment aimed to investigate the behaviour of He at the transition between super¯uid and liquid state [41]. Low Temperature Microgravity Physics Facility (LTMPF) is a NASA science project built to provide long duration low temperature and microgravity conditions onboard the ISS. The unpressurised module will be attached to the Japanese experimental module (JEM) and is scheduled for ¯ight in 2003. LTMPF will be based

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819

on a liquid helium Dewar with a lifetime goal of about 5 months [42]. Finally we should mention FACET (fast alternative cryogenic experiment test-bed), a proof of concept study for the development of a compact and less expensive ¯ight facility for low temperature experiments, ®tting into the space of two Space Shuttle hitchhiker cans. The module would use a cryostat with a solid neon guard and super¯uid helium guaranteeing an orbit lifetime longer than 6 days. FACET should allow to perform a larger number of cryogenics experiments before the entry into service of LTMPF [43]. 2.4. Applications In this paragraph we review the space missions which have made or are planning to make use of cryogenic equipment within the ®elds of telecommunications and earth observations (here considered as applicative areas). We should notice that in these two ®elds the required operating temperatures are higher than for scienti®c missions (typically >10 K). Considering the recent progress made by cryogenics, such a requirement is not too dicult to meet, allowing a number of possible applications and related design solutions onboard satellites. 2.4.1. Telecommunications The recent progress in the area of HTS has opened new perspectives for the fabrication of RF superconducting devices, such as ®lters, delay lines, resonators and antennas. Although superconductors have `zero' DC resistance, at ®nite frequencies there are losses which remain smaller than those of a normal metal for all frequencies up to the millimetre wave region. For instance, the superconductor YBCO (yttrium±barium± copper-oxide) at a temperature of 77 K and at a frequency of 10 GHz has a surface resistance that is 30 times smaller than that of copper at the same operating frequency and temperature. Such a reduced surface resistance implies lower insertion losses and the possibility to reduce considerably (of order 10 times) the size of the devices (e.g. ®lters) without signi®cant eciency degradation. In engineering terms this implies the possibility to signi®cantly improve the energy eciency of high frequency telecommunications systems or to signi®cantly reduce their size and weight without increasing the insertion losses [44]. Given a high frequency system, the more the components bene®t from a low operating temperature the more attractive will be the selection of a cryogenic design. Such a system would simultaneously employ passive HTS components (taking advantage of the low insertion losses) as well as STJ-based cryoelectronics and cooled semiconductor active devices (such as HEMTs and Schottky diodes). In addition, the

807

superior Q-factors in the HTS ®lters allow for improved spectral eciency via reduced guard bands [45]. The ®nal decision to adopt a cryogenic design for a satellite subsystem will be based on a trade-o€ analysis considering the mass savings, the total power consumption and the reliability of the di€erent solutions. A more detailed discussion will be presented in Section 3. We should mention that the total energy balance depends heavily on the eciency of the cooling system: a passively cooled subsystem will have a positive energy balance, while an actively cooled unit (e.g. Stirling cooler) has a more critical energy balance. The choice between passively and actively cooled systems will depend on the required operating temperature and on the type of orbit of the satellite. As an example we can discuss the case of band-pass ®lters for output multiplexers (OMUX) at about 4 GHz (C-band) [45]. The insertion losses are inversely proportional to the resonator quality factor QHTS , which are considerably higher than in conventional technology (Qconv ). The decreased insertion losses translate into a lower RF ampli®er power, and thus into lower power consumption. Due to the additional cooler power, power saving can only be achieved if QHTS =Qconv > 1 ‡ gamp =gcooler ;

…1†

where gamp and gcooler are, respectively, the power eciency of the RF ampli®er and of the cryogenic cooler. Using state-of-the-art in ampli®er …gamp of order 0.5) and cooler …gcooler of order 0.06) technology, we derive that if QHTS > 1:5  105 we can achieve savings in power consumption in addition to the miniaturisation of the payload (potentially leading to increased system capacity). The US NRL, in collaboration with the US Air Force, developed HTSSE, a technology demonstration programme aimed at proving the viability of HTS devices for space applications (see Section 2.4). Several organisations have participated in this programme [34], including numerous industries. HTSSE I (lost due to a launch failure) aimed at the testing of single YBCObased devices, such as resonators, ®lters, delay lines, cavities and patch antennas (operating between 1.4 and 10 GHz). HTSSE II focused on more advanced HTS components and subsystems, including hybrid (HTSsemiconductor) receivers, multiplexers and A/D (analogue/digital) converters. In Europe, Bosch Telecom is currently developing a 3channel transponder including a front-end receiver with HTS noise reduction ®lters and cooled low noise ampli®er. The demonstrator includes a 3-channel HTS input and output multiplexers. The system is scheduled for delivery to the ISS in 2001 [46]. Cooling will be provided by two redundant cryo-coolers with the objective of evaluating the expected mass and power savings. ESA has also initiated some actions in the frame of the

808

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819

ARTES-5 development programme, in order to identify components, cooling techniques and system architecture related to the use of HTS in telecommunication satellites. 2.4.2. Earth observation and meteorology satellites The ®eld of Earth observation (i.e. the remote sensing of our planet from space for civilian purposes) has grown considerably in importance over the last 10±15 years. Several missions have been developed with the objective of monitoring earth's natural environment and to study the natural phenomena related to the planet water cycle. Cryogenics is required because of the utilisation of detectors capable of imaging the earth surface in the NIR and MIR (typically operating around or just below 100 K). Due to the low altitude orbits of these satellites, the large thermal ¯ux emitted by the earth precludes the utilisation of purely passive means, thus obliging to make use of mechanical coolers. Upper Atmosphere Research Satellite (UARS) is the ®rst major element of NASA's Mission to Planet Earth (MTPE) and was launched in 1991 to carry a comprehensive study of the upper atmosphere in terms of chemical composition, temperature, winds and energy balance [47]. Four instruments were devoted to spectroscopic measurements and among these cryogenic limb array Etalon spectrometer (CLAES) was designed to measure the atmospheric radiance and the concentrations of several molecular species [48]. The instrument consists of a solid-state Fabry±Perot spectrometer coupled with a re¯ective telescope and a photo-conductive detector. The wavelength coverage is from 3.5 to 12:7 lm. A solid CO2 (123 K) and a solid Ne (16 K) cryostat maintained the detector at about 16 K, the spectrometer at 50 K and the telescope optics at 150 K. The instrument operated for a total of about 19 months. LANDSAT 7, launched in 1999, is the latest NASA satellite of a series of earth observation missions dating back to 1972 [49]. It o€ers signi®cant improvements over its predecessors, maintaining the characteristics of a thematic mapper, with ground resolutions of 30 m (visible range) and 60 m (NIR range). Adequate NIR performance can be achieved by passively cooling the imaging system at about 90 K. Terra (formerly known as EOS AM-1) is the ®rst mission of NASA's Earth Observing System [50], launched in December 1999. Terra's objectives are the study of seasonal and climate changes over a nominal lifetime of 5 years. Advanced spaceborne thermal emission and re¯ection radiometer (ASTER) is a high spatial resolution imaging instrument onboard Terra, and a co-operative e€ort between NASA and Japan's Ministry of International Trade and Industry. ASTER includes a thermal infrared imaging unit, based on HgCdTe detectors cooled at 80 K by a low vibration and long lifetime Stirling cooler.

Aqua (formerly EOS PM) is the second mission of NASA's Earth Observing System. Due for launch in December 2000, Aqua objective is the multi-disciplinary study of the Earth's interrelated processes (atmosphere, oceans and land surface) and their impact on earth system changes. Atmospheric infrared sounder (AIRS) is an IR spectro-photometer onboard Aqua. The focal plane of the spectrometer is based on an array of HgCdTe detectors cooled to 60 K by a combined Stirling/pulse tube cryo-cooler, while the entire optics is cooled to 150 K by a two-stage radiative cooler. MODIS is a wide band radiometer also onboard Aqua, with two focal plane assemblies cooled down to 85 K by a high performance passive radiator cooler [51]. Aura (formerly EOS Chem) is the third mission of NASA's Earth Observing System. Due for launch in 2002, Aura objective is the study of the Earth's atmosphere chemical composition, with speci®c attention to the monitoring of the ozone layer. The instrument HIRDLS (high resolution dynamics limb sounder) is an infrared scanning radiometer onboard Aura, designed to scan the upper layers of the atmosphere in the wavelength range between 6 and 17 lm. An array of 21 detectors is cooled down by a mechanical cooler to about 65 K [52]. The Earth Observation Programme of the European Space Agency is based on a number of missions for the monitoring of our planet's atmosphere, oceans and land. The ESA satellites ERS 1 and ERS 2 (ESA Remote Sensing Satellites) were developed to provide information on the earth and its environment and were launched, respectively, in 1992 and 1995. Onboard ERS 1 and ERS 2 the IR radiometer ATSR (along track scanning radiometer) was equipped with Stirling cycle coolers (Oxford University) cooling the focal plane assembly to about 100 K [53]. Cryogenic Infrared Spectrometers and Telescopes for the Atmosphere (CRISTA) is a limb-scanning satellite experiment developed by the University of Wuppertal and ¯own on the free-¯ying ASTRO-SPAS satellite (DASA, D). CRISTA 1 ¯ew in 1994 (STS-66) and CRISTA 2 in 1997 (STS-85). The spectrometers were cooled by liquid He with a lifetime of the order of 10 days. Meteosat Second Generation (MSG) continues the legacy of the previous Meteosat missions, with largely improved performance. Three satellites (MSG 1 to 3) are going to be procured by ESA on behalf of Eumetsat and launched to guarantee uninterrupted coverage from 2000 to 2012. Onboard MSG 1 two instruments have focal plane assemblies operating at low temperature, the imager SEVIRI (spinning enhanced visible and infrared imager) and the radiometer GERB (Geostationary Earth Radiation Budget), both cooled by passive radiators and operating at about 80 K [54].

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819

Envisat 1 is a large multidisciplinary mission, dedicated to the study of the earth and atmosphere environment and having both science and application objectives [55]. Onboard this spacecraft (due for launch in July 2001) three instruments require cryogenic temperatures to operate. These are MIPAS, AATSR and Sciamachy. MIPAS (Fig. 5) is a fourier transform spectrometer operating in the wavelength range between 4 and 16 lm and using photo-conductive and photovoltaic HgCdTe detectors. Optics and detector assembly are cooled to 70 K by a pair of Stirling cycle coolers [56,57]. AATSR is an IR±visible radiometer whose focal plane assembly is cooled to 80 K by another pair of Stirling coolers. Finally, Sciamachy is an imaging spectrometer operating between 0.2 and 2.4 lm; it uses silicon and InGaAs detectors cooled down passively to di€erent temperatures ranging between 235 and 130 K. ESA, EUMETSAT, CNES and NOAA are co-operating in the development of meteorological operational (MetOp), a new generation of weather satellites [58]. MetOp will continue part of the ESR mission, complement the results provided by Envisat and allow for scienti®c investigations as well as weather forecasts. MetOp, presently undergoing development, will carry instruments similar to Envisat, with matching cryogenic requirements. 2.5. Cryo-electronics and large-scale applications 2.5.1. Cryo-electronics In the last 10 years electronics systems operating at cryogenic temperature (often indicated by the term cryoelectronics) have found numerous applications in several ®elds, including ¯ight hardware onboard spacecrafts.

Fig. 5. The instrument MIPAS before integration onboard the ESA satellite Envisat. The spectrometer operates between 4 and 16 lm and its detector assembly is cooled down to 70 K by a pair of Stirling cycle coolers. MIPAS is about 1.4 m long, 1.0 m large and 1.5 m high (including optics).

809

We will make here a distinction between semiconductorbased devices operating at low temperature in order to reduce their noise ®gures and alternative devices (such as superconductor-based devices) which can only be operated at low temperatures. In the ®rst category we should mention the front-end electronics (FEE) of several scienti®c payloads especially in the area of astrophysics. In many cases the JFET ampli®ers responsible for the initial signal conditioning are anchored to a heat sink at about 20 K and operated at about 120 K when powered on. This con®guration allows to achieve the lowest possible noise levels. As examples we should mention the JFET pre-ampli®ers of XRS onboard Astro-E and the detector assemblies of ISOPHOT, ISO-SWS and ISO-LWS, onboard the ESA satellite ISO (Infrared Satellite Observatory). In the latter cases, the FEE had to be positioned as close as possible to the detectors and to operate at about 1.6±1.8 K. In the case of ISOPHOT, these requirements were met by developing a speci®c MOS IC which demonstrated low noise, low dissipation, multiplexed operations [59]. The use of cold read-out electronics is planned by most of the science missions especially in the sub-mm and FIR wavelength range. The NASA mission ARISE is also planning to make use of cryo-electronics in the form of low noise ampli®ers and receivers (in the 5±8 GHz frequency range) operating at about 20 K and, for this reason, requiring the presence of a cryogenic cooler. Over the last 10 years a remarkable e€ort has been invested in the development of a new generation of high speed digital electronics capable of outperforming the present semiconductor technology. Superconducting electronic (SCE) devices have shown interesting performance and the potential to provide the speed required by future high density and large speed computing systems. Such devices allow increasing packing density without the power dissipation problems su€ered by the present technology. Two main e€ects are used by SCE in the ®eld of digital applications: (1) ¯ux quantisation and (2) Josephson e€ects [46]. Such devices can be used as basis for the logic gates and have very fast intrinsic switching speeds (of order 1 ps) and low power dissipation. The switching energy of LTS and HTS technologies is typically of the order of 10 17 ± 10 18 J against 10 13 J of CMOS devices operating at room temperature. More recently a new technology has been investigated using single-¯ux-quantum (SFQ) devices and based on the movement of single quanta of magnetic ¯ux rather than on the di€erent voltage levels. The SFQ technology does not require any hysteretic STJs and o€ers even lower energy dissipation [60]. Another important category of cryogenic devices is represented by SQUIDs. A SQUID is based on two STJs connected in series to form a closed loop; such a device is the most sensitive magnetometer known to date, reaching sensitivities of order of a few femtoTesla

810

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819 0:5

(fT)/…Hz† at frequencies of a few Hz (1 fT ˆ 10 11 G). SQUIDs are used in several applications including very sensitive gravity gradiometers, fast digital electronics and detector read-out circuitry. While the lowest noise performances are achieved by LTS devices, it is envisaged that HTS-based units (T > 70 K) will be used in applications not requiring very high sensitivities and could be appropriate for high throughput digital systems [46]. 2.5.2. Large-scale applications Large-scale applications involving cryogenics in space include two main categories, energy storage in large superconducting magnets and gas storage by liquefaction for life support systems and propulsion purposes. Energy can be stored in the form of an intense magnetic ®eld generated by a large current ¯owing in a superconducting magnet with very low resistive losses. Such a storage technique o€ers interesting advantages in situations requiring the sudden absorption of large energy amounts that cannot be supplied by the more traditional combination of solar cells and batteries onboard spacecrafts. Similar systems have been considered in the context of the Strategic Defence Initiative sponsored by the US Department of Defence. Typical operating temperatures would be close to 4.2 K in the case of magnets based on traditional LTS windings. The storage of gases in lique®ed form is well known and has been used since many years in the context of spacecraft propulsion systems. Such a technique o€ers obvious advantages for the storage of large amounts of gas in a limited volume tank. In addition to propulsion units, we can envisage the extension of this technique to life support systems for inter-planetary missions (such as the planned mission to Mars) or for the ISS [61]. Required operating temperatures are less demanding in comparison with other cryogenics applications and vary depending on the di€erent gases (essentially N2 and O2 ) ranging between 70 and 90 K. This particular ®eld is clearly of interest also for any future manned station installed on the moon surface.

which are used to process signals coming from the earth (earth observation, meteorological, telecommunication) or space (astronomy). The instruments can either have their own optics or share a common unit (e.g. main telescope). Cryogenic installations have a strong impact on the architecture of a spacecraft or of an instrument; the key factors coming into play are summarised below. · A cooler needs to be used (i.e. cryogens, radiators, mechanical coolers). The cooler must have a heat lift compatible with the satellite size and the available power resources (see Fig. 7 for an overview of the different coolers). · The low temperature equipment must be properly supported, insulated from the room temperature satellite bus, and protected from solar and/or earth radiation. The lower the operating temperature, the higher the demands on the thermal insulation. · The cold parts have to be accessed (e.g. optical access to the detectors, signal wires, temperature sensors and heaters) and wiring needs to be routed between the cold payload, and the satellite bus for further processing (A/D conversion, data handling) before transmission to earth via telemetry. · Cryogenic ancillary equipment needs to be used to operate the cryogenic payload (e.g. heat links, heat switches, ®lters, thermometry). · System testability (instrument performances and payload cooling system) is likely to have an impact on the architecture of the payload. · The complete system must survive the vibrations induced by the launcher: this requirement has a strong impact on the cooler and on the instrument design (e.g. a compromise is required between the large sup-

3. Cryogenics and spacecraft engineering In this section we provide a summary of the engineering issues related to cryogenics in space. Overall spacecraft architecture, coolers, thermal insulation and ancillary equipment are described separately. 3.1. Architecture of cryogenic spacecrafts A spacecraft is usually composed of a satellite bus (or service module) and a payload module (Fig. 6). The payload module carries one or several instruments

Fig. 6. Typical architecture of a cryogenic satellite for space science. Three main sections can be identi®ed: the telescope (cooled below 100 K), the payload module (with focal plane detectors maintained below 4K) and the satellite bus (or service module maintained at room temperature).

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819

Fig. 7. Space coolers. The range of operations of each cooler is represented in terms of cooling power vs. operating temperature. Di€erent temperature ranges imply the utilisation of di€erent technologies.

port cross-section required for the launch, and the thermal insulation requirements). · The cooler has to operate in zero gravity for a period of time of order of a few years. · The payload needs to be built with materials compatible both for space and cryogenic environment. · The lifetime (or MTBF, main time between failures) of the equipment should be compatible and possibly exceed the mission duration. 3.2. Space coolers 3.2.1. Principle of cooling systems for space Coolers provide a cold heat sink, by removing the heat in the cold area, and dissipating it in the warm area. In the case of a satellite, isolated in space, the energy will be ®nally radiated to space. The process of cooling is well described by elementary thermodynamics: either the energy is directly radiated to space (via radiators) or some work has to be performed to pump the energy between two temperature levels, from a cold to a warm level to then be more easily radiated away. Such a heat pumping operation can be done according to an open cycle con®guration or in a closed cycle. The open cycle corresponds to the use of stored cryogens, where the work is performed before the mission, on ground, by a lique®er. The cold heat sink is provided by evaporation of liquid or solid cryogens. In this case, there is no energy to radiate, but gas is to be released. The lifetime of the system is thus governed by the heat losses and by the mass of cryogen that can be ¯own. The closed cycle corresponds to the use of mechanical coolers, where the work is done continuously during operations (see Fig. 8). Existing space coolers can provide about 1 W of cooling power in the temperature range 50±100 K (Stirling coolers, pulse tubes), about 100 mW in the range 15±20 K (double-stage Stirling) or a few mW at 4 K (Joule±Thomson). Very low temperature (VLT) coolers (e.g. 3He cryo-sorption refrigerators, dilution, ADR) rely on the pre-cooling systems mentioned

811

Fig. 8. Block diagram of a space cooler. The cold end of the cooler is interfaced to the focal plane (detectors), while its active part and control units are linked to the satellite structure and ultimately to the radiators. The heat load is minimised by thermally isolating the cryogenic area from the rest of the system.

above to reach even lower temperatures (typically between 100 mK and 1 K). In all cases, some electronics is required to monitor the temperature, maintain it constant, or drive the cooler mechanisms. For higher temperature systems …T > 50 K†, a single stage can be sucient. For lower temperature systems, multiple-stage coolers or a chain of various types of coolers have to be used. Fig. 7 provides an overview of the di€erent coolers and of the related cooling power and operating temperatures. 3.2.2. Type of coolers Radiators are the most ecient, simplest and more reliable space coolers. They are based on the fact that all objects emit infrared radiation proportionally to their area S, emissivity e, and to the fourth power of their temperature T , and on the fact that the environment temperature (deep space) is very cold (blackbody at T0 ˆ 2.73 K). The net cooling power is thus Qrad ˆ rSF e…T 4 T04 †  rSF eT 4 , where r is the Stefan constant and F  1 is the shape factor. Radiators are ecient above 100 K, but have limited performance at low temperatures (where the parasitic loads through the insulation increase) and have limitations related to their size (it is usually dicult to get more than a few m2 on a spacecraft). Fig. 9 shows the actual performance of satellite radiators against the theoretical heat rejection capabilities. Another limitation of radiators is their orientation: they need to be shaded from the solar radiation (1.4 kW/m2 ), and from the earth infrared and albedo radiation (about 300 W/m2 for the Earth), looking at the dark space in order to eciently radiate. This is a severe limitation, which can be managed only by constraining the spacecraft attitude and manoeuvring, together with a careful design of ba‚es and shields to reject the unwanted radiation. In addition, it is

812

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819

Fig. 9. Heat lift of radiators as a function of temperature and area. Actual radiators deviate from the theoretical expectation due to the actual emissivity of their surface. In practice it is dicult to run a passive radiator at T < 60 K even at orbits far away from Earth.

often necessary to have multiple stage radiators, thus increasing their complexity. There is therefore a lower limit to the temperature and cooling power that can be achieved with radiators. For low earth orbits (e.g. Earth Observation Satellites), the temperature limit is about 100 K, with a cooling power lower than 1 W/m2 . For geostationary orbits (at 36 000 km, e.g. telecommunication satellites), the temperature limit can be reduced to 75±90 K. For far away orbits (e.g. Lagrangian points), the earth radiation constraint vanishes, and the radiator architecture becomes simpler with lower temperature and better performance. In the case of Planck, it is expected to have a cooling power of about 2 W and a temperature of about 50 K; in the case of NGST or DARWIN it is estimated to have a cooling power of 200 mW at about 35 K. A stored cryogen cooler is composed of a cryogen tank, a vacuum vessel (isolating the cryogen tank before and during launch), ®lling and venting lines, heat shields/MLI, and some interface or volume for instrument accommodation. In the absence of gravity, the ¯uid needs to be maintained inside the tank by a phase separator (based on capillary forces or fountain e€ect for super¯uid helium). For ground Dewars, the cryostat neck is normally used as ®lling and venting line in addition to supporting the inner cryogen tank. Space Dewars, due to the dynamic loads present during the launch, are not compatible with this architecture. A separate venting line is used to eciently use the gas enthalpy to cool the shields, and to release the gas without applying momentum to the spacecraft (two jets symmetrically opposed around the centre of gravity). In space it is also possible to cool the whole vacuum vessel by radiation to space and by the venting line (as opposed to ground, where the tank must be at room temperature to avoid condensation). The ISO vacuum vessel (Fig. 10) once in space was at 110 K; the Herschel vessel is expected to be at 77 K, and the one of SIRTF at 5 K. In

Fig. 10. Cross-section of the ISO vacuum vessel. The drawing shows the main parts of the vessel including the di€erent thermal shields and the key structural components.

addition, the bath equilibrium pressure is not 1 bar as on ground, but it is vented to the space vacuum. This allows to pump on the cryogen bath and to use solid cryogens, which usually have a sublimation heat much larger than their latent heat. The proper design of the exhaust nozzle allows to tune the base temperature (vapour pressure) of the cryogen bath by adjusting the pressure drop. The volume of cryogen to be carried depends on the mission duration and on the heat input. The choice of the cryogen to be used depends on the base temperature required. The cryogens available do not provide a continuum of temperatures, but rather discrete values in

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819 Table 4 Cryostat/cryogen choice on a number of spacecrafts Missions

Cryogen

IRAS, COBE, ISO, Herschel, SIRTF IBSS, STEP WIRE XRS on ASTRO-E NICMOS

Superfluid 4 He Supercritical 4 He Solid H2 Solid Ne Solid N2

di€erent ranges. The most widely used are super¯uid or supercritical helium, solid H2 and solid Ne. An overview of the choices made on di€erent missions is presented in Table 4. For low temperature systems, in order to optimise the mass of cryogen, it is more interesting to use a bi-cryogen system, such as N2 and He, or H2 and He. However, the design is more complex, as all lines and valves have to be doubled. In most of the cases, a single cryogen is preferred, despite the mass penalty. Coolers based on cryogens are cooled on ground and topped up just prior to the launch. When on ground, the low vapour pressure is maintained by pumping on the bath through the vent line. This is usually possible only up to a few days before the actual launch. Such a period of `Launch Autonomy' might become a design driver for the cryostat, as the heat losses are signi®cantly higher on ground that in orbit, and increase as the heat shields are not vented. This requires an extra volume of cryogen or an auxiliary tank to compensate for such extra losses. In a mechanical cooler (or active cooler) mechanical work produced by moving parts is transformed into refrigerating power. There are many ways to classify active coolers. The most widely used is to distinguish between regenerative cycles (Stirling, pulse tube, Gi€ord coolers) and recuperative cycles (Joule±Thomson or Brayton coolers). The regenerative coolers are based on a pressure wave generated by a compressor (usually mechanical), and a cold ®nger, using a mobile (Stirling, Gi€ord) or a ®xed (pulse tube) regenerator. The heat is extracted at the cold end when the gas expands, and rejected at the warm end when the gas is compressed. The recuperative cycles use the enthalpy di€erence between high and low pressure gas. The Brayton cycle cooler use a cold turbine to expand the gas, whereas the Joule±Thomson (JT) coolers use the expansion through an ori®ce, and the properties of real gas to get the cooling e€ect. Being irreversible, the JT cooler (normally coupled to Stirling units) is less ecient than Brayton's, but it is simpler. Recently some e€ort has been invested in developing a new generation of mechanical coolers based on small turbines (typically running at 500 000 rpm) for both expansion and compression. Such coolers have the potential to o€er reasonable coecient of performance (COP) without the low frequency vibration problems of coolers with linear displacers.

813

The main di€erence between ground and space coolers is the required lifetime; a useful review is provided in [62]. A lifetime of 5 years is a typical requirement for most space applications. This means that no friction can be tolerated between moving parts. This leads to the development and quali®cation of coolers based on the `Oxford compressors': the compressor uses a linear drive, while the leak tightness of the compressed volume is guaranteed by a tight clearance seal (about 10 lm). A diaphragm spring is then used to maintain an alignment compatible with such a small gap, while allowing the axial motion of the piston. Life tests as long as 8 years have been performed with this system. Many such coolers are currently ¯ying. Another important limitation for space coolers is the electrical power demand. A typical power allocation for a cryo-cooler is between 50 and 200 W. Most mechanical coolers (typically based on the Stirling cycle) have an eciency of order 2±5% of the ideal Carnot cycle, implying a cooling power of a few mW at 4.2 K with an input power of about 100 W. Fig. 11 provides a summary of the COP of di€erent active coolers as a function of temperature, in the form of the ratio between cooling power and absorbed electrical power. The COP is compared to the eciency of the ideal Carnot cycle. Mass also represents a critical parameter in the evaluation of space coolers, since the typical allocated values are of order 100±150 kg. In addition, coolers should not generate vibrations degrading the performance of the sensitive detectors they are supposed to cool down. Vibrating forces are generated as reaction to moving masses within the cooler and such forces may cause the elastic deformation of the instrument structure, either a€ecting its alignment or causing electrical interference in the form of micro-phonic pick-up. To date most of the mechanical coolers proposed for space applications are based on the Stirling cycle or on the Joule±Thompson expansion, but more recently pulse

Fig. 11. COP of di€erent coolers. The solid line represents the ecient expected from an ideal Carnot cycle. The typical eciency achieved at 10±20 K is of order of 0.001.

814

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819

tube refrigerators have been proposed as an interesting alternative, due to the lack of moving parts and to the reduced level of vibration. Most of the development e€ort is concentrated on improving the eciency of such coolers. A small PTR developed by Lockeed±Martin in collaboration with NIST (US) has ¯own onboard the Space Shuttle mission STS-90, delivering 50 mW at 100 K, with an input power of order 10 W (COP ˆ 0.005). TRW (US) has also delivered several PTR units, among which a system used on the instrument AIRS is to be ¯own onboard the mission Aqua (formerly EOS-PM) with a cooling power of 1.75 W at 55 K (see Table 3). Finally we should mention that closed-cycle hydrogen sorption coolers are being developed by JPL (US), with the potential to o€er a vibration-free alternative to mechanical cooler in the 50±20 K temperature range; such an approach is presently baselined for the ESA mission Planck [63]. 3.2.3. Very low temperature coolers (T < 1 K) In many scienti®c satellite applications it is necessary to achieve even lower temperatures, well below 1 K [64]. Such a temperature range (see Figs. 1 and 7) can be achieved by using closed-cycle 3He sorption coolers (down to 250 mK), by dilution refrigerators (50 mK) and by ADR (50 mK). 3 He sorption coolers o€er interesting performance due to the simplicity of operations, the lack of moving parts and the possibility to work in a closed-cycle con®guration with an ecient duty cycle …3 He condensation phase vs. hold time at base temperature). Typical cooling power is of order 10 lW at 300 mK. Sorption coolers have already ¯own onboard balloons (Boomerang, Maxima, Archeops), sounding rockets and on the satellite SFU (IRST ± Infrared Telescope in Space instrument) [65,66]. Sorption coolers will be used onboard Herschel instruments SPIRE and PACS (see Table 2). Dilution refrigerators, based on the quantum mechanical properties of 3 He±4 He mixtures, are routinely used on earth to achieve temperature below 100 mK, with cooling power exceeding 100 lW. This technique is now being adapted for space applications and it is meant to ¯y on Planck (cooling power of order 0.1 lW) [67]. The absence of gravity and running the conventional mixture circulation in space are the main challenges for such a development. The proposed approach avoids the use of circulation pumps by working in open-loop mode, thus requiring a very large amount of gas mixture and o€ering a lifetime limited by the gas reservoirs. Alternative techniques may combine the capillary liquid con®nement with a closed-loop system based on cryosorption pumps [68]. Adiabatic demagnetisation refrigerators have already been used onboard sounding rockets and scienti®c satellites (Astro-E) [69]. They produce base temperatures of order 50±100 mK by reducing the entropy associated to

the electronic spins of the atoms of paramagnetic salts. Forcing the electronic spins to align in a single direction via a magnetic ®eld of order a few Tesla reduces the entropy. Cooling powers of about 10 lW are achieved [1]. ADRs o€er very low base temperatures with simple operations and good duty cycle eciency. The main challenges presented by ADRs are the need for large magnetic ®elds (implying large currents and potential EMI issues) and for high performance and high reliability thermal switches. The use of an ADR system is baselined by ESA for the future mission XEUS (Table 3). Solid-state coolers, analogous to Peltier elements but operating below 1 K, are also being investigated. They are based on the metal±insulator superconductor which provide cooling of the lattice by relying on phonon± electron coupling and removing the hottest electrons present in the normal metal electrode of the device [70]. Cooling of membranes from 0.3 to 0.1 K has already been achieved [71]. Such coolers are developed with the aim of building self-cooling detectors (bolometers or STJs) with simpler pre-coolers (e.g. 3 He sorption coolers). 3.3. Thermal insulation and ancillary equipment 3.3.1. Insulation technology for space The goal of thermal insulation is to limit the heat loads on the cold stage to a level compatible with the heat lift of the cooler. Due to the limited COP of space coolers, minimisation of the heat load is crucial to meeting the mission requirements. In space conductive and radiative coupling represent the loss mechanisms to be reduced by thermally insulating the cold stages. Low conductive supports such as metallic and composite supports are the key elements of the structure of any cryogenic payload, especially in view of the dynamic loads to be sustained during the launch phase. While steel and titanium alloys are still frequently utilised, Kevlar, glass and carbon ®bre-based materials are being extensively investigated and used in ¯ight hardware. Such materials are light and have low thermal conductivity associated to high tensile resistance. Kevlar strings are often used to ensure a high degree of sti€ness combined to small contact surfaces. Plastics (such as Te¯on, Nylon, Vespel) can also be used whenever the structural load requirements are less critical [72]. Struts and tension straps play an important role in supporting the weight of Dewars and cold stages. Orbital disconnect struts (ODS) are being investigated to limit heat losses by removing unnecessary thermal bridges once the spacecraft has reached its station orbit. A review on cryogenic structural supports and materials can be found in [73]. Supporting VLT elements add additional constraints to the supports: also minimise the heat dissipated by damping of vibrations (for instance generated by pre-cooler compressors).

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819

Multi-layer insulation (MLI) is commonly used on ground as well as in space. MLI consists of a stack of polyester (Mylar) or polyamide (Kapton) foils that are embossed, crinkled or separated by a spacing material such as a thin net. The foils are aluminised on one or both sides to reduce the radiative transfer of heat. MLI is used to protect the spacecraft from intense solar radiation as well as to insulate cold stages located inside a cryostat. As far as the radiative transfer is concerned, the use of N layers allows reducing the surface emissivity e to an e€ective value eeff by a factor proportional to N 1. In practice, losses due to holes, edges, seams impose a correction to the theoretical estimate of the e€ective emissivity eeff [74,75]. Space applications require speci®c attention to the outgassing of the satellite during testing and during its launch ascent. Very e€ective utilisation of MLI is critical to extend the mission lifetime of missions based on cryogen reservoirs (e.g. ISO, Herschel). V-groove shields are also used onboard spacecrafts, for instance to shield a low temperature radiator from warmer parts. V-groove shields are based on a few (typically <5) angled and highly re¯ective shields open to space. The principle is to reject to space radiation after a number of re¯ections between the angled shields rather than trapping it between layers (such as in MLI). V-groove shields also imply limitations on the satellite attitude control. Additional details can be found in [76]. 3.3.2. Ancillary cryogenic equipment for space Cryogenic ancillary equipment plays a critical role during ground testing and must be adapted and quali®ed for space utilisation. It is worth mentioning the key items which are required onboard spacecrafts. High thermal conductivity links (K > 1 W/K) are required to link the focal plane to the cold stage of the cooler. Such links often need to be ¯exible, to provide electrical insulation as well as damping eventual vibrations exported from the cooler to the detectors. Cryogenic heat pipes are presently being developed to guarantee high thermal conductivity with the added capability to function as thermal switches [77]. Heat switches play a critical role in several coolers including 3 He sorption coolers and ADR systems. An ideal cryogenic heat switch for space applications should have no moving parts, ensure high insulation in OFF mode and high conductance in ON mode, with negligible power dissipation, high reliability and a good level of redundancy. To date gas-gap switches (based on cryosorption) and electromechanical switches are used, but the performance and reliability levels need to be drastically improved [78]. High heat capacitance devices acting as cryogenic heat reservoirs (e.g. 4 He gas volume providing a 4.2 K heat sink when condensed) can play an important role to

815

stabilise the temperature of further stages or providing additional cooling power at absorption peaks. Space quali®ed temperature sensors are a critical item, especially onboard instruments using bolometers, which are very sensitive to temperature ¯uctuations. An overview of cryogenic thermometry is provided in [79]. Critical issues are the radiation hardening of the sensors, their calibration and associated heat load. A new generation of thermometers based on the Coulomb± Blockade e€ect has been proposed with the perspective of providing a simple primary temperature standard that could be quali®ed for space [80]. Pressure, level and ¯ow meters are required to monitor the performance of cryogenic equipment onboard spacecrafts. There are almost no such space quali®ed equipments. We should mention here the direct liquid content measurements (DLCM) applied on ISO to determine the residual cryostat lifetime during the mission [81]. Cold ®lters/windows are required to shield the focal plane from any unwanted radiation while transmitting the desired portion of the electromagnetic spectrum. Typical applications involve the rejection of any visible/ IR radiation for X-ray bolometers with the reduction of the heat load applied to the focal plane. Also in this case highly reliable elements with sharp spectral attenuation are required. Cryo-mechanisms (e.g. choppers, ®lter wheels, grating devices) also play an important role onboard demanding space science instruments and requiring a high level of reliability throughout the envisaged mission lifetime [82]. Special paints, coatings are required to achieve either very low or high emissivity on given surfaces: typical applications involve special paints used to absorb any IR radiation (i.e. high ) impinging in the proximity of the focal plane to minimise stray-light re¯ections [83]. Cryogenic cables and wiring with low electrical resistance, low thermal conductivity and low capacitance are crucial for the development of cryogenic instrumentation for Space applications. Such items are even more critical when the focal plane is operating at temperatures below 1 K. Ribbon cables have demonstrated the capability to provide in excess of 1000 lines (metal or superconductor) with a total DC resistance below 20 X and a total load on the lowest temperature stage lower than 20 lW at 0.3 K [84,85]. The development of connectors with large number of pins would be necessary to co-achieve practicable high density ribbon cryogenic cables. 4. Key technologies 4.1. Present situation and future needs On the basis of the needs highlighted in the preceding sections, a number of critical technologies have

816

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819

been identi®ed. In this paragraph we brie¯y describe the main issues involved and the envisaged development needs. Passive radiators (in the range between 100 and 40 K) play a major role on board scienti®c satellites, reducing the requirements imposed on active cooling systems. They need to guarantee high emissivity at low temperature (see the role played by micro-cavities), while the related thermal isolation technology requires improvements; design as well as testing tools are required. Active cooling systems (between 100 and 50 K) are slowly moving from the pioneering phase of the technology demonstrators to a mature commercial phase. Due to the lack of a large customer base, the existing space quali®ed coolers are very expensive (order of million EUR) and quite heavy; in addition they remain a major source of vibrations and their eciency needs to be improved. Twenty K coolers also have an important role to play as pre-cooling stages within more articulated and lower base temperature cryogenic systems. The Stirling cyclebased coolers are not easily accommodated on board spacecrafts, while Joule±Thompson coolers are less ef®cient but more ¯exible, allowing the use of radiative pre-cooling. Modularity and cooling power scalability are also important qualities to be pursued. Cryo-sorption-based systems represent an attractive alternative to be explored. Two to four K coolers should provide larger cooling power (>50 mW) in view of supporting lower temperature stages. Absence of (or low) vibration levels is also required in applications involving very low temperature coolers, sensitive detectors and high accuracy spacecraft pointing and/or positioning (e.g. astrophysical observatory). VLT coolers (T < 1 K) are becoming more and more important to space missions due to the utilisation of very sensitive cryogenic detectors. A large e€ort is required to develop closed-loop, space quali®ed coolers (such as ADR, DR, sorption coolers) providing subKelvin temperatures and o€ering reliable performance and long lifetime. Miniaturisation also represents an important trend, since it should allow reducing heat losses, power consumption and sensitivity to vibrations. New activities are aimed to verify the possibility to use micro-machining technologies to develop both active and solidstate miniature coolers. Finally ancillary equipment and devices should not be neglected. It is the case of high conductivity thermal busses (e.g. heat pipes) and connections; low thermal conductivity and orbital disconnect supports; heat switches (important also to the VLT coolers); cryo-mechanics; temperature stabilisation devices and low temperature measurement techniques.

4.2. Economic return of cryogenics in space Cryogenics can be used in space, albeit at a signi®cant cost. Any potential application is bound to address the issue of the economic return of cryogenic equipment as opposed to conventional instrumentation. A good example is set by satellite communications, commercially operated and for whom viable alternatives to cryogenic equipment are available (as opposed to the case of space science missions). Economic considerations might favour the use of superconducting components if miniaturisation can result in both mass and power consumption. Payload launch costs can be as high as 50 000 EUR/kg, while the mass to radiated power relation is about 0.2 kg/W. In order to compensate for the larger development and manufacturing costs, compact space quali®ed coolers are required, with low mass, the highest possible eciency and high reliability. In addition, HTS devices operating above 77 Kare needed. Additional economic return can be generated by increased performance, such as increased transmission capacity in a given bandwidth. Due to increased crowding in the allocated frequency bands, HTS devices might become more attractive due to their superior performance in terms of sharp separation between adjacent bands and related interference suppression [46]. 4.3. Technology road-map In Table 5 we have summarised the key areas to be explored in the future to produce signi®cant advances in the ®eld of cryogenics for space applications. The content of the table re¯ects what is discussed in the previous sections with the addition of consideration on the development timescale and on the temperature range involved by each speci®c technology. The timescale considered is limited to the next 10±20 years. The European Space Agency, within its Technology Research Programme, the General Support Technology Programme and speci®c Projects, is active in most areas indicated in the table. It is worth mentioning the potential advantages o€ered by an improved co-operation between ESA and other institutions of the European Union (e.g. in the ®eld of materials science such as advanced composites). The small size of the market of cryogenics for space applications (e.g. quali®ed mechanical coolers) has determined the high costs and has drastically reduced the number of suppliers. In order to improve the situation it is necessary to promote the maximum possible compatibility between space and ground products, thus widening the potential market base. To this end, future development activities should include the space quali®cation of cryogenic systems largely based on or derived from commercial o€ the shelf (COTS) units, originally developed for ground purposes.

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819

817

Table 5 Technology road-map for cryogenics in space Area

Critical technologies

Time (yr)

T (K)

Coolers

High eciency, low T passive radiators Improved eciency, large size Dewars for space Low vibration, high COP Stirling coolers High COP, space quali®ed PTR Space quali®ed compressors based on turbines High COP, miniaturised active coolers.

>5 >5 >10 >10 >10 >15

<60 <4 <10 na <10? <50?

VLT

Optimisation of space quali®ed ADR Development of closed-loop DR for space Space quali®cation of sorption coolers Solid-state coolers based on NIS devices Orbital disconnect supports Very low emissivity coatings Improved V-groove shields

>5 >10 >2 >10 >5 >2 >2

<0.1 <0.1 <0.5 <0.3 <10 <50 <100

Ancillary equipment

Cryogenic heat pipes for Space Cryogenic heat switches for Space IR absorbing paints Space thermometry/in-¯ight calibration Pressure/level/¯ow meters Cryo-mechanisms (e.g. ®lter wheels) Cryo-optics, large area cooled mirrors Cryogenic wiring for low amplitude signals High heat capacitance devices. Testing facilities (e.g. vibrating table at low T )

>5 >10 >10 >10 >5 >5 >5 >5 >2 >2

<10 <10 <10 <4 <4 <10 <50 <10 <10 <10

Materials

HTSs ®lms/wires High temperature superconductor devices Low temperature superconductor devices Advanced composite materials for cryogenics

>5 >5 >5 >10

>80 >8 <10 300±1

Thermal insulation

5. Conclusions Cryogenics has made remarkable progress over the last 15 years, moving from laboratory prototypes to commercial applications in several ®elds. Such a progress, coupled with the advanced performance o€ered by cryogenic and superconducting devices, has triggered a virtuous cycle of ever-growing initiatives and new applications. Reliability and simplicity of operations have opened the possibility to use cryogenics in space, albeit at the price of additional complexity and larger costs. Continuous improvements have resulted in longer lifetime and reduced risks, with a number of design solutions, from cryostat to mechanical cooler-based systems capable of covering a large range of base temperature requirements. The utilisation of cryogenic devices onboard spacecrafts, such as photon detectors, has allowed unprecedented results, especially in the ®eld of space science. Over the last 10±15 years several missions have demonstrated that these devices outperform any competing technology. HTSs have the potential to play a signi®cant role in telecommunications, providing improved performance and higher eciency. In this ®eld, managed in commercial terms, cryogenic equipment is faced with the

need to prove economically viable when compared to more conventional systems. Such a condition implies the capability to demonstrate high reliability, improved power eciency and lower mass in addition to equivalent or improved performance with respect to alternative technologies. An emerging trend is the development of complete cryogenic payloads, in which the use of cryogenic devices is extended to FEE (e.g. low noise ampli®ers and input multiplexers) and back-end electronics (e.g. output multiplexers and superconducting digital electronics), in addition to the more traditional detectors and optics applications. The development of cryogenic payloads for space calls for a system approach, thus involving, from the very beginning of the project, the complete spacecraft design. Clear examples are set by several space science observatories, which are built around their cryogens tank, or by the crucial role played by the spacecraft geometry and by mission control in the case of passively cooled instruments. Several technologies need to be improved to extend the application of cryogenics in space, including active mechanical coolers, thermal insulation techniques and the miniaturisation of equipment, with the goal of reducing heat losses and the sensitivity to vibrations. The development of high conductivity thermal busses and

818

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819

connections (e.g. heat pipes), low conductivity or disconnect supports, heat switches, temperature stabilisation units and low temperature measurement techniques is all necessary to improve further cryogenic payloads. In the previous sections we have discussed the dominant space engineering trends and the guidelines along which cryogenic technologies are expected to develop in the next 10 years. The e€ort produced by the leading space organisations shows without any doubt that cryogenics is going to play a strategic role on board future space missions.

References [1] Pobell F. (Ed.), Matter and methods at low temperatures. Berlin: Springer; 1996. [2] Kessler MF, et al. Astronom Astrophys 1996;315:L27±31. [3] Collaudin B, Passvogel T. Proc SPIE 1998;3358. [4] Ogawara Y. International Astronomical Union, Symposium No. 188. Kyoto, Japan: Kluwer Academic Publishers; 1998. p. 75±8. [5] Timbie PT, et al. Cryogenics 1990;271:30. [6] Bavdaz M, et al. Proceedings of the Second European Symposium on the Utilisation of the International Space Station. ESA SP-433, 1999, p. 621±27. [7] Tananbaum H, White NE, Bookbinder JA, Marshall FE, Cordova FA. Proc SPIE 1999;3765:62±72. [8] Cihlar J, Belward A, Govaerts Y. Meteosat Second Generation opportunities for Land Surface Research and Applications. EUMETSAT 1999 ± ISSN 1561-140X, ISBN 92-9110-031-5. [9] Phillips WA. Proceedings of ESA Workshop on Space Applications of High Tc Superconductors. ESA WPP-052, 1993, p. 219. [10] Fowler AM. Proceedings of ESA Symposium on Photon Detectors for Space Instrumentation. ESA SP-356, 1992, p. 129±35. [11] Lemke D, Klaas U, Abolins J, et al. Isophot ± capabilities and performance. Astronom Astrophys 1996;315:L64±70. [12] van de Stadt H. Proceedings of ESA Symposium on Photon Detectors for Space Instrumentation. ESA SP-356, 1992, p. 197± 200. [13] Thompson RI, Rieke M, Schneider G, Hines DC, Corbin MR. ApJ Lett 1998;492:L95. [14] Rando N, et al. NIM A 2000;444:441±4. [15] Kelley RL, et al. NIM A 2000;444:170±4. [16] Jakobsen HST, Ultrviolet-optical space astronomy beyond HST. ASP conference series, vol. 64. 1999. p. 397±404. [17] Verhoeve P. Proceedings of the WOLTE-4 Conference. ESA WPP-171, 2000, p. 205±10. [18] Clarke J. NATO ASI series, vol. F59. Berlin: Springer; 1989. p. 87±147. [19] Langford D, Simmons J, Ozawa T, Long E, Paris R. SPIE Proc 1983;445:244. [20] The COBE mission design, spacecraft, and orbit, from Bennett et al., In: Thronson HA, Shull JM, editors. The evolution of galaxies and their environment. Proceedings of Third Teton Summer School, 1992. [21] Kessler MF, et al. Astronom Astrophys 1996;315:L27±31. [22] Cornelisse JW. Classical Quantum Gravity 1996;13:A59±65. [23] Hacking P, et al. Proceedings XVI of Moriond Astrophysics Meeting on Microwave Background Anisotropies, March 1997; Moriond, France. [24] Porter FS, et al. Proceedings of LTD-8. NIM A 2000;444:179±83. [25] Tauber J. In: Bouchet FR, Gisperteds R, editors. Proceedings of Moriond Astrophysics Meeting on Extragalactic Astronomy in the Infrared, Moriond, 1997.

[26] Pilbratt GL. In: Favata F, Kaas AA, Wilson A, editors. Proceedings of the 33rd ESLAB Symposium. ESA SP-445, 2000. [27] Timbie PT, et al. Cryogenics 1990;30:271. [28] Winkler C, Hermsen W. Proceedings of the Fifth CGRO Symposium. New York: AIP; September 1999, in press. [29] Seery BD, Smith EP, Mather JC. Proceedings of 34th Liege International Astrophysics Colloquium, 15±18 June 1998; Liege, Belgium. ESA SP-429, 1998, p. 13±22. [30] Kuzmin L, et al. Proceedings of the Second European Symposium on the Utilisation of the International Space Station, 16±18 November 1998; Noordwijk, ESA SP-433, 1999, p. 127±34. [31] Ulvestad JS, et al. In: Jackson N, Davis R, editors. High sensitivity radio astronomy. Cambridge (UK): Cambridge University Press; 1997. p. 252±55. [32] Penny AJ, et al. In: Reasenberg R, editor. Astronomical interferometry. Proc. SPIE 1998;3350:666±71. [33] Beichman CA, Woolf NJ, Lindensmith CA, editors. The terrestrial planet ®nder (TPF): a NASA origins program to search for habitable planets, May 1999. JPL publication 99-003. [34] Miseno€ M, et al. Proceedings of ESA Workshop on Space Applications of High Tc Superconductors. ESA WPP-052, 1993, p. 119±27. [35] http://lasp.colorado.edu/strv/index.shtml. [36] Chui PTC, Lipa AJ, Nissen AJ, Swanson DR. Lambda point experiment in microgravity. Cryogenics 1994;34(5):341±7. [37] DiPirro MJ, et al. Cryogenics 1994;34(ICEC Suppl):267. [38] Gilman DC, Russo SC, Sugimura RS, Lessons learned during the integration phase of the NASA CSE. In: Proceedings of the Eighth International Cryocooler Conference. New York: Plenum Press; June 1995. [39] Bowman RC, et al. Brilliant eyes ten-Kelvin sorption cryocooler experiment (BETSCE). Final Report, JPL Publication 97-14, Pasadena, CA, September 1997. [40] Amundsen RM, et al. Development of the materials in devices as superconductors (MIDAS) experiment. NASA/TM-1998-208440, Langley Research Center, Virginia, May 1998. [41] Lipa JA, et al. Phys Rev Lett 1996;76:944±7. [42] An overview is provided at http://ltmpf.jpl.nasa.gov/. [43] Nash A, et al. The fast alternative cryogenic experiment testbed. Presented at 1999 Space Cryogenics Workshop, Quebec, Canada, July 1999. [44] Chaloupka H, et al. Proceedings of ESA Workshop on Space Applications of High Tc Superconductors. ESA WPP-052, 1993, p. 9±25. [45] Baumfalk A, et al. IEEE Trans Appl Supercon 1993;9:2857±61. [46] Braginski AI. IEEE Trans Appl Supercon 1999;9:2825±48. [47] Reber AC. The upper atmosphere research satellite uars. Trans Am Geophys Union 1990;71:1867. [48] Kumer BJ, Mergenthaler JL, Roche AE. CLAES CH4 , N2 O, and CCl2 F2 (F12 ) global data. Geophys Res Lett 1993;20:1239±42. [49] Mowle EW, Dennehy CJ. The Landsat 6 and 7 system ± the third generation of civil US, land remote sensing. International 42nd Astronautical Congress, 5±11 October; Montreal, Canada, 1991. [50] Bordi F, et al. Presented at the 48th Congress of the International Astronautical Federation, 6±10 October; Torino, Italy, 1997. [51] Barnes WL, et al. IEEE Trans Geosci Remote Sensing 1998;36(4):1088±100. [52] Dials AM, et al. SPIE 1998;3437:84±91. [53] Murray MJ, et al. Geophys J Res 1998;103:8153±65. [54] Aminou ADM, et al. Proceedings of SPIE. Europto series, vol. 3221, 1998, p. 19±31. [55] Envisat-1 mission & system summary. Ocial project brochure, available at http://envisat.estec.esa.nl/. [56] Endemann M, et al. ESA Bull 2000;101(February). [57] Burkett P, et al. IEEE Trans Geosci Remote Sensing 1983;GE21:345. [58] Edwards PG, Pawlak D. ESA Bull 2000;102:7±18.

B. Collaudin, N. Rando / Cryogenics 40 (2000) 797±819 [59] Dierickx B, et al. Proceedings of the 30th ESLAB Symposium on Submillimetre and Far-Infrared Space Instrumentation. ESA Publication SP-388, 1996, p. 61±65. [60] Rowell JM. IEEE Trans Appl Supercon 1999;9:2837±48. [61] Kohout LL. NASA TM 101980. Prepared for the International Conference on Space Power, June 5±7; Cleveland, OH, 1989. [62] Ross RG. Cryogenics 1990;30:233±45. [63] Jones JA, Bard S, Schember HR, Rodriguez J. Cryogenics 1990;30:239±45. [64] Jewell C, Rouille G, Rando N, Peacock A, van Dordrecht A. ESA Bull 1993;75:55±61. [65] Freund MM, Dubond L, Lange AE, Matsumoto T, Murakami H, Hirao T, Sato S. Cryogenics 1998;38:435±43. [66] Duband L, Hui L, Lange AE. Cryogenics 1990;30:263. [67] Benoit A, Pujol S. Physica B 1991;169:457±8. [68] Roach PR, Gray KE. Adv Cryog Eng 1998;33:707. [69] Hagmann C, Richards PL. Cryogenics 1995;35:303±9. [70] Nahum M, Eiles T, Martinis J. Appl Phys Lett 1994;65:3123. [71] Manninen AJ, Suoknuuti JK, Levo MM, Pekola JP. Appl Phys Lett 1999;74:3020±2. [72] Roach PR. Rev Sci Instrum 1992;63:3216±8. [73] Reed RP, Golda M. Cryogenics 1997;37:233±50. [74] Bapat SL. Cryogenics 1990;30:700±10.

819

[75] Bapat SL. Cryogenics 1990;30:711±9. [76] Bard S. Development of a high performance cryogenic radiator with V-groove radiation shields. J Spacecraft 1987;24(3). [77] Gilman DC, Russo SC, Sugimura RS. Proceedings of the Eighth International Cryocooler Conference. New York: Plenum Press; June 1995. p. 869±82. [78] Torre JP, Chanin G. Rev Sci Instrum 1984;55:213. [79] Rubin LG. Cryogenics 1997;37:341±56. [80] Kauppinen JP, Loberg KT, Manninen AJ, Pekola JP, Voutilainen RA. Rev Sci Instrum 1998;69:4166±75. [81] Seidel A, Passvogel T. SAE. Intersociety 20th Conference on Environmental Systems, July 9±12; Williamsburg, VA, 1990. [82] Lemke D, Grozinger U, Krause O, et al. Focal plane chopper for the PACS instrument aboard the far-infrared space telescope. In: Strojnik M, Andresen BF, editors. Infrared spaceborne remote sensing VII. Proc. SPIE, vol. 3759, p. 205±13. [83] Ungar S, et al. Infra-red black paint for room and cryogenic temperatures. In: Infra-red technology XV. SPIE, vol. 1157, 1989, p. 369. [84] Cunningham CR, Hastings PR, Strachan JMD. Cryogenics 1995;35:399±402. [85] Warner BA, Breon SR. Adv Cryog Eng 1994;39:1161±8.