Galileo Probe

JPL. Public domain (NASA / US government work).

Overview

A giant planet probe has to survive the highest entry speeds and heating of any planetary mission, descend through an atmosphere of unknown deep structure, and return data over a link that cannot be tracked from Earth. The Galileo probe is the only vehicle to have done this. It separated from the Galileo orbiter, coasted to Jupiter and entered on 7 December 1995 at 47.4 km/s, the highest entry speed any spacecraft has flown [1]. It returned an atmospheric profile from the top of the atmosphere down through the cloud decks, and it carried an instrumented heat shield whose in-flight recession was measured, which remains the only such dataset from a giant planet entry. The entry, descent and relay sequence it flew is still the reference case cited in general treatments of giant-planet entry-descent-landing science [9].

Specifications

Parameter	Value	Source
Diameter and height	1.27 m by 0.91 m	[19]
Mass at entry	337 +/- 4.0 kg [1]; 339 kg on the NASA program page [19]	[19]
Mass after entry	248.1 +/- 8.7 kg	[1]
Spin rate	10.5 rpm	[1]

Mission profile

Parameter	Value	Source
Carrier	Galileo orbiter, launched 18 October 1989 from NASA KSC on Atlantis, STS-34	[1], [19]
Probe release from the orbiter	12 July 1995	[19]
Entry date	7 December 1995
Entry velocity	47.4 km/s [1], 47.6 km/s on the NASA program page [19]; the highest any spacecraft has flown	[19]
Entry flight path angle	84 +/- 1 degrees
Angle of attack at entry	10 degrees
Peak heating rate	about 30 kW/cm2
Total heat load	about 300 kJ/cm2
Peak deceleration	250 g
Velocity at 450 km altitude	47.406 km/s
Velocity at 50 km altitude	0.833 km/s
Time between those points	111.02 s

Rows with no marker are from [1]. Secondary accounts differ on several of these values. The NASA chronicle dates probe release to 13 July 1995 and gives a maximum deceleration of 228 g and a transmitter failure 61.4 minutes after entry at 22.7 atmospheres [12]; the navigation record gives an entry speed of 47.6 km/s [14]. The sources do not reconcile these differences; the table keeps the reconstruction values.

Navigation and delivery

The entry flight path angle at 450 km above the 1-bar level was constrained to -8.6 +/- 1.4 degrees atmosphere-relative with 99 percent probability [14]. The delivery error was 0.15 degrees, and the reconstructed angle was accurate to 0.10 degrees against a 0.15 degree requirement. The orbiter was deflected away from the probe trajectory by a 61 m/s maneuver on 27 July 1995 and relayed the descent data to Earth, since the probe could not be tracked directly [14].

Architecture

The probe divides into a descent vehicle and an aeroshell [2]. The descent vehicle carries every science instrument and every support subsystem, telecommunications, power, control and thermal, into the atmosphere under parachute. The aeroshell, comprising a forward heat shield and an aft backshell, protects the descent vehicle through cruise, coast and entry. Both the pilot and main parachutes are disk-gap-band canopies, the same family characterized in wind-tunnel wake testing behind Viking-type aeroshells for other planetary entry vehicles [10]. On Galileo that structure had a 22.2 cm nose radius and a 44.86 degree cone half angle [1].

The aeroshell has five functions, all five as listed by [2]:

Function	Requirement
Aerodynamic stability	Remain stable through hypersonic and supersonic entry into an H2-He atmosphere while spin-stabilized about the symmetry axis
Thermal protection	Protect the descent vehicle from entry heating and thermomechanical load
Structural	Carry the deceleration load from the descent vehicle through hypersonic entry
Transition	Provide a stable transition from supersonic entry to subsonic descent
Separation	Release heat shield and backshell from the descent vehicle on a g-switch with timer backup

The deceleration the third of those functions has to carry was 250 g on Galileo [1].

Some of the probe system stays behind. The release and separation mechanism and the probe telemetry receiver remain on the carrier spacecraft [2]. The probe is entirely self sufficient once released: batteries support coast, wake-up, health checks, entry and descent, and provide autonomous thermal control during coast. A coast timer loaded before release issues the wake-up that starts the entry power-on sequence, initial warm-up, instrument and subsystem health checks and pre-entry calibrations.

Thermal protection system

The heat shield is the part of the probe with no substitute in the record, the one entry system component every ice giant probe study still has to size against [3].

Element	Material	Source
Forebody nose cap	Chopped-molded carbon phenolic	[1], [3]
Forebody frustum	Tape-wrapped carbon phenolic	[1], [3]
Aft heat shield	Phenolic nylon	[1]

Geometry	Value
Forebody thickness at centreline	14.6 cm
Minimum thickness at front of frustum	5.1 cm
Nose radius	22.2 cm
Cone half angle	44.86 degrees

Source: [1].

Recession was measured in flight by an array of ablation detectors. Total recession was 4.45 +/- 0.25 cm at the nose, fell to a minimum of 2.74 +/- 0.15 cm at mid-frustum, then rose again toward the base, at a maximum recession rate of 0.4 cm/s [1]. Of the 337 kg entry mass, 79.0 +/- 4.0 kg was lost to forebody ablation, 1.5 +/- 0.5 kg to pyrolysis and 8.4 +/- 4.2 kg from the aft shield.

The measured shape does not match the design prediction. The mid-frustum minimum is the disagreement, and it is the reason the experiment was flown at all [1]. The entry environment was underpredicted and the design thickness was barely adequate, so a future ice giant probe should be designed with more margin rather than to the Galileo precedent [3].

The manufacturing base is also gone. The carbon phenolic thermal protection system used on Galileo and Pioneer Venus is no longer available, having lost its manufacturing processes, and NASA developed the Heatshield for Extreme Entry Environment Technology as the replacement rather than attempting to revive it [2], [3]. HEEET is a 3-D woven ablator and is about 50 percent more mass efficient than the heritage carbon phenolic system [2].

Entry environment

The atmosphere the probe met was not the one it was designed against. Nominal entry assumed 89 percent H2 and 11 percent He for a molecular weight of 2.22; the measured helium mole fraction was 13.6 +/- 0.3 percent, giving 2.35 [1].

Altitude	Pressure	Temperature
450 km	7.21e-8 bar	527.8 K
250 km	1.60e-5 bar	163.2 K
100 km	7.00e-3 bar	160.6 K
50 km	7.51e-2 bar	112.4 K

Source: [1].

Before any spacecraft reached Jupiter, the planet’s radiation and atmosphere were bounded from Earth-based and flyby data alone, in design criteria such as the 1971 NASA monograph that set the encounter environment probe designers worked against [7]; the accompanying radiation design levels required hardening estimates for the electronics of any thermoelectrically powered outer-planet spacecraft [8].

That reconstruction, from the Atmospheric Structure Instrument, is what Jupiter-GRAM is built on, and the model still carries neither composition nor winds [5]. Titan-GRAM and Neptune-GRAM by contrast represent variability as a min-average-max density envelope because measurement uncertainty at those bodies is comparable to real variability [6].

Payload

The instrument set is the reference design for every giant planet probe proposed since [2]:

Instrument	Measurement
Atmospheric Structure Instrument	Temperature, pressure and density vertical structure
Mass spectrometer	Composition, high molecular mass organics, isotopic composition
Helium Abundance Detector	Helium abundance from atmospheric refractive index
Doppler Wind Experiment	Zonal wind speed and direction against altitude
Nephelometer	Cloud structure, solid and liquid particles
Net Flux Radiometer	Thermal and solar energy flux

Source: [2].

A seventh instrument was engineering rather than science: the Analog Resistance Ablation Detector array in the heat shield, which is what produced the recession record above [1].

The Helium Abundance Detector works by measuring refractive index over the 2 to 10 bar pressure range [2]. Because the jovian atmosphere is more than 99.5 percent H2 and He, that refractive index is a direct measure of the He/H2 ratio. Galileo used a Jamin-Mascart interferometer, chosen for a simple and compact design at high accuracy, with one beam through a reference gas and one through the atmosphere.

The Doppler Wind Experiment uses the probe-to-carrier radio link itself, on the assumption that a probe in terminal descent under parachute moves with the wind [2]. Both ends carry an ultrastable oscillator. Recovering wind requires accurate reconstruction of the probe location at the start of descent, from measured accelerations during entry, and of the descent speed against time, from Atmospheric Structure Instrument pressure and temperature; expected link frequencies are then differenced against measured ones and the residuals inverted. The measurement also captures probe motion from turbulence, aerodynamic buffeting, convection and waves.

Communications

The probe telecommunication system is transmit only, with two redundant channels transmitting orthogonal polarizations at slightly offset frequencies for isolation, driven by an ultrastable oscillator so the link frequency is stable enough for radio science [2]. The mission ended when that link was lost at 22 bar [4]. Because the descent-phase link began only after entry deceleration and heating had passed, Galileo avoided the plasma blackout that can sever a UHF or X-band entry link near peak heating on other vehicles [11].

Frequency selection is set by atmospheric microwave opacity rather than by link budget convenience. Opacity from trace absorbers such as H2O, NH3, H2S and PH3 rises roughly as the square of frequency, which drives the choice as low as practical, often UHF [2]. At Jupiter the lowest practical frequency is L-band, because below that the synchrotron radiation environment is too intense. The choice then propagates into transmit antenna type, size, gain and beamwidth, and into pointing requirements for the receiving antenna on the carrier.

The atmosphere below the cloud tops remains largely unknown in thermal, compositional and dynamical structure, so a difference between the model atmosphere and the real one degrades relay performance directly and has to be carried in the frequency choice [2]. The Galileo reconstruction is the only in-situ constraint available for that model [1].

Relay geometry is a mission design constraint, not a detail. At a giant planet the probe mission competes with orbit insertion: insertion is most propellant efficient close to the planet, but that trajectory combined with a shallow probe entry angle that keeps heating and inertial loads low yields an impractically short relay window [2]. Delivering the probe to an aim point about 180 degrees from the orbiter’s, at a small cost in orbiter delta-v, allows a moderate entry angle and a relay window of up to 2 hours.

Results and their limits

The probe descended into a 5 micrometer hot spot near the boundary between the Equatorial Zone and the North Equatorial Belt, a meteorologically anomalous site [4]. Water was subsolar and still increasing at 22 bar, where the radio signal was lost. Nitrogen and sulfur were about 3 times solar at about 10 bar [4].

The consequence, general to giant-planet entry science, is that a single entry does not characterize a planet [9]. Juno microwave radiometry of the equatorial zone later gave water at 2.5e3 ppm, 2.7 +/- 1.7 times protosolar, against a probe mass spectrometer estimate at the lower limit of its range, 566 +/- 216 ppm against a signal-derived 700 +/- 100 ppm [4]. Any model of Jupiter’s envelope enrichment built on the probe value alone rests on one location.

The mass spectrometer measured composition over the descent, down to 21.1 bar [18]. It returned a 4He/H2 mixing ratio of 0.157 +/- 0.030, CH4/H2 of 2.1e-3 +/- 0.4e-3, 13C/12C of 0.0108 +/- 0.0005, D/H of 2.6e-5 +/- 0.7e-5 and H2S of 7.7e-5 +/- 0.5e-5, with H2O at 5.6e-5 +/- 2.5e-5 and only upper limits of 3e-5 for neon and 2.3e-3 for ammonia [18]. H2O and H2S rose with depth through the descent, which is why the water value is a lower bound on the deep abundance.

The elemental ratios the probe did establish are the anchor for successor payloads: He/H 7.85e-2 +/- 0.16e-2, C/H 1.19e-3 +/- 0.29e-3, N/H 3.32e-4 +/- 1.27e-4, O/H 2.45e-4 +/- 0.80e-4, S/H 4.45e-5 +/- 1.05e-5 and Ar/H 9.10e-6 +/- 1.80e-6 [2].

Computing

The probe flight computer was a dual microprocessor system whose function was largely confined to sequencing, with an architecture similar to the Mariner, Viking and Voyager systems. The probe mission was managed by NASA Ames Research Center and the entry probe was assembled by Hughes [13]. The orbiter, by contrast, was a distributed multi-computer system whose software was judged more complex than Apollo’s because of the number of intercommunicating machines [13].

Other entry probes and what has not flown

The design lineage is Pioneer Venus: the Galileo entry probe was based on the large probe of the Pioneer Venus multiprobe [12]. The only other atmospheric entry probe to descend through a thick outer-solar-system atmosphere is Huygens, which entered Titan at about 6 km/s, deployed its main parachute at 155 km altitude and landed at 5 m/s on a 93.7 K, 1470 mbar surface [16]. Its descent winds were rebuilt from Earth-based Doppler tracking of the probe signal after the on-board wind experiment data were lost, giving a zonal profile from about 140 km to the surface [20]. Its entry speed is an order of magnitude below Galileo’s, so it validates the descent and relay sequence and not the heat shield. Dragonfly’s entry team later reran the Huygens descent in a six degree of freedom simulation to validate its Titan atmosphere model [17].

Probes to Saturn, Uranus and Neptune have been proposed and none has flown. The 2011 decadal survey ranked a Uranus orbiter with a small atmospheric probe as a flagship priority, and its cost evaluation found that much of the mission risk could be retired by preparing for probe entry [15]. Its cost figures are pre-decisional and the mission was not flown as described [15]. Sizing for those entries still starts from the Galileo heat shield record [3].

What is being tried now

Titan entry, descent and landing (flight in preparation). Dragonfly is the next outer-solar-system entry system to fly. Its concept of operations has a 4.5 m aeroshell entering at 7.41 km/s at -46.8 degrees, peak deceleration limited to 11 g, an 8.25 m disk-gap-band drogue, about 100 minutes of drogue descent, and a 16.67 m ringslot main parachute deployed at 5 km pressure altitude [28]. Every value is a design or simulation output made before integration and test [28]. The rotorcraft concept itself rests on analysis of a 5.4 kg/m3, 94 K atmosphere, with no hardware test at Titan conditions [31]. The modeling chain is validated against Huygens: the 2025 overview covers the trajectory tools, atmosphere models and Monte Carlo margins [29], and a multi-body parachute model reproduces Huygens flight accelerations at pilot and main deployment and at heat shield jettison, attributing nadir angle excursions to zonal wind profile slope changes rather than to the parachute model [27]. The Huygens design record lists the three-parachute control chain and a 1995 balloon drop test of the deployment sequence, but is written in the future tense and omits flight outcomes [25]. Huygens Doppler and tilt data bound the Titan wind environment: the strongest large-scale shear was about 5 m/s/km, below the 16 m/s/km preflight limit, from one descent at one location and season [26].

Engineering atmosphere models (ground, in use). The NASA Global Reference Atmospheric Model family supplies entry designers with density, temperature and pressure profiles. Titan-GRAM and Neptune-GRAM trace to Voyager-era radio science and spectrometer retrievals [23], [22]. Uranus-GRAM is built from three Voyager 2 papers, has no wind model because Uranus winds are unknown, and does not capture seasonal variation of the Voyager-era profiles [21]. Jupiter is the only one of these bodies with an in-situ constraint, and it is the Galileo probe [5].

Ice giant navigation and aerocapture (simulation only). A covariance analysis of onboard optical navigation using satellites as beacons gives Uranus tour position knowledge of 5 to 100 km 1-sigma and Neptune 1 to 20 km, but only if current satellite ephemeris uncertainty improves by a factor of 100; at present uncertainties the scenarios are not navigable autonomously [24]. A Uranus aerocapture study with Monte Carlo navigation reports onboard optical navigation delivering entry flight path angle to 0.013 degrees 3-sigma against 0.020 degrees for a ground-based schedule, in simulation of a concept that does not exist, and aerocapture has not been demonstrated at any body [30]. The Galileo entry angle accuracy of 0.10 degrees [14] is the flown comparison.

Heat shields (ground development). HEEET replaces the lost carbon phenolic process and is about 50 percent more mass efficient [2], with the Galileo recession margin as the design lesson [3].

Technologies developed

The probe established the entry, descent and deployment sequence that every proposed giant planet probe still uses: coast timer wake-up, hypersonic entry on a spin-stabilized aeroshell, pilot chute above the tropopause pulling the main chute, heat shield release, and relay to a carrier spacecraft under parachute [2]. It established that a carbon phenolic heat shield can survive a 47.4 km/s entry, and simultaneously that the margin was thinner than intended [1], [3].

It also produced the only in-flight ablation dataset at these conditions [1].

What is not established

The helium and elemental ratios, the winds and the cloud structure come from one descent at one site, a meteorologically anomalous hot spot, and the probe’s data end at 22 bar [4]. Water at depth, and the global enrichment of the envelope, are not constrained by the probe, and the Juno retrieval that addresses them covers one latitude band and is conditional on a model with parameters the data do not constrain [4]. The two probe-era ammonia estimates differ by about a factor of two and the sources do not resolve which is correct [4]. Jupiter-GRAM has no wind, perturbation or constituent model because the probe data do not support one [5]. Entry speed, release date and peak deceleration differ between the reconstruction, the mission page and the NASA chronicle, and the sources do not settle the difference [1], [19], [12]. No ice giant probe has flown, so the carbon phenolic margin lesson has not been tested against a second entry [3], [15].