The Huygens Surface Science Package
J. C. Zarnecki, ' M. Banaszkiewicz,5 M. Bannister2 W. V. Boynton6 P. Challenor7 B. Clark,'
P. M. Daniell, ' J. Delderfield,2 M. A. English, ' M. Fulchignoni,' J. R. C. Garry, ' J. E. Geake,3
S. F. Green, l B. Hathi, ' S. Jaroslawski,' M. R. Leese, ' R. D. Lorenz, �e3 6 J. A. M. McDonnell, '
N. Merryweather-Clarke, 2 C. S. Mi1L3 R. J. Miller, l G. Newton,' D. J. Parker,2 P. Rabbetts,'
H. Svedhem,4 R. F. Turner' & M. J. Wright'
�eUnit for Space Sciences and Astrophysics, Physics Laboratory, University of Kent at Canterbury,
Canterbury, Kent CT2 7VR, United Kingdom
E-mail: j.c.zarnecki@ukc.ac.uk Far: +44 1227 762616
2Rutherford Appleton Laboratory, Chilton, Didcot, Oxon OX11 OQX, United Kingdom
3University of Manchester Institute of Science and Technology, Sackville Road,
Manchester M60 lQD, United Kingdom
4ESA Space Science Department, ESTEC, Postbus 299, NL-2200 AG Noordwijk, The Netherlands
5Polish Academy of Sciences, Space Research Centre, ul. Bartycka 184, 00-716 Warsaw, Poland
6Lunar and Planetary Laboratory, The University of Arizona, Tucson, AZ 85751, USA
?James Rennell Research Centre, Gamma House, Chilworth, Southampton,
Hampshire SO15 7NS, United Kingdom
�eMartin Marietta Aerospace, Denver Aerospace, PO Box 179 (MS-0560), Denver, Colorado 80201, USA
90bservatoire de Paris Meudon, Departement de Recherches Spatiales, 5 Place Jules Janssen,
F-92195 Meudon Cedex, France
The Surface Science Package (SSP) consists of nine independent subsystems with
the primary aim of characterising Titan's surface at the end of Huygens' descent
through Titan's atmosphere in late 2004. In addition, many useful atmospheric
measurements will be made during the descent phase. Seven of the subsystems are
mounted inside or on the lower rim of a cavity in the Probe's foredome, and are
thus exposed to Titan's atmosphere and surface material. Of these subsystems, the
penetrometer is mounted in front of the Probe in order for it to be the first point
of contact with the surface, and the acoustic sounder is mounted pointing
vertically downwards to view Titan's surface. Two subsystems, which do not
require exposure to the atmosphere or surface, are mounted on the electronics box
on the experiment platform upper surface. The system is controlled and its data
processed by a microprocessor-based system that allows the varied data types to
be transmitted in packet format on the Probe's dual-redundant interface and
transmitters. The instrument weighs 4.2 kg and consumes approximately 10 W of
power during descent.
177
SP-1177
1. Scientific Objectives The Surface Science Package (SSP) consists of a suite of instruments to conduct in situ
measurements of Titan's surface after Huygens' impact. In particular, the package will:
l determine the physical nature and condition of Titan's surface at the landing
site
l determine the abundances of the major ocean constituents, placing bounds
on atmospheric and ocean evolution
l measure the thermal, optical, acoustic and electrical properties and density
of any ocean, providing data to validate physical and chemical models
l determine wave properties and ocean/atmosphere interactions
l provide ground truth for interpretating the large-scale Orbiter Radar Mapper
and other experimental data.
The SSP will also conduct measurements in Titan's atmosphere on a best-effort
basis. The thermal properties, acoustic velocity and electrical permittivity subsystems
may provide useful science measurements of various atmospheric properties.
The performance of the SSP on Titan's surface is biased towards a liquid landing but
will provide useful data for the full range of possible surface states. Each sensor can
determine its own designated property to high accuracy, resulting in a physical
description of any surface liquid. The thermal conductivity, density, speed of sound,
refractive index and electrical permittivity are all basic physical properties that can
help to constrain the surface composition. Beyond the combined scientific return, each
measurement will yield a simple property of the surface of another world. The
accuracy of the various sensors and their ability to determine the composition of a Titan
ocean is discussed in detail in Section 4.
It is the accelerometers and penetrometer that will return the most interesting data
on impact with a solid surface. Analysis of the penetrometer output will yield a
measure of the granularity of the surface regolith or, in the case of a cohesive surface.
the resistance to penetration. If the surface is slushy or fluffy, the accelerometer (which
has a lower sample rate than the penetrometer) can furnish an idea of the surface
cohesivity through the deceleration profile. The tilt sensors will also return the Probe's
final attitude.
For both the liquid and solid surface scenarios, the acoustic sounder will provide
an idea of surface topography on a scale intermediate to that provided by analysis of
the Probe's radar altimeter data or from Orbiter remote sensing.
Huygens' survival, and by implication that of the SSP, cannot be guaranteed because
of the surface's unknown nature, but it is a design aim. Experience from other missions
and analysis of Huygens impact dynamics (Lorenz, 1994b) and the Probe drop test
(Jakel et al., 1996) suggest that survival is very likely for most surface types.
2. System Design 2.1 General
The SSP addresses these science objectives through a suite of nine measurement
subsystems selected to provide at least a partial characterisation of either a solid or
liquid surface. Seven require intimate contact with the surface and are housed in a
100 x 100 mm square cross-section cavity, known as the Top Hat, cut out of the Probe's
foredome and extending to the main experiment platform (Fig. 1). Venting of this Top
Hat is achieved through a 5.5 mm inner diameter tube from the Top Hat to the Probe's
top surface. The Top Hat sensors are summarised in Table 1, and those mounted on
the electronics box in Table 2. The resources allocated to SSP are summarised in
Table 3.
178
Scientific Instruments
API-S
Table 1. Top Hat measurement subsystems.
External Accelerometer
Acoustic Properties
- Velocity of Sound
ACC-E
API-V
Acoustic Properties - Sounder API-S
Density DEN
Permittivity PER
Refractometer REF
Thermal Properties THP
Piezoelectric force transducer to characterise impact
Piezoelectric transducers for the determination of the
velocity of sound in Titan's atmosphere and any
liquid surface
Sonar-type sounder. Atmospheric and surface
sounding, depth of any surface liquid
Archimedes flotation sensor to determine density of a
liquid surface
Parallel plate sensor to determine electrical
permittivity of atmosphere and any liquid surface.
resistance measurement to check for polarised
particles in liquid surface
Linear critical angle refractometer to determine the
refractive index of a liquid surface or any
condensates during descent
Transient hot-wire sensor to determine the
temperature and thermal conductivity of Titan's
atmosphere and surface
Fig. 1. Surface Science Package layout.
Table 3. SSP allocated experiment resources.
Table 2. Electronics box measurement subsystems.
Subsystem ACRJIZWI Description
Internal Accelerometer ACC-I Commercial piezoelectric transducer (single axis)
Tilt TIL Commercial electrolytic tilt sensor (two axis)
Mass
Power
Data Rate
4.2 kg
IO w
63 bit/s during Mode 1
189 bit/s during Modes 2-3
696 bit/s during Modes 4-6
179
SP-1177
2.2 Mechanical structure
The SSP Top Hat structure consists mainly of components manufactured from
advanced composite material - a glass-reinforced epoxy resin, selected for its thermal
conductivity characteristics. The structure consists of a box-like section (Top Hat)
open at both ends, a support cone to attach the structure to the Probe, and a metallic
funnel to attach the main structure to the vent tube. All these components are
adhesively bonded (and mechanically fastened). The structure's design allows the
ingress of Titan's atmosphere and ocean into the Top Hat.
The sensors are attached to the Top Hat using threaded inserts adhesively bonded
to the structure. The sensors are electrically connected to the electronics box through
a copper flying lead harness. The whole of the structure is surrounded by Basotect
thermal insulation foam through which the harness is routed. The foam is formed into
machined blocks that fit together like a three-dimensional jigsaw puzzle around the
harness. New methods were developed to machine the foam into blocks. The foam is
surrounded by a fine Kapton mesh to prevent the shedding of particles during vibration,
and this is covered with aluminised mylar to provide a suitable thermal interface.
A seal is effected at the interface between the Top Hat and the Probe's foredome.
Below this interface is the SSP Electromagnetic Compatibility (EMC) screen. This is
a lightweight metallic mesh providing a continuation of the Probe's Faraday cage and
EMC protection for the instrument, while allowing access to the Titan's environment.
The support cone provides the main interface to the Probe's lower experiment
platform. The cone is perforated for mass reduction and also to provide a low
conductance thermal path. There is a simple mechanical interface between the vent
tube exhaust and the Probe's top platform.
The electronics box (SSPE) is of conventional design and provides the accommodation
for most of the SSP electronics. The seven printed circuit boards (PCBs) are
supported in card guides with a motherboard backplane. The box is finished in Chemglaze
2306 black thermal control paint. The TIL sensor is on top of SSPE and the
ACC-I sensor is on an extension to one of the mounting feet.
2.3 Electronics
The SSPE is a 7-board electronics box. The PCBs are allocated as: power supply.
Probe interface, processor, redundant Probe interface, acoustic sensors electronics,
high resolution measurement electronics and low resolution measurement electronics.
The box size is kept to a minimum, given the fixed processor board size, by direct
connection of the sensor and Command and Data Management System (CDMS)
harness to PCB-mounted D connectors at the front of the box. Each board can be
individually removed from the box.
The processor (a Matra Marconi Low Power Processor) interfaces to the front-end
electronics by a redundant data routing bus. This is configured to provide maximum
reliability, together with the provision of redundant elements on the front-end boards
where possible. The component count is reduced as much as possible by the use of
the processor to provide clocking for the sensor electronics. The SSP realtime clock
counter (resolution 1 ms) is synchronised to the Probe's broadcast pulse. The processor
detects the impact time and signature, and switches data collection modes appropriately.
It provides data storage and compression for several sensors, particularly for the
impact event. The mode selection algorithm will also use the broadcast data (time and
altitude) during the descent phase.
The electrical design also takes into account the need to provide calibration and/or
test stimuli to simplify ground testing and in-flight checkout. Modular DC-DC
converters are used to provide redundant 12 V and 5 V supplies from the Probe's
+28 V power supply. Filter units are included to reduce electromagnetic emissions
and susceptibility. The SSP electrical system is shown in Fig. 2.
180
Scientific Instruments
EXTERNAL
ACCELEROMETER
8 BIT ADC 512x8 FIFO I,
,
CDMU B INTERFACE
P 1I - T
IN I tlWAI;t
INTERNAL
ACCELEROMETER ' III
2.4 Thermal design Fig. 2. SSP electrical system.
Mathematical thermal models have been developed to predict temperatures and heat
fluxes during the cruise and descent mission phases. During cruise, the thermal
boundary conditions are assumed to remain constant (stationary) in time. However,
during descent, the experiment's thermal environment changes significantly as the
Probe moves from the Orbiter's controlled environment to Titan's atmosphere. That
cold atmosphere gives rise to unusual problems in the thermal design and analysis of
the descent phase. The descent's most significant thermal constraints are the mean and
maximum heat loss rate from the experiment to Titan's atmosphere, which must be
minimised. To achieve this, the Top Hat and vent tube are surrounded by Basotect
foam, the thickness optimised to minimise the heat leak. As the Top Hat's profile is
square and the foam's outer profile is circular, a geometric shape factor for heat
transfer through the foam was calculated using a 3D finite element model. A wiring
harness is routed through the foam connecting the Top Hat sensors to the SSPE
electronics box. As the sensors experience cryogenic temperatures of - 80K and the
electronics box remains at about 290K, the heat leak down the harness must be
carefully controlled. To facilitate this, the harness cables were originally proposed to
181
SP-1177
be manufactured from cupro-nickel (60% Cu/40% Ni) and copper. However, due to
financial constraints and technical difficulties involved with the use of cupro-nickel
(constantan), an all-copper harness (with the cross-sectional area of the harness wires
minimised) was adopted for the flight model. The effective thermal length of the
harness mass is determined from finite element models of the harness embedded in
the foam.
Although there are seals to minimise the rate that cold atmospheric gas enters the
warm experiment interior, it must be assumed that cold gas comes into contact with
the foam's warm surface, and hence conductive heat transfer from the experiment
interior to Titan's atmosphere will occur. This conductive link is large and increases
during the descent.
The foam surrounding the Top Hat cannot be installed in one piece. Instead, many
blocks of foam are glued together around the Top Hat structure. The presence of cracks
and voids in the foam will enhance the heat leak from the Top Hat to the atmosphere
by convection. Detailed modelling of this problem suggests that the heat transfer
through the foam is increased by approximately 40%.
3. Measurement 3.1 ACC-E
Subsystems A penetrometer, based on a piezoelectric force transducer, is used to measure the
small-scale mechanical properties of Titan's surface. The sensor, described more fully
in Lorenz et al. (1994), carries the historical designation �eACC-E' (Accelerometer-
External) since it was originally envisaged (Zarnecki, 1992) to be based on an
accelerometer. However, measurement of force directly eliminates most of the effect
of the structural response of the Top Hat and Probe, which would compromise the use
of an accelerometer for this measurement.
The sensor measures the force exerted by the surface material on a small (14 mm
diameter) titanium alloy hemisphere attached ahead of the SSP Top Hat by a short
(5 cm) pylon. Forces in the range 5-2000 N are recorded at a sampling rate of 10 kHz,
giving a depth resolution of 1 mm at the nominal (- 5 m/s) impact speed. The force
is measured by a disc of piezoelectric ceramic PZT-SA, which generates a charge
proportional to the applied force. PZT-5A was chosen for its good temperature and
ageing stability.
The charge generated by the transducer is dumped into a capacitor (to reduce the
voltages present in the circuit, which could otherwise be high) and the voltage across
the capacitor amplified by a pseudo-logarithmic amplifier (to reduce the data volume
requirement). The digitised (8 bits) output is stored in a circular First In First Out
(FIFO) buffer of 512 samples (corresponding to -50 ms, or 25 cm depth) which is
frozen 6.4 ms after a preset voltage threshold is exceeded.
As the charge generated by the transducer in response to a constant force leaks away,
it is difficult to calibrate the transducer simply by applying static loads (e.g. by
attaching weights). A known, but varying, force must be applied. This is achieved with
a small pendulum, which impacts a block of material (synthetic rubber, or for larger
forces, PTFE) on to the transducer at a known speed, corresponding to the drop height
of the pendulum. This calibration method produces a force pulse with a width typically
of a few milliseconds; matching the observed voltage pulse with computer-generated
simulations allows the uncertainty in the Young's Modulus of the impacting material
to be reduced, and allows the sensor to be calibrated to an accuracy of -5% (Lorenz,
1994a). For flight calibration, a pendulum incorporating a reference force transducer
has been developed, which should further improve the calibration accuracy.
Investigations have indicated that the transducer's sensitivity drops (principally due
to change in the dielectric constant of the PZT material) by a factor of -1.8 at liquid
182
Scientific Instruments
nitrogen temperatures (77K, close to the operating temperature at Titan's surface of
- 94K) compared with room temperature operation.
Interpretation of the force profiles for the impact with Titan is based on comparison
with impacts on to representative planetary surface materials in the laboratory. The
sensor was attached to a set of weights and dropped from a height of - 1.3 m (giving
impact speed at 1 g of -5 m/s; the measurement is relatively insensitive to small
changes in impact angle or speed) into a receptacle of material (typically sand, gravel
or clay), and the signal recorded. The mass of weights used (typically - 5 kg) was not
important - it matters only that the transducer/weight ensemble has sufficient inertia
not to be slowed appreciably during the first -5 cm of penetration.
Some sample force profiles are shown in Fig. 3. Sand exhibits an increase in
mechanical resistance with depth, while clay has an approximately constant resistance.
Granular materials (gravel with mean particle diameter of - 8 mm and - 15 mm are
shown) give spikey signatures, with both the amplitude and spacing of the spikes
correlating with particle size. If the particle size of the surface material is fairly
uniform (as for fluvially-sorted sediments, for example), then it may be possible using
both peak spacing and amplitude information to discriminate between low density (e.g.
ice) and high density (rock) particle materials (Lorenz, 1994a).
3.2 ACC-I
The ACC-I (Accelerometer-Internal) subsystem comprises an Endevco 2271AM20
piezoelectric accelerometer, mounted on the SSP electronics box. Its prime function
is to record the Probe's deceleration history at impact, at a 500 Hz sampling rate, to
infer the mechanical properties of the surface material.
The sensor is a simple, commercial device. Its charge output is proportional to the
acceleration, and is simply measured and digitised at 12 bits resolution. Like ACC-E,
therefore, it is not DC-sensitive and is unable to monitor slowly-varying accelerations
(e.g. during entry). However, it is able to measure much higher g-loads than the
Huygens Atmospheric Structure Instrument (HASI) accelerometers, which are
optimised for higher sensitivity for atmospheric measurements and can measure
accelerations only up to 20 g. The measurement range of ACC-I, up to 100 g, spans
100
00.
Setnpk Number Sample Number
0
0 100 200 300
0
400 0 100 200 300 400
Sample Number Sample Number
Fig. 3. Examples of experimental ACC-E
profiles, with the ordinate in arbitrary units of
time (the full range corresponds to about
0.5 s). The abscissa is in arbitrary units of
force. Note that different units are used in
profiles (a) to (d). Materials are (a) sand, (b)
clay, (c) fine gravel (particle size about 8 mm)
and (d) medium grade gravel (particle size
about 15 mm). Adapted from Lorenz (l994a).
183
SP-1177
all impacts that are likely to be survivable. Like ACC-E, the transducer output is stored
in a circular buffer, frozen when the integrated deceleration exceeds a preset threshold.
A significant complication for impact accelerometry on Huygens is that the Probe is
not infinitely rigid, but itself deforms on impact. Lorenz (1994b) performed computer
simulations of Probe impact with various solid surfaces, taking the Probe's finite
strength into account with a many-element non-linear spring model.
For soft materials (e.g. fluffy aerosol deposits, perhaps resembling terrestrial snow),
the decelerations are of the order of 10 g, and are not sensitive to the Probe's rigidity.
Harder materials, such as regolith or sand, give loads of 25-30 g. The deceleration
histories (and peak values) are somewhat affected by the Probe's deformation. Surface
materials that are harder still will give higher decelerations, but may lead to
mechanical failure of the Probe before meaningful data at high-g could be collected.
As discussed in Lorenz (1994b), the variation of mechanical strength with depth may
be estimated by examination of the - 50 ms-long deceleration profile. It seems
possible to discriminate cohesive (sludgy or wet/sticky) materials from non-cohesive
materials. Impact into a hydrocarbon ocean would give splashdown loads peaking at
about 10 g, although the rise time for the load is much faster than for a fluffy material
giving a similar peak load. Thus, plotting peak load against rise time allows a quicklook
identification of the surface material (Lorenz, 1994a).
3.3 API
The primary scientific objectives of the Acoustic Properties Instrument (API)
subsystem are to:
- measure the speed of sound in the atmosphere as a function of altitude.
From the speed of sound the mean molecular weight can be inferred
- measure the surface topography for a solid or liquid surface
- measure the depth of any surface ocean
- measure the speed of sound in the liquid surface (if relevant)
- measure the Probe's lateral speed, which will be strongly correlated with the
horizontal wind speed.
The secondary objectives are to:
- measure the true Probe altitude over the descent's final 100 m
- measure the direction of the lateral velocity.
Also of interest but not yet proven to be feasible are to:
- provide information on precipitation (snow, rain or hail) and turbulence in
the atmosphere below 50 km
- measure the surface acoustic impedance
- measure liquid acoustic impedance.
The API is divided into two parts. One (API-V) will measure the speed of sound in
the atmosphere during the descent and in the liquid in the case of landing in an ocean
or lake. The other (API-S) consists of an acoustic sounder/altimeter.
The speed of sound, c, in an ideal gas is given by
(1)
where y is the ratio of the specific heats, R is the molar gas constant, T is the
temperature and M is the mean molecular mass.
The temperature can be accurately measured by other instruments and thus the
184
Scientific Instruments
molar mass can be calculated from the speed of sound measurements. Using this, for
an atmosphere composed of only two major gases, their mixing ratio can be estimated
to an accuracy of better than 1%. For three or four major constituent gases, limits on
their mixing ratios can be set.
An atmospheric profile for the speed of sound based on the present nominal Titan
atmospheric model is shown in Fig. 4.
The acoustic sounder may provide information on precipitation (snow, rain or hail)
and turbulence in the atmosphere below 50 km. The acoustic energy returned by rain
or hail is proportional to the drop diameter to the sixth power and to the number of
drops in the pulse volume. The pulse volume is the physical volume that in a certain
instant carries the acoustic energy.
The sounder will give information on surface roughness and acoustic back-scattering
properties of Titan's surface during the last few hundred metres. The surface return
signal is a function of the surface acoustic impedance, topography and incident angle.
The topography of solid surfaces cannot easily be described analytically, so statistical
models have to be used. A statistical model can be characterised by an r.m.s. surface
deviation from the average level and a correlation length for these deviations. Liquid
surfaces can, in some cases, be modelled by a simple wave pattern where the wave
height and wave length are the characteristic parameters. Acoustic and electromagnetic
wave propagation and interaction at interfaces show many similarities. This enables the
use of the same equations, models and databases, after minor modifications, to be
applied to both cases. The surface roughness measurements complement data from the
Huygens radar in range and resolution. The API-S provides higher resolution but is
active only over a limited area around the landing site (= 1000 m2, assuming first
contact at 100 m altitude).
If the Probe lands in an ocean, the depth (up to 1 km) will be measured. The
maximum depth that API-S can sound depends on the composition of the liquid.
Titan is particularly well suited for acoustic experiments at high frequency because
of the lack of water vapour and the low temperature. This keeps the atmospheric
attenuation to a minimum compared to Earth's atmosphere. In addition, the higher
surface pressure and low temperature increase the atmosphere's acoustic impedance,
which allows the use of small, simple sensors.
200
180
160
-
1
140
v 120
$ 100
2 80
60
40
20
n
�gi60 is0 ioo - i20 i40 i60 i80
Speed of sound (m S1)
Fig. 4. Nominal profile of speed of sound in
Titan's atmosphere.
185
SP-1177
The speed of sound is measured by a short ultrasonic pulse transmitted and received
by two resonant piezoelectric transducers and an accurate timing circuit that can
determine the delay over a fixed distance. The transducers have been electrically and
mechanically tuned to optimise bandwidth and sensitivity. The pulse carrier frequency
is 1 MHz and the pulse length is about 5 ƒÊs. The time delay is measured for the pulse
in opposite directions to compensate for any motion of the gas with respect to the
transducers, caused by the motion of Huygens through the atmosphere. The data are
sampled every 1 s, resulting in a spacing of the data points of about 10 m. The API-V
is operated in the same mode throughout the mission.
The sounder consists of an array of 10 resonant piezoelectric plates that form an
acoustic beam with a full beamwidth of about 20�‹. A short 15 kHz pulse is transmitted
and received by the same transducer and the return echo is sampled at 1 kHz and preprocessed
onboard. The sounder operates in two modes: an atmospheric mode above
7 km, and a proximity mode below. The proximity mode is also used in the ocean but
with a modified beam acquisition scheme.
In the atmospheric mode, the pulse length is 10 ms and the data from within 50 m
below the Probe is integrated, binned and stored. Data are sampled once every 2 s.
In the proximity mode, the pulse length is initially 10 ms, but between 1-0 km
altitude (given by descent data broadcast, DDB) is reduced to 2 ms (near-surface
mode). The data points around the average surface distance are sampled at maximum
resolution and data points further away from this point are binned together. The shorter
pulse length gives improved resolution in distance at the expense of reduced sensitivity.
This is not a problem because, at close distance, the signal is strong enough. Data are
sampled every 3 s.
The API-V and API-S flight model transducers are shown in Fig. 5.
3.4 DEN
The density sensor of the SSP applies Archimedes' principle to measure the density
of any Titan ocean. A small buoyant body is attached to the SSP's wall by a flexible
beam. When immersed in a fluid, its upthrust bends the beam. This bending moment
induces strain on the beam surfaces, which is then measured using strain-gauges. The
basic principle is shown in Fig. 6. (The actual float used is cylindrical, and is seen
here in cross-section).
Fig. 5. API-V and API-S flight model
transducers.
186
Scientific Instruments
F - Bouyant Force
Low density cylindrical float
Strain induced on surface of beam
I Rigid supporting wall
In order to satisfy the constraints of vibration, mass and science return, the beam
was cast from epoxy resin. This material allows low mass, high flexibility and high
mechanical strength. Torsion effects, due to uneven fluid flow (for example) were
ameliorated by splitting the beam into two parts, one at each end of the float. This also
aids in reducing torsional mode oscillations. As the sensor output relies upon the
effects of the buoyant force, which is proportional to the difference in density of the
float and the liquid, the float is cast from a low density sealed cell foam.
Four strain gauges are bonded to the beam surfaces, two on the upper faces, and two
on the lower. These are connected in a Wheatstone bridge circuit, which not only
allows a greater sensitivity of measurement, but also a degree of temperature
compensation. Implementation of the wiring is in the supporting plate for the beam,
which is directly behind the sensor. Output from the sensor was calculated, and
measured, to be linear with the density of the liquid in which the sensor is immersed.
The resolution of the final instrument meets the science requirement of &2 kg/m3 (at
a density of 550 kg/m3 and a measurement range exceeding the 400-700 kg/m3 range,
calculated for pure methane to pure ethane at Titan surface temperatures). This was
specified to give the ethane mole fraction of binary methane to ethane mix of the
proposed ocean to within 1%.
Further work (English, 1995) has shown that if nitrogen is included as a solute in
liquid methane (Kouvaris & Flasar, 1991), and propane as a second major photolytic
product (following Raulin, 1987), then a measure of density will still give the ethane
mole fraction (English, 1995). The presence of argon (recently limited in the
atmosphere to <O.l by mixing ratio by Strobe1 et al., 1993) in the ocean may add a
further free parameter (due to its non-detection) in any ocean density model. Fig. 7
shows how ocean density is highly argon-dependent. The abundance of atmospheric
argon (as measured by the Gas Chromatograph and Mass Spectrometer, GCMS, for
example) would constrain this oceanic component.
The resolution of the flight DEN at cryogenic temperatures is +l kg/m3, which
allows the ethane mole fraction to be determined to better than 0.01, once the presence
of argon has been quantified.
The development to flight model included a study of the effect of radiation (to
40 krads) and temperature effects on the properties of the materials used in its
construction (English, 1995). It was found that radiation had no measurable effect. The
temperature effects were explicitly included in DEN's design to arrive at the final
density resolution. Though the sensor cannot measure the atmospheric density, it is
activated from SSP switch-on to end of mission, in order to monitor temperature effects
on the sensor offset voltage.
Fig. 6. The basic layout of the Density sensor.
187
Fig. 7. Density against ethane mole fraction for
several Titan ocean compositions.
- 92SK Solutes + Nitrogen
..�e..�e....1. OlK Solutes + Nitrogen + Argon
650 ; ---- 92.5K Tertiary mix
,..�h,_ ...-
.. .�h,. .�e.
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Ethane mole fraction
3.5 PER
The prime function of the PER sensor subsystem is to measure the static permittivity
of Titan's surface to an accuracy of hO.003. This measurement corresponds to
determining the mixing ratio of ethane to methane to 1%. In addition, the sensor
performs a resistance measurement that will investigate the presence of polar
molecules that could be present in any surface liquids (Raulin, 1987).
The Clausius-Mossotti function (see equation 2) is characterised by being almost
constant over a wide range of temperature and pressure, thus the electrical permittivity
measurement can yield density in moles per unit volume. This can be combined with
the density (mass per unit volume, measured by the DEN subsystem) to provide an
estimate of the mean molecular weight.
3E (E -1)
cY= Or
P3(5+2)
(2)
where p is density in moles/unit volume, (Y is molecular polarisability and NA is
Avogadro's number.
The refractive index of the Titan ocean, CL, (determined by the refractometer
subsystem) would equal ep�h if there were no change in E, between DC and optical
frequencies (assuming a magnetic permittivity of unity). In reality, a comparison of
these two measurements may give an indication of the molecular and orientational
polarisations of the ocean constituents.
The permittivity sensor consists of a 22-plate open capacitor with 1 mm spacing and
a functional area of 550 mm'. The range 250-275 pF is sensed to an accuracy of
0.25 pF, corresponding to an ethane to methane mixing ratio determination of 1%. In
addition, the sensor has a pair of exposed planar metal plates to sense the presence
of any ionic substances. The PER sensor houses a temperature sensor that gives
approximately 0.5K resolution.
188
Scientific Instruments
3.6 REF
The SSP refractometer (REF) is a linear critical-angle refractometer, as described by
Geake (1969, 1970, 1975). It uses a sapphire prism with a cylindrical active face, and
covers the refractive index range 1.250 to 1.450 with a discrimination of 0.001,
corresponding to measuring, to better than 1%, the relative abundance of methane and
ethane in the expected mixture in the liquid on Titan's surface. Two LEDs at 635 nm
provide alternative modes of illumination, internal and external, the former being for
general use and for in-flight testing, and the latter being useful if the liquid is found
to contain scattering material. The refractometer output image shows a cut-off edge,
whose position is a linear function of refractive index. The image intensity profile is
sensed by a photodiode array and digitised; it is then differentiated numerically, as this
is found to enhance the discrimination and may also give information about any
suspended matter. The differentiated profile shows a sharp peak at the point of
maximum gradient, and this is found to occur at the true position of the cut-off edge;
the pixel number at which this occurs is proportional to the refractive index of the
liquid, and constitutes the output signal. All the data processing and control functions
are carried out by software in the Low Power Processor (LPP) of the SSP
It is intended to transmit just the peak position, giving the refractive index value,
as the first quick-look output, followed by transmission of the whole differentiated
profile, if survival time permits, as this may contain further information about the
liquid, such as turbidity and the nature of any suspended matter. In addition to its use
on Titan's surface, it is intended to switch REF on during the descent through Titan's
cloud layers, in the expectation that cloud droplets and condensate collected by the
prism will enable useful measurements. Fig. 8. REF operation principle.
The instrument is as shown in Fig. 8. Its essential feature is a prism with a
cylindrical active face, and it is this that gives it its linearity and other useful
properties. The prism is made of sapphire because of the high refractive index (about
1.766), durability and radiation resistance. The instrument structure is of titanium,
which has a coefficient of thermal expansion very close to that of sapphire's. This is
advantageous in preserving the integrity of the prism mounting, as the prism
temperature is expected to fall to about 95K on Titan's surface. A thermal barrier is
provided between the prism and the upper box containing the detector and head
electronics, by truncating the prism to provide a gap, and by supporting it on thinwalled
tubes.
The detector is a self-scanning linear photodiode array of 512 elements used in pairs.
It is mounted directly on the PCB carrying the head electronics. Light is provided by
two LEDs, via glass-fibre light guides, giving alternately internal or external
illumination, depending on which LED is switched on. Crucial to the performance of
REF is a system of baffles, used to block unwanted rays and to minimise internal
scattered light.
REF was constructed at UMIST in collaboration with RAL; a brief description is
given in Geake & Mill (1992). It is the space version of an instrument designed for
laboratory use, which is described in Geake et al. (1994); this gives a fuller explanation
of the principle and properties of this new type of linear differentiating refractometer.
In the flight model REF sensor, the detector and its pre-amplifier circuit board are
mounted within the foam of the Top Hat. The flight model sensor (Fig. 9) has been
tested and shown to function at cryogenic temperatures. Three diode temperature
sensors, mounted at the linear photodiode array and at the top and bottom of the
sapphire prism, will enable the measurements to be corrected for temperature effects.
3.7 THP
The Thermal Properties subsystem (THP) will measure the temperature, thermal
conductivity and perhaps thermal diffusivity of Titan's atmosphere during descent and, Fig. 9. Flight model of the REF sensor.
189
SP-1177
if possible, surface liquids after impact (Birchley et al., 1992).
The THP sensor consists of two redundant pairs of platinum wire sensors, one pair
optimised for atmosphere measurements, and one for liquid measurements. These
elements are configured as four wire resistance thermometers, with a high drive
current available to provide power for measurements using a transient hot wire
technique.
The expression for the temperature rise of a long thin wire heated with a constant
power is given by (Healy et al., 1976):
(3)
where M = temperature rise of wire, h = thermal conductivity of medium, K =
thermal diffusivity of medium, q = power per unit length of wire, a = radius of wire,
t = time, C = 19 , where g = Euler's constant. If a graph of AT vs In(t) is plotted,
then the thermal conductivity is given by:
and thermal diffusivity is given by:
a2C B
K = ~ exp -
4 0 A
where A = gradient of AT-vs-in(t) graph, and B = intercept of AT-vs-In(t) graph.
The platinum wire sensors are housed inside individual cylindrical cells. These
perform two main functions. Firstly, they protect the wires during integration and fluid
in-rush following impact. Secondly, they reduce forced convection, which can cause
errors in the thermal properties results.
3.8 TIL
The tiltmeter is primarily intended to provide information on ocean wave properties
on Titan. Should the landing occur on a solid surface, the Probe's repose will be
returned, perhaps providing an indication of the metre-scale relief. In either case, the
tiltmeter will also provide information on the Probe's motion during descent, perhaps
allowing information on winds and atmospheric turbulence to be determined.
Probe motions after impact on a liquid surface can be used to determine a variety
of ocean and atmospheric properties. After the impact oscillations have damped out,
the Probe can be regarded as a buoy. Thus by measuring its vertical acceleration, using
the low frequency accelerometer in the HAS1 experiment, and its deviation from the
local horizontal in two axes with the tiltmeter, it will be possible to calculate the wave
motions. Much of our understanding of the properties of surface gravity waves, and
in particular the growth rate of wind waves, is based on empirical observations taken
in laboratory tanks or from measurements taken at sea. By extending the parameter
range covered by observations, it will be possible to test how general these empirical
observations are. Modelling of waves under possible Titan conditions (Srokosz et al.,
1992) indicates that a Titan ocean will occupy a very different parameter space from
terrestrial oceans, with typical waves having much longer wavelengths and larger
amplitudes for a given (plausible) wind velocity (maximum of a few m/s).
The sensor is also operated during the descent phase. Work by Lorenz (1994a) has
shown that the reconstruction of wind gusts from the Doppler Wind Experiment
190
Scientific Instruments
(DWE) data can be improved by using the TIL data and analysing the complex
pendulum motion of the Probe on its parachute system. TIL data may also provide the
Probe spin rate if this is greater than the maximum rate measured by the Radial
Accelerometer Sensor Unit (RASU, 15 rpm).
The selected device consists of two orthogonally mounted tilt sensors, comprising
small glass vials partially filled with an electrically-conductive fluid. When the sensor
is at its equilibrium position, the fluid's electrical resistance from the centre electrode
to each of two outside electrodes is balanced. Tilting the sensor disturbs this condition,
and the resistances change in proportion to the angle of tilt. The design accuracy for
the sensor subsystem is &0.50�h in both axes.
Accuracy for the separate sensors was specified in terms of the determination of ocean 4. Analysis of SSP Ocean
ethane content. This science requirement (see Table 4) was based on the assumption
of a binary mix for the ocean, with a determination of the ethane fraction to 1% .
Composition Measurement
Further work (Lorenz, 1994a), has shown that at best performance with no argon
present, but with solutes (nitrogen, propane and other hydrocarbons), the ethane
fraction may be determined to about + 2%. This is from the combined data of all
sensors, and is shown in Fig. 10. With argon present up to 4%, the uncertainty in
ethane fraction increases to &7 %
Beyond knowledge of the surface's physical state, an approximate age for the ocean
can be derived from the ocean depth (as measured by API-S), and the methane-toethane
ratio if the liquid is a large-scale ocean and not an isolated lake (for which age
Table 4. The original SSP science requirements, with worst case estimates of operational range.
Adapted from Lorenz (l994a).
Sensor Design Spec. Justijication
PER *0.001
THP f 0.001 W/m/K
10.003
+0.005 W!m/K
1% ethaneimethane ratio specification
1% ethaneimethane ratio specification. Degraded
accuracy due to turbulence effects. See Birchley
et al., 1992
API-V *5 mls *20 mls Speed of sound 2000 m/s: transducer separation
(- 140 mm) known to f 0.5 mm
DEN
REF
l I kg/m3
l O.OOl
5 kg/m3
*0.002
1% ethaneimethane ratio specification.
Susceptible to fluid motion
1% ethanejmethane ratio specification. This is
also the �eas built' performance of prototype
Accuracy
I 0.003 I
PER
DEN
API-V
THP
REF 0.002
0.001
I I I I I I I I I I I I I
75 70 65
Fig. 10. Accuracy of ocean composition
determination by SSP sensors when immersed
in a methane-ethane ocean with a 70% ethane
molar fraction. Bars indicate the range of
composition that are compatible with each
measurement, for the sensor's stated accuracy.
Adapted from Lorenz (l994a).
191
SP-1177
determination is lost). Rearranging equation (5) from Dubouloz et al. (1989) leaves an
expression for the age of the ocean (assuming constant photochemical rates for
dissociation of methane and production of ethane) in terms of measured depth and
methane-methane ratio:
(6)
where At is ocean age, x is mole fraction, F is the ethane production rate, NA is
Avogadro's number, V, is the molar volume, and h is the measured depth (subscript
�em' denotes methane, �ee' ethane).
5. OpC!rating Modes The SSP experiment operates in the following modes during Huygens' descent:
Mode 1 (Upper Atmosphere) Mode 1 is initiated at experiment switch-on (tc,)
(nominal altitude: 153 km) and continues until t,,+lO min (nominal altitude: 12 km),
i.e. the switch in overall SSP data rate from 3 packets per cycle to 8 packets per cycle.
(See Section 6 for a description of the packet structure).
Mode 2 (Mid Atmosphere) Mode 2 is initiated at b+lO min and continues until
tc+85 min (nominal altitude: 18 km), i.e. the switch in SSP data rate from 8 packets
per cycle to 11 packets per cycle.
Mode 3 (Lower Atmosphere) Mode 3 is initiated at ta+85 min and continues until
Probe altimetry data indicates that Huygens has reached a nominal altitude of 7 km.
Mode 4 (Proximity Mode) Mode 4 is instigated internally by SSP from altimetry data
when the Probe reaches a nominal altitude of 7 km. The mode ends at impact detection
or on a backup timeout. For the API-S subsystem, there is a sub-mode instigated from
Probe altimeter data for a nominal altitude of 1000 m to 0 m.
Mode 5 (Surface Mode) Mode 5 is initiated by impact detection (timpact), internally
by SSP, and continues until tlmpact+3 min.
Mode 6 (Extended Surface Mode) Mode 6 begins at tlmpact+3 min with an internal
SSP telecommand and continues until the end of the Huygens mission (loss of link with
Orbiter).
Each SSP subsystem will sample data at the rate appropriate to the scientific aims, with
certain subsystems not operational during some modes. The modes during which each
sensor will be sampled are indicated in Fig. 11.
6. Onboard Data Handling The SSP experiment has a MAS281 processor incorporated to perform data acquisition
and Telemetry and spacecraft commanding/telemetry handling. This processor is radiation-hard due
to its Silicon-On-Sapphire fabrication, and adheres to MIL-STD-1750a, which defines
the processor architecture. The design of the software had to take into account the nine
different sensors, each with their own data acquisition and timing requirements. To do
this, a small, multi-tasking kernel was written that could process each task independently
and check the timings for each sensor to ensure that they did not interfere with
each other.
The data are transmitted to the main spacecraft in 126-byte packets. The maximum
192
Scientific Instruments
Monitor ACC-I Readout ACC-I
1 API-V 1 X I < ________________________S_a_m__p_le_d_ a_l_ 1_H__z (in alternate directions) ____________________------>- --I- -----
API-S X Sampling in atmospheric mode Sampling in
proximity mode
Sampling in
surface mode
Extended surface
mode
DEN < _____________________________________S_a_m_p--l-e-d-- a--t- -1--H--z- throughout ____________________---------------------------->-- -----
l-IL < ____________________2_ _a_xe_s_ _a_t _1_H_z sampling rate_ ____________________>_ _<_ _____2__a _x es at 2Hz sapling rate_ ___>_
I PER I
< _____________________________________S_a_m_p--l-e-d-- a--t- -1--&-- - throughout ____________________---------------------------->-- -----
I
packet rate for SSP is 11 packets/cycle (16 s). The low data rate coupled with the Fig. 11. SSP data sampling requirements
disparate data rates of the sensors meant that a technique had to be found to transmit (x = not sampled during this mode).
the sensor data efficiently. To do this, statistical multiplexing techniques developed for
data communications were used. The data from a sensor are fed into a data stream
associated with that sensor. When a packet can be transmitted, a packet builder polls
each data stream in turn to see whether there is enough data volume in the data stream
to fill a packet. If there is. data are read from the data stream and put into the packet.
A word in the packet header identifies the origin of the data and the sequence count
for the sensor; this enables the source of missing packets to be identified. The packet
is then ready for transmission. The data for each sensor can be recovered simply by
de-multiplexing the packets and reconstructing the data streams back on the ground.
Identical data for SSP are transmitted on the prime and redundant CDMU channels.
A delay is introduced between the two channels by the Probe before transmission to
the Orbiter.
The EGSE for the Surface Science Package is designed to:
1. provide for full testing of the flight sensors during the pre-launch project
phases
2. assist in the flight operations by providing SSP checkout and realtime
telemetry data processing
3. provide data processing facilities for post-flight analysis of mission data.
The Special Check-Out Equipment (SCOE) is built around an Instrument Workstation
(IWS) and a stimuli subsystem. The Instrument Workstation is a powerful PC
providing commanding, communications, data processing and display facilities. The
display facilities may also be available on a secondary IWS connected to the primary
IWS. The stimuli subsystem itself comprises a central computer (Stimuli Monitoring
Unit, SMU), plus the API-S stimulation unit. By monitoring pre-integrated sensors
and generating control signals from stimulation units, the SSP electronics can be
developed and fully tested. For flight operations, the stimuli subsystem becomes
redundant but, due to the commonality of all SCOE computers, may be reconfigured
as either a spare IWS or an extra display station.
7. Ground Segment
193
SP-1177
Fig. 12. SSP Electrical Ground Support
Equipment (EGSE).
8. Calibration
9. Project Organisation
�e, EGSE LAN f
PIFS
(pre-integration) -
PROBE <
(post-integration)
The flight model SSP sensors were calibrated individually before Top Hat integration
in a calibration facility at UKC. As a minimum, sensors were calibrated in liquid
methane and liquid ethane (where relevant) over the Titan surface temperature range.
Following the flight model programme, flight-representative prototype sensors (and.
after launch, flight spare sensors) will be calibrated over as wide a range of ocean
compositions as proves possible.
A further, more detailed description of the UKC calibration facilities and the results
for particular sensors can be found in Garry & Zarnecki (1996).
The SSP design, manufacture, test and calibration results from the collaborative efforts
of several institutions. Table 5 lists the institutions and their responsibilities. In
addition to those activities listed, all groups are expected to take part in the data
analysis, concentrating where appropriate on data from sensors for which they have
specific responsibility.
Table 5. Institutional members of the Surface Science Package consortium.
Institutim Responsibili~
University of Kent at Canterbury, UK
Rutherford Appleton Laboratory. UK
University of Manchester
Institute of Science and Technology
ESA Space Science Department.
ESTEC. The Netherlands
Space Research Centre.
Academy of Sciences, Poland
Observatoire de Paris, Meudon, Paris, France
Institute of Oceanographic Sciences, UK
University of Arizona, USA
Martin Marietta Astronautics Group, USA
Principal Investigator, project management, sensor
development and calibration (ACC-E, ACC-I. DEN.
THP and TIL), ground support equipment
Structural. thermal and electrical design and
manufacture. sensor development (PER), software
development, integration and test facilities
Sensor development (REF)
Sensor development (API)
Electrical design, sensor development (THP)
Correlation with HAS1 experiment, calibration
facilities
Titan ocean modelling
Data analysis
Data analysis
194
Scientific Instruments
The work at the University of Kent, the Rutherford Appleton Laboratory and the Acknowledgements
University of Manchester Institute of Science and Technology is supported by the UK'S
Particle Physics and Astronomy Research Council.
Birchley, P. N. W.. Daniell, P. M., Zarnecki, J. C. & Parker, D. J. (1992). Laboratory References
Determination of the Thermal Conductivity of Liquid Methane and Ethane. In
Proceedings of the Symposium on Titan, ESA SP-338. ~~311-314.
Dubouloz, N., Raulin, F., Lellouch, E. & Gautier, D. (1989). Titan's hypothesized
ocean properties: The influence of surface temperature and atmospheric
composition uncertainties. Zcarus 82, 81-96.
English, M. A. (1995). The development of instrumentation and modelling for the
understanding of Titan. PhD Thesis, University of Kent at Canterbury.
Garry, J. R. C. & Zarnecki, J. C. (1996). A laboratory simulator for Titan's atmosphere
and surface. Presented at SPIE 96, Denver, Colorado, USA.
Geake, J. E. (1969). A linear refractometer for liquids. J. Sci. Instrum., Series 2 2,
477-484.
Geake, J. E. (1970). Optical properties of the linear critical-angle refractometer.
Optical Actcl 17. 363-380.
Geake, J. E. (1975). Linear refractometers for liquid concentration measurement.
Chem. Eng. 297, 305-308.
Geake. J. E. 8i Mill, C. (1992). The Huygens SSP Refractometer (REF). In
Proceedings of the Symposium on Titan, ESA SP-338, ~~343-346.
Geake, J. E., Mill. C. S. & Mohammadi. M. S. (1994). A linear differentiating
refractometer. Meas. Sci. Technol. 5. (5), 531-539.
Healy, J. J.. de Groat, J. J. & Kestin, J. (1976). The Theory of the Transient Hot-Wire
Method for Measuring Thermal Conductivity. Physica 82, 392-408.
Jakel, E., Rideau, P., Nugteren, P. R. 8i Underwood, J. (1996). Drop Testing the
Huygens Probe. ESA Bulletin 85, 51-54.
Kouvaris, L. C. & Flasar, F. M. (1991). Phase equilibrium of methane and nitrogen
at low temperatures: Application to Titan. Icarus 92, 112-124.
Lorenz. R. D., Bannister, M., Daniell, P. M., Krysinski, Z., Leese, M. R., Miller,
R. J., Newton, G., Rabbetts, P., Willett, D. M. & Zarnecki, J. C. (1994). An Impact
Penetrometer for a Landing Spacecraft. Measurement Science and Technology 5,
1033-1041.
Lorenz, R. D. (1994a). Exploring the Surface of Titan. Ph.D. Thesis, University of
Kent at Canterbury.
Lorenz. R. D. (1994b). Huygens Probe Impact Dynamics. ES4 Journal 18, (2), 93-117.
Raulin, F. (1987). Organic chemistry in the oceans of Titan. Ads in Space Res. 7, (5),
71-81.
Srokosz, M. A., Challenor. P., Zarnecki, J. C. & Green, S. F. (1992). Waves on Titan.
In Proceedings of the Syposium on Titan, ESA SP-338, ~~321-323.
Strobel, D. F., Hall. D. T., Zhu, D. T. & Summers. M. E. (1993). Upper Limit on
Titan Atmospheric Argon Abundance. Icarus 103, 333-336.
Zarnecki, J. C. (1992). Surface Science Package for the Huygens Titan Probe. J. British
Interplanetary Sot. 45, 365-370.
195