3. Cassini Arrives

It’s July 1, 2004, and Cassini’s long trek to its destination is over. In just one hour and fifty-two minutes, the spacecraft will be closer to Saturn than at any other time during its mission. The engine burn that will curtail Cassini’s interplanetary trajectory and deflect it into orbit around Saturn—Saturn Orbit Insertion, or SOI—is only minutes away. This burn will be the most critical event of the mission since launch. But first the spacecraft is crossing through the ring plane. To minimize the risk from a collision with a ring particle, it makes its passage through the rings in the large gap between the F and G rings. Shortly after Cassini has crossed above the rings, the main engine burn begins at 01:12 Coordinated Universal Time (UTC) and continues for ninety-six minutes.

 

RALPH’S LOG, JULY 1, 2004

I spent part of the morning filming with the BBC, driving up and down Pasadena’s Colorado Boulevard pretending to “commute,” and then pretending to check my e-mail on a laptop in a motel room (not my own). Just like the hundreds of contingency plans, backup sequences, or observation designs that don’t get used on a space mission, the media take a lot more footage than they actually use, and thankfully this pedestrian stuff fell onto the cutting room floor.

Saturn Orbit Insertion would be somewhat tense. A couple of Mars missions had been lost at just such arrivals, when fuel pipes exploded or bad navigation hurled them into the planet. In the emptiness of space, spacecraft are usually quite safe, but maneuvering in close proximity to a planetary body is a risky time. This is the moment when controllers find out whether the rocket motor will fire properly. For Cassini, there was also the apparent hazard of crossing the ring plane.

The media, with the space agency public relations machine as a willing accomplice, talked up the risk, and cameras zoomed in on the tense faces of controllers. In reality, there was little to control—seventy light-minutes away, whatever was going to happen was going to happen, and no heroic engineer was going to grab a joystick and be able to swerve Cassini out of trouble, even if we knew that some rogue ring particle loomed in our path.

And so, everything went just the way it was supposed to—pretty much exactly as portrayed in the drawings in the Phase A report back in 1988. Cassini passed across the ring plane, fired its motor for the required time, and ducked back across the rings. The only information we got was a radio tone, shown as a sequence of points on a huge screen. A green diagonal line showed what was supposed to happen, which was the frequency of the radio signal suddenly starting to change as the changing speed of the spacecraft during the engine burn altered the amount of Doppler shift. Every so often, as expected, the received data points would diverge from the “predicted” line as the denser parts of Saturn’s rings blocked the signal. But then the signal would come back, to cheering. And finally, when the green diagonal line straightened back to horizontal at the end of the motor burn, the red data points followed and did just that. Although some of the cheering and high-fiving seemed a bit forced, the relief and excitement were nonetheless genuine.

Saturn had a new satellite. We had arrived.

 

Cassini was on the brink of an unprecedented tour of exploration and discovery. But the saga that had successfully brought Cassini to this point, with its highs and lows, crises and triumphs, had already been going on for more than twenty years.

Figure 3.01. Artist’s impression of the release of the Huygens probe by Cassini. In reality, Titan would be much farther away, a mere dot. (NASA)

THE IDEA FOR CASSINI

The Cassini story began in the early 1980s, soon after tantalizing Voyager encounters with the Saturnian system in November 1980 and August 1981. The European Space Agency called for ideas in 1982, and an international group of three scientists responded. Wing Ip of Germany, Daniel Gautier of France, and Toby Owen from the United States proposed a mission they called Cassini, which would be a Saturn orbiter carrying a Titan probe.

This concept was examined by NASA and ESA jointly. NASA’s Galileo mission to Jupiter was carrying a probe, which would separate from the main spacecraft and descend through Jupiter’s atmosphere in 1995. There was a spare of this probe, and the original notion was to use it as the probe to go down onto Titan; ESA would provide the orbiter. However, the roles were later reversed. NASA engineers designed a new generic spacecraft, or “bus,” called Mariner Mark II. The idea was to save development costs by using the same basic vehicle for a series of missions. The logic there was straightforward, but taking on the responsibility for a planetary probe was a big step for ESA, which had not built one before. The closest thing had been a comet flyby spacecraft, Giotto, in 1986. It would require important technological capabilities new to ESA, for entry shields, parachutes, and so on. An assessment study was undertaken in 1984. It was followed by a “Phase A” study, begun in 1987 after a delay of a year to synchronize the schedules and budgets of NASA and ESA.

By late 1988, the technical challenges involved in constructing not only the probe vehicle but also the instruments to be carried were sufficiently understood that it was possible to consider the probe as a real mission rather than just a study. In November 1988, subject to NASA approving Cassini, ESA selected the Titan probe to be funded and named it Huygens. Time was now of the essence. The preferred route to Saturn and Titan would use the gravity of Jupiter to sling the spacecraft on its way, and Jupiter would be favorably positioned for this gravity assist maneuver only between 1994 and 1997. The U.S. Congress approved the start of Cassini in 1989. The joint mission was on the road.

Teams of scientists on both sides of the Atlantic considered what experiments they could carry out with Cassini and Huygens. The experiments would be competitively selected—the stakes were high. At this time, planetary missions were elaborate and carried many instruments, but were few and far between. The “faster better cheaper” era introduced by NASA Administrator Dan Goldin lay in the future.

INSTRUMENTS FOR HUYGENS

The Cassini project got going in earnest in autumn 1990, with the announcement of the selected payloads. Six experiments were chosen for Huygens:

1. The gas chromatograph/mass spectrometer (GCMS) led by Hasso Niemann of NASA’s Goddard Space Center

2. The descent imager/spectral radiometer (DISR) led by Marty Tomasko of the University of Arizona

3. The Doppler wind experiment (DWE) led by Mike Bird of the Radio Astronomy Institute at the University of Bonn

4. The aerosol collector/pyrolyzer (ACP) led by Guy Israel of the Service d’Aeronomie, Paris

5. The Huygens atmospheric structure instrument (HASI) led by Marcello Fulchignoni of La Sapienza University in Rome, Italy—later at the Paris Observatory in France

6. The surface science package (SSP) led by John Zarnecki, formerly at the University of Kent, but later at the Open University in the United Kingdom

At 18 kg, the GCMS was the most massive instrument, and in many ways the most important experiment. It would analyze the chemical composition of Titan’s atmospheric gases and of gases liberated by the ACP (discussed later). The GCMS would be able to distinguish between and identify the great variety of compounds in Titan’s atmosphere, not only by their molecular mass but also by their affinity to special coatings inside thin tubes, along which different compounds take different times to travel. It was a complex instrument to tackle a difficult job.

The DISR was Huygens’s camera that would take pictures looking both directly down and to the side as the probe descended. But in addition, it had the ability to record the spectra of sunlight filtering through the haze and reflected up from the ground. It could also measure the sunlight scattered around the Sun (the aureole), to obtain data about properties of the haze.

The DWE was one of the simplest. The only hardware for it on the probe was an ultra-stable oscillator on one of the two channels of the probe’s radio link. The principle behind it was to measure the change in frequency of the signal received by the orbiter by comparison with a reference oscillator on the orbiter. From this it would be possible to determine the probe’s motion. Of the measured Doppler shift in frequency, most would be the result of the orbiter’s rapid approach toward Titan, and some would be due to the probe’s vertical descent. These two components could be calculated and removed. Any remaining component of the probe’s motion would be due to wind.

The ACP was a novel instrument designed to trap aerosol particles by sucking atmospheric gas through a tiny filter held out in front of the probe. The filter would then be pulled inside the instrument and heated in a set of temperature-controlled ovens. Aerosol material breaks down when heated. The gaseous products released were to be transferred through a pipe to the GCMS instrument for analysis.

The HASI was to measure the basic properties of Titan’s atmosphere, such as pressure and temperature, during the probe’s entry and descent. A capability for HASI to digest the signal from the probe’s engineering altimeter was added later.

The SSP was a collection of small sensors. As well as an accelerometer and penetrometer to characterize an impact on a solid or soft surface, the package included instruments to take measurements in the event of a “splashdown” in liquid. Tilt sensors would measure the probe’s orientation and bobbing motion in any waves. An ingenious refractometer would measure the ethane/methane composition of the liquid, with additional information provided by thermal, density, and electrical permitivity sensors. An acoustic sensor would act as a sonar to measure the depth.

The hope was that together, the six experiments would maximize the data return during the few precious minutes of Huygens’s descent, within the constraint of the size and mass capacity of the probe.

 

RALPH’S LOG, AUGUST 1994

Each of the instruments involved the efforts of dozens of scientists, engineers, technicians, and students. One of them was me.

One function of the SSP was to measure the mechanical properties of Titan’s surface by recording the impact of the probe, if it survived. Since I had a background in aerospace engineering, as well as a detailed knowledge of the probe, having worked in the ESA project team for a year already, this investigation seemed like an ideal fit with my interests and expertise and would make a good PhD project.

It is rare that such a junior individual gets to play such an identifiable role in a major space endeavor, but someone has to do it. The budget for the SSP was tight, and as a PhD student, I was cheap labor. So, at the tender age of twenty-two, I was made responsible for specifying what sensors SSP should use—how fast and how accurately they would have to be recorded, and so on. There had been preliminary designs, of course, but these had been proposed without knowledge of the final probe design.

Between 1991 and 1994, the SSP team decided that two sensors would make the impact measurement. For very soft landings, or splashdown into a liquid, an accelerometer mounted on the SSP electronics box would do. I had read up on all the details of splashdown testing from NASA records of the Apollo program. That bit was easy. An off-the-shelf accelerometer would be used, rather like those used to actuate air bags in cars. But for impacts into harder materials, the acceleration recorded would be affected more by the structural properties of the probe itself than by the surface hardness we wanted to measure. And so I designed a penetrometer, a force sensor that would stick out of the bottom of the probe on a rigid mast. If the probe landed flat, this would be driven into the ground, generating a force signal.

We had to check that the sensor would work after a lethal dose of radiation, such as the one the probe would get flying by Jupiter, and that it would function acceptably at Titan’s surface temperature. A way had to be found of checking the instrument’s health en route to Titan—these and dozens of other details were worked out on paper before it could actually be built.

Eventually, the design details were settled and all the parts arrived—titanium alloy for the penetrometer head, the custom-made piezoelectric ceramic disks. Even the wires in SSP were special: stainless steel instead of copper, to minimize how much of the probe’s warmth would leak out along them.

And so, here I was one summer day in the University of Kent’s Electronics Workshop, feeding a little wire through a little hole. I held it in place as Trevor Rees applied the soldering iron and bonded the wires to the transducer, ready for their long, cold trip. (I was not permitted to do the soldering myself—that had to be done by someone who had been space-certified after taking a special ESA course in high-reliability soldering.)

In fact, we built five penetrometer heads. The best one, the one where the wire seemed to bend a little more smoothly, where the solder joint seemed just that bit neater, was designated the flight model—the one that would go to Titan. Another two, almost as good, were also double-bagged and placed in the secure “flight cupboard” as flight spares, in case something were to go wrong. The other two, which worked fine but weren’t quite as sound, would be used in lab tests and student projects.

 

HUYGENS: DESIGNED FOR ITS JOB

The design of the probe was to deliver a payload of 50 kg of instruments into Titan’s atmosphere and have them descend to the surface in about 135 minutes, while keeping them warm, supplied with power, and transmitting their data to Cassini. Additional requirements were that the probe should start its descent around 160 km altitude, that it notify the experiments when the altitude dropped below 10 km, and that the probe spin at a few revolutions per minute during descent so the optical instruments could pan around. Further, it should support the payload for a minimum of 3 minutes after impact. However, survival beyond impact was not a requirement. There was no way it could be, without knowing more about Titan’s surface.

In order to do all that, of course, it would also have to survive the noise, acceleration, and vibration of being launched into space on a rocket, spending seven years in the vacuum and radiation of space while attached to Cassini, and another twenty-two days coasting by itself. But perhaps the biggest challenge of all, the toughest one to be sure about, was entry. The twenty-two-day coast sounds benign, but what this means is hurtling at 6 km per second toward Titan. At this speed, the kinetic energy of the probe is the same per kilogram as a high explosive like TNT. Somehow the probe would have to shed that energy.

To reduce its speed, the probe was made with a round-nosed conical airbrake or decelerator. This would act like a rigid parachute, slowing the probe down by air drag. But at hypersonic speeds, the air would glow with the violent energy of entry, and a bare structure or parachute would melt. So the decelerator was coated with a high-temperature insulator or heat shield.

Figure 3.02. A schematic representation of the Huygens descent sequence. After hypersonic entry in its heat shield, a pilot chute pulled off the back cover and allowed the main chute to inflate. This slowed the probe down and allowed the front shield to fall away; thereafter, the main chute was released and the probe descended under a small “stabilizer” parachute. It is pictured here finally operating on the surface. (ESA)

The prime contractor for the probe, selected through competitive bidding, was Aerospatiale of Cannes, France (later to become part of Alcatel and finally, as Europe’s post—cold war aerospace industry consolidated, Alenia-Alcatel Space). It headed a consortium of suppliers from several European countries, including Germany, Spain, Italy, and the United Kingdom. As with other European Space Agency projects, the industrial contracts had to be shared among the member states, the value of the contracts going in rough proportion to the contributions each country makes to ESA’s budget. On this basis, France, Germany, and Italy were to take the lion’s share of the work, followed by Britain, Spain, and the many smaller countries. It is a challenge for the project managers in ESA, who are usually based at its technical center, ESTEC, in Noordwijk, the Netherlands, to juggle the bids from various companies to make sure that France is supplying 35 percent of the project, and Denmark its 2 percent, and so on. It doesn’t necessarily lead to the most efficient technical solutions, but it is a system that works surprisingly well, bearing in mind the diversity of ESA’s members.

Aerospatiale would lead the technical design and be responsible for accommodating the experiments. They would also supply the all-important heat shield, based on tiles of a silica material used in French ballistic missiles. Alenia and Laben of Italy provided the data handling and communications systems. A British company, Martin-Baker, developed the descent control system, the set of pyrotechnics, lines, and parachutes that would pull the probe away from its heat shield and bring it safely down to the surface in the allotted time. The metal structure of the probe was to be supplied by CASA of Spain.

Figure 3.03. An engineer in a white clean-room suit performs some final assembly tasks on the Huygens probe at Cape Canaveral. The front shield is visible behind the white support fixture. Note the cold air hose in the foreground. (NASA/KSC)

A spacecraft, like any other modern aerospace project, is not just hardware. Important elements included the onboard software and the procurement services to secure the space-rated pedigree of its components (provided by Logica and ITT in the United Kingdom). The probe itself would be assembled in Germany, where extensive tests would take place to verify its ability to function in the vacuum of space and in the deep cold of Titan’s atmosphere, to check its ability to tolerate lightning strikes, and so on.

All of this, as well as the important interfaces with the Cassini project in the United States, had to be coordinated. Literally hundreds of face-to-face meetings and teleconferences would negotiate and agree to the various technical and contractual details—all, of course, to be recorded in tons of thick documents, changes to which would be meticulously tracked. And almost any detail or function on a spacecraft has to be tested. Sometimes designing and executing the test is as much of a challenge as designing and building the thing in the first place! It was a great deal of work. At the beginning, in 1990, much of the communication was by fax; e-mail was not as prevalent as it is today, nor had Microsoft Office software yet emerged as a global standard. And the World Wide Web did not even exist! But this large-scale system engineering activity and project management process—arguably a more significant spin-off of space exploration than the nonstick frying pans that are often (erroneously) cited—brought it all together in the end. Time zones, computer incompatibilities, different units of measurement, different sizes of paper, different languages and ways of doing business are all just barriers to be overcome by patience and hard work.

Figure 3.04. The day-to-day business of spacecraft and instrument engineering is not as surgical as figure 3.03 implies. At left is the Huygens probe “engineering model,” the probe equipment deck festooned with wires to verify electrical function. The metal structure is not present. At right is a test of the DISR instrument’s Sun sensor on the roof at the University of Arizona. (ESA and R. Lorenz)

Beyond the cultural differences within Europe and between Europe and the United States, another cultural divide must be overcome on a project like this—namely, that between scientists and engineers. Although both are technical disciplines, they have very different styles. Engineers are more team-oriented, used to the hierarchical organization of engineering organizations. Scientists are more often prone to prima donna behavior but sometimes less constrained in how they work.

Many people trained in science go on to engineering careers; rather fewer people go the other way. The scientist as the “customer” wants to supply a massive, power-hungry instrument that looks in all directions and generates a mountain of data that the scientist and his or her colleagues and successors can sift through for years afterward. The engineer, with similar demands from all directions and faced with ultimate limits on the cost of the project, the size of the launcher, and so on, must work to keep the payload and its needs manageable. The engineer’s job would be easiest if a scientific instrument were just a small box that could be put anywhere on the probe, and didn’t need any power or data. The dynamic tension between the two disciplines results in intense negotiations but contributes to an optimum solution emerging.

A peculiarity of ESA’s system (compared with NASA’s) is that the agency does not pay for the instruments that fly on its missions. The costs of both hardware and scientific personnel are paid for by individual member states and consequentially can be subject to all kinds of uncertainty. In contrast, NASA pays for both the platform and the instruments, and thus has some measure of overriding control.

Although the Cassini orbiter was for the most part built in the United States, and the Huygens probe in Europe, and their payloads likewise, there were components of each from the other side of the Atlantic. Some components on the Huygens probe, like batteries and accelerometers, came from the United States, but the orbiter’s high-gain antenna was built in Italy. On the scientific side, the two most complex experiments on the probe, the gas chromatograph/mass spectrometer and the descent imager/spectral radiometer, were led in the United States, while the teams working on the Cassini dust analyzer and the magnetometer on the orbiter were led in Europe. Cassini—Huygens was therefore a monumental joint effort, requiring patient coordination.

CASSINI’S TOOLS FOR TITAN

In parallel with the probe’s development taking place primarily in Europe, the much larger orbiter spacecraft was being put together in the United States. The instruments selected for Cassini were announced soon after the list for Huygens. They included several that would be used to study Titan during the dozens of close flybys it would make through the course of the mission.

Figure 3.05. Configuration of the Cassini spacecraft. (NASA/JPL)

Sadly, the contracting economy of the early 1990s took its toll on plans for planetary exploration. The “two for the price of one and one-half” logic of Mariner Mark II faltered against a NASA budget crunch in 1992 as NASA struggled to pay for its part of the International Space Station. And so Cassini’s sister mission, CRAF (Comet Rendezvous and Asteroid Flyby) was canceled. Cassini survived this financial crisis by the skin of its teeth, chiefly because of its international nature, which would have made cancellation look very bad. But Cassini was slimmed down, many of its instruments having their capabilities reduced (“de-scoped”). Crucially, the scan platforms—devices to point instruments independently of the body of the spacecraft—were deleted to save money. By any rational life-cycle costing, this would be seen as a false economy, the savings in construction in 1992 being eaten up and more by the additional operations complexity years later because the whole spacecraft would have to be slewed around to point its different instruments at targets as it whipped by.

Even with descoped instruments, Cassini had a formidable science payload. Disentangling the bewildering array of phenomena in the Saturnian system—the interactions of the rings, the satellites, the magnetosphere, and the atmospheres of Saturn and Titan—requires simultaneous measurements of many kinds. It is not the sort of exploration that can be tractably addressed with a small mission.

Cassini was the first outer solar system mission to be equipped with radar. Unlike most radar-mapping spacecraft, which operate over a fairly limited range of altitudes, Cassini would be making radar observations from as far away as tens of thousands of kilometers to within 4,000 km of the surface. To cope with this, the design needed to be highly versatile. All of Titan was to be mapped with low resolution, but in addition, strips totaling about one-quarter of the entire surface would be charted much more finely. In sweeps during Cassini’s close passes over Titan, the radar would look to the side with five beams that form a line. Dragging this line across the surface covers a strip a couple of hundred kilometers across and several thousand kilometers long. The signals are processed in a clever way so that pixels only 400 m across can be measured.

The visual and infrared mapping spectrometer (VIMS) was to be another versatile instrument with applications to Titan. Its field of view is a narrow strip, but each strip is turned into a high-resolution spectrum. By sweeping the strip around, or by simply letting the spacecraft’s motion drag the strip, an image cube can be built up—a stack of images of the same scene in many hundreds of different colors. Effectively, VIMS records the spectrum of many pixels in an image simultaneously.

Cassini’s imaging instrument, known as the ISS (imaging science subsystem), consists of two telescopic cameras—one wide-angle and one narrow-angle. Both have sensitive CCD detectors and an array of filters. Some of the filters have been specially matched to the 0.94-micron “window” in Titan’s atmosphere, to probe Titan’s surface.

A key instrument for Titan was the ion and neutral mass spectrometer (INMS). Even though Cassini would be flying 1,000 km above Titan’s surface, there would still be enough atmosphere at that altitude for its composition to be measured by this instrument, pointed forward like a scoop, literally counting the molecules of different masses.

The ultraviolet imaging spectrometer (UVIS) would be used to detect airglow—the means by which Voyager had detected molecular nitrogen at Titan. It would also be used in occultations of the Sun and of stars to profile the amount of gas and haze in the upper atmosphere.

As the Galileo spacecraft did on its mission to Jupiter, Cassini carries a set of sensitive magnetometers. Water, especially if it is salty, conducts electricity—unlike ice and rock, which do not. Moving through a giant planet’s magnetic field, a subsurface ocean on a moon generates a small magnetic field of its own. Galileo detected the appropriate magnetic signature for Europa as expected, but when Callisto showed the same magnetic signature, it came as something of a surprise. Cassini’s magnetometer would be able to carry out a similar test on Titan.

Cassini was equipped with a radio communications and tracking system more elaborate than any other planetary spacecraft has ever had, and it would be possible to probe Titan’s atmosphere by means of its signals. Passing the radio signal through the upper atmosphere of Titan could determine the density of free electrons. Lower down in the atmosphere, the signal is refracted by the denser layers. The radio signal is also affected by Earth’s ionosphere and the tenuous gas between the planets. However, by measuring its three frequencies simultaneously, these effects can be removed to get much more accurate measurements. Cassini also carries long, sensitive radio antennas to search for electrical phenomena in the Saturnian system and to listen for lightning at Titan.

LAUNCH

Cassini was launched from Cape Canaveral in the early hours of the morning (local time) of October 15, 1997, by a Titan IV/B launch vehicle, the most powerful launcher in Western inventory. Just as sailors must take the tide or remain stranded, planetary exploration is governed by windows of opportunity, when planets are aligned correctly. Cassini had rather stringent constraints on its launch, since it relied on no less than four planetary encounters to hurl it on to Saturn. The launch window would last only a month or so. If there were a problem that took longer than that to fix, the launch would have to slip by months but the arrival at Saturn would be delayed by years, and fuel that could be used at Saturn would be consumed in getting there.

The probe and its experiments, shipped from Germany, had already been in Florida for months, undergoing various interfacing and compatibility tests with the orbiter, trucked over from California.

A last-minute hiccup did delay the launch. Huygens was well insulated against the cold of space and could get hot during tests on the relatively warm Earth. So a duct blew cold air continuously into the probe to keep it cool. However, somehow the airflow on this duct got set far too high, and the gale-force blast of air shredded some of the insulating foam on the inside of the probe. The risk that particles of foam might have penetrated some sensitive component couldn’t be taken. The probe had to be removed from the launch pad. Opening the probe up and vacuuming its insides took nearly a couple of weeks, half the available window.

Figure 3.06. Assembly of Cassini at Cape Canaveral. The instrument pallet carrying the cameras and spectrometers, wrapped in gold-colored Kapton insulation, is being lowered toward the Cassini orbiter structure in the scaffold. (NASA/KSC)

After this slip, launch was rescheduled for October 13. A large crowd of scientists and engineers gathered in the early hours to watch the launch. But, as can often happen, a computer glitch and high winds at altitude prevented the launch from taking place. There would be another two-day delay.

In fact, although one wouldn’t want to do this as a matter of policy, in case the weather turned sour or some mishap occurred on the launch pad, the two-week delay in launch actually had a major benefit. The launch window represents a period during which conditions are acceptable—the planets are in the appropriate alignment and so on. But not all moments in the window are the same, the middle of the window usually being closest to optimum. And so it was with Cassini. The mission would require more fuel to reach Saturn if launched at the beginning or end of the window. The delay caused the launch to take place in the middle of the window, when the fuel costs were less. This was good news. If the rest of the mission went without major problems, the fuel saved could be used to prolong the mission, or to perform a more aggressive tour with more close flybys of Titan.

Fortunately, on the fifteenth of October 1997, Cassini got off the ground safely. This night launch was quite spectacular. Not only was the Titan rocket the largest conventional launch vehicle in U.S. inventory at the time, but the monster solid rocket motors strapped to its side have an extremely luminous yellow exhaust. Even from miles off, the ascent was brilliant, the rocket itself invisible next to the dazzling columns of flame. Nature added an aesthetic flourish, a fluffy cloud over the launch pad—not a thunderstorm, thankfully, which would have scrubbed the launch. As Cassini ascended through it, the cloud was lit up from inside like a Chinese lantern. It seemed a good omen.

The roar of the engines took many seconds to reach the assembled spectators, presaged by a bit of a rumble that propagated through the ocean. Thus, the engine ignition took place in apparent silence, soon replaced with cheers and applause. It was a cathartic moment, a huge relief. Even the most trustworthy rockets are only 97 percent reliable, and the upgraded solid motors on Titan, which would save Cassini’s fuel for use at Saturn, were comparatively untried. But everything worked perfectly, and after a minute or two, the burned-out husks of the solid motors tumbled like comets toward the sea, their job done, while Cassini sailed off into the darkness.

The scientists gathered at the launch site still had several days of meetings—to thrash out the design of the tour, the specification of archive data products, the latest updates to planning software, and so on. But the launch was a watershed in the project, marking the transition from hardware to operations. Many engineers would move on to new jobs. And although some of the scientists would remain busy in planning the tour, calibrating the instruments in flight, designing observations, and so on, many would disappear from the scene for a few years until their expertise was called on to analyze Cassini’s data.

 

RALPH’S LOG, 1998

DS-2

Although I was heavily involved in the various planning tasks, Cassini was not enough to support my research full-time, and so I pursued some other opportunities. A project that made a neat fit with my Huygens work came up in early 1998. This was a small Mars Penetrator project called the DS-2 Mars Microprobes, part of NASA’s New Millennium Program of technology validation missions, the idea being to alleviate the risk in trying new technology on big science missions by showing they work in fast, cheap projects first. At 28 million, it was a tiny fraction of the cost of Cassini.

The two microprobes were like tiny versions of Huygens—about the size of a basketball. They would piggyback to Mars on the Mars Polar Lander (MPL) mission, to arrive in December 1999. Each microprobe weighed about 4 kg and was encased in a lightweight but brittle heat shield. There was no parachute; these probes would slam into the ground at four hundred miles an hour. But they were designed for that—the heat shield would shatter, and these specially toughened probes would penetrate about half a meter into the ground, decelerating at some 20,000 g. There they would cook a tiny sample of Martian soil to see how much water was locked up in the ground (this was, of course, long before the Mars Odyssey spacecraft discovered lots of permafrost remotely in 2002).

Figure 3.07. Launch! Cassini blasts off on the Titan IV launcher in October 1997, a spectacular night launch. (NASA)

My role in the project was to determine how we could measure the ice content of the ground, find any layering, and measure the depth of penetration using the deceleration profile recorded at impact. This was jolly good fun, and involved blasting mock-ups of the probes into the ground with a huge air cannon on the back of a truck at a ballistics facility in New Mexico, with dead helicopters and tanks with holes in them littered around. Many of the technical problems that had to be confronted were similar to those on Huygens, but the style of the small project was very different, and the schedule almost ridiculously fast.

Martian scientists have it easy. There is almost instant gratification: you build your instrument or spacecraft, launch it, and in less than a year, you are there, getting data. After a smooth launch in January 1999 (a daytime launch on a small Delta rocket—impressive, but not quite as awesome as Cassini’s nighttime Titan), MPL and its two diminutive sidekicks, since named Scott and Amundsen in a student contest, neared Mars on December 3, 1999. The DS-2 probes would last only a couple of days before their batteries ran out, and the DS-2 science team geared up to present some first results at the AGU conference the following week.

JPL (the Jet Propulsion Laboratory) was a zoo. Even the fact that much of the MPL science operations would be performed from UCLA a few miles away did not seem to take the heat off—landing on Mars is big news! Scientists had to use the remote wilds of the car park because the prime spots were taken by VIPs and television trucks. Suddenly the policy of not tailgating someone through the access-controlled door to the operations building was enforced by a humorless guard with a gun. We got ready, briefed on the best color scheme to use in plots of data so that they show well on TV. We gave interviews, discussing what we hoped to find. The data would arrive in the early evening, so we tried to get some extra sleep ahead of time so we could work through the night.

And then, at the appointed hour, silence. Nothing. No signal from either of the microprobes, or worse, from MPL either. Well, maybe the battery was cold, and the spacecraft will report in during the next window in a few hours. No such luck. Engineers run through the various scenarios—well, if this failed, then this would switch in automatically eight hours later, and the lander should respond to ground commands. The scenarios played through, but still no response.

After a couple of days, we just had to give up and go home. We knew DS-2 was a risky mission and surely expected that one might fail. But to lose both, and MPL as well, was a horrifying shock. No exciting new findings, no demonstrating new, efficient ways of delivering payloads to the Martian surface. Nothing but some test reports and some lessons learned. That’s show business.

God, I hope Huygens doesn’t end up like this.

 

PLANNING THE TOUR

It would take seven years for Cassini to reach Saturn, but even that interval would be too short to work out all the details of what to do once it got there. First the teams had to agree on the “tour,” the best orbital path around the Saturnian system. The nominal length of the mission was four years, and during that time, Cassini would make over seventy orbits of Saturn. The challenge was to devise a tour that would optimize the opportunities for all the experiments on board to gather observations of their targets, whether Saturn itself, the rings, or moons. And of course, early on in the tour, there would be the release of Huygens.

Once that was agreed (and trying out dozens of possibilities, progressively reaching a better design, took over five years), then the minute-by-minute sequence of pointing the spacecraft and designing the observations themselves (exposure times for the cameras, bandwidths for radio instruments, and so on) had to be done. This process was, of course, made all the harder by the compromise in Cassini’s design that had to be made in 1992, the deletion of the scan platforms. Instead the cameras were bolted to the side of the spacecraft, which meant that pointing them would require the whole spacecraft to be turned around. It also meant that it would not be possible to point the antenna at Earth while images were being taken. Operations would have to be arranged to take place sequentially rather than simultaneously. So even as the hardware moved to Florida in the latter half of the 1990s, and then into space, there was plenty of work to do.

Though Cassini left the launch pad in October 1997, it would not actually leave the inner solar system until almost two years later. Even the most powerful rocket available, the Titan IV/B, was not capable of imparting enough speed to Cassini to propel it to Saturn. Cassini was a monster of a spacecraft, at 6.8 m long among the largest ever launched and weighing in at 5.5 tons. The solution was the gravity assist technique—picking up speed in a close encounter with a planet to create a kind of slingshot effect. Cassini’s flight plan required not just the close flyby of Jupiter that constrained the overall window of opportunity for the mission, but three helping hands in the inner solar system. So it was that Cassini headed first toward Venus. In two complete orbits around the Sun in the inner solar system, Cassini swung past Venus twice, on April 26, 1998, and June 24, 1999, and then once by Earth, on August 18, 1999. Passing Earth at a distance of 1,180 km added a valuable 5.5 km/s to Cassini’s speed, sending it on its way at about 25 km/s.

PUTTING THE PETAL TO THE METAL

Once at Saturn, there would be a near-infinity of options. In the broadest terms, the first order of business was to deliver the probe and get the orbital period down by bleeding off energy at the first few Titan flybys—rather like the slingshot technique in reverse. Then successive Titan encounters would use Titan’s gravity to slowly change Cassini’s orbit around Saturn. At each flyby, the orbit could be changed in one of several ways. The orbit could be shrunk or grown, changing its period—usually from one resonance with Titan’s sixteen-day period to another, such as from a thirty-two-day orbit to a sixteen-day one, since Cassini would have to reencounter Titan to make further changes. Or the orbital plane could be changed—raising or lowering the inclination. Or the orbit could be rotated within its plane, the so-called petal rotation.

The architecture that evolved, after literally dozens of trials, was as follows. After the first couple of orbits, during which the Huygens probe would be delivered so as to arrive with the Sun in the right part of the sky for the camera to work, and so on (the second orbit being backup for the first), the inclination would be brought down to zero, putting Cassini in the Saturnian ring plane. This would be good for several things, notably for many close encounters with Saturn’s other satellites and an edge-on view of the rings, as well as occultations of Saturn’s atmosphere. Then the petal would be rotated from the dayside to the nightside, permitting an exploration of the Saturnian magnetotail. After that, in what would be late 2006, the inclination would first be increased with orbits that encountered Titan as Cassini flew away from Saturn (outbound encounters), then decreased with inbound encounters. This novel maneuver, called a 180-transfer or “cranking over the top,” flipped the orbit back into the dayside like a pancake. Finally, the orbital inclination would be progressively increased to allow Cassini to look down on Saturn’s pole and the rings toward the end of the nominal mission, finishing up with an inclination of some 75° in June 2008.

Figure 3.08. Cassini’s interplanetary trajectory took it two and one-half times around the Sun, on four planetary flybys, before it reached Saturn. (NASA/JPL)

All in all, there would be forty-four close Titan encounters, numbered T1 to T44. A less “Titanocentric” system also referred to the orbit or “rev” numbers; yet a third system referred to sequences, the several-week-long blocks of commands sent to the spacecraft. Thus, Titan flyby T8 occurred on rev 17, in sequence S15 on day 301—better known as October 26, 2005.

Figure 3.09. Cassini’s tour around the Saturnian system, designated T18-5. This particular variant is T18-5-JD4. The left view is from the north pole, with the direction to the Sun toward the bottom. The right view is from the Sun, north up, showing the inclination of the orbits varying throughout the tour. The lowermost, slightly kinked petal is the first orbit, the kink being the periapsis raise maneuver. (NASA/JPL)

In the original plan, Huygens was to be released on the first Titan flyby (T1), planned for November 27, 2004. However, the mission was redesigned (see the following section) to include an additional orbit around Saturn and for the probe to be released a little later, so it would reach Titan in January 2006, on a third flyby. Because so much work had gone into planning observations in the tour, and countless documents already used the existing sequence numbers (it is easier to reprogram a spacecraft a billion miles away than it is to reprogram a bunch of scientists on Earth!), the new tour began with flybys TA, TB, and TC before picking up at the old T3.

A BIT OF TROUBLE WITH HUYGENS

During the Earth flyby, engineers took the opportunity to test out various instruments and systems. One of those tests was of the radio link system that would operate between Huygens and Cassini during the probe’s descent. A signal was transmitted from Earth, simulating the probe. It soon became apparent that all was not well.

It emerged that a design compromise years before in the Europe-supplied radio receiver on Cassini that would receive Huygens’s signal led to it being highly sensitive to the exact rate at which it received the radio bitstream of data. If the received frequency were slightly off, the bit-stream would become desynchronized and the data would be corrupted. Unfortunately, although the overall receiver design accommodated all the anticipated factors on the frequency and had been tested on the ground, the Doppler shift due to Cassini’s motion relative to Huygens was enough to trip up the bit synchronizer, and this rather difficult test had not been attempted.

When the results came back from the Earth flyby, the catastrophic problem was soon diagnosed. There were many places where the design could have been fixed, in hardware or software, but only when the probe was on the ground. The crucial parts in the radio were not remotely reprogrammable. A recovery team was convened, with experts in the mission and in digital radio design to hand. Although the situation looked bleak, the team members at least had time on their side. There were still five years to go.

The solution involved several changes to the mission. First, the probe would be switched on several hours early. It looked as if there was enough battery power to do so, since the batteries seemed to be holding their charge well. This would cause the radio transmitter to warm up, shifting its frequency slightly and improving the synchronization. The software on the probe, and on the various experiments, had to be modified accordingly. Some changes in ground software would recover a few of the otherwise-rejected data packets. But the biggest change was to the orbiter mission. In the original plan, Huygens was to be delivered on the first orbit. By delaying the delivery of the probe for a few months, and flying by at a greater distance, the Doppler shift due to the relative motion of the probe (dangling under its parachute in Titan’s atmosphere) and the orbiter, whipping past at 6 km/s, would be reduced to a level where the bit synchronizer could lock on correctly.

Simulations of the radio performance showed under what conditions the receiver would lose data. One particularly instructive plot of signal strength against frequency showed a jagged region, the “shark’s teeth,” where it would fail, and tests on the spacecraft as it cruised Saturnward agreed perfectly with the model. With various software patches made, the system was tested and tested again, and all the indications were that the problem was solved. Of course, it cost some of Cassini’s precious fuel to adjust the orbit, and a considerable amount of time and effort to develop the solution, to make the various software patches (which involved bringing some individuals out of retirement), and to verify that the temperatures and new longer probe mission would not introduce yet new problems. But the exercise also had the effect of focusing the teams’ attention on the probe and how it would work.

Figure 3.10. The revised Huygens delivery trajectory. The orbiter and Huygens are on near-parallel paths, the orbiter about two hours behind and 60,000 km to one side of the probe. This kept the Doppler shift of the radio signal small enough for the weak receiver design to cope with and recover the data. (NASA/JPL)

BY JOVE!

The Jupiter flyby was an important opportunity to exercise the capabilities of the teams and the spacecraft to design and execute instrument sequences, in much the same way as they were going to have to at Saturn. When Voyager flew by a planet in the 1980s, it would take up to six months to generate the detailed observing plan, to work out which instruments were pointed where, and when, and when the ground station would be available to receive the data. Cassini was to have around one hundred such encounters, so it was crucial to develop a streamlined process for generating a timeline of operations. Jupiter was the first real chance to try this out and find what had to be improved. In order to spread out the budget of a space project, the development of some capabilities (such as having the spacecraft repoint itself to make mosaics of multiple images) is deferred until after launch.

This meant that the science teams were learning as they went along, using software that was still under development, being written and improved even as it was being used for the first time—a frustrating situation! Even though the Jupiter flyby observations were officially for instrument checkout, there was, of course, the prospect of new scientific discovery, so the teams worked against the clock to get the sequences ready.

An important bonus was that the Galileo spacecraft, in orbit around Jupiter since 1995, was still operating. Making simultaneous measurements with two spacecraft would leverage the scientific value of each—for example, by measuring in situ the conditions of the solar wind just upstream of Jupiter, and by measuring the effects of the changing wind on Jupiter’s magnetosphere.

The Jupiter encounter provided a beautiful sequence of thousands of images of the planet. These filled an important gap in Jupiter science, since the failure of Galileo’s antenna had meant that it did not have the capacity to send back many pictures. A novel measurement used Cassini’s radar to sniff out the faint microwave glow from Jupiter’s synchrotron radiation belts, revealing that high-energy electrons (5–20 MeV) were more abundant than had been predicted.

Although the sequence was derailed for a few days because one of Cassini’s reaction wheels, used for fine pointing, seized up, overall the Jupiter encounter was a great success. The data are even now only just beginning to be analyzed in detail.

NEARING THE TARGET

Even in October 2002, when Cassini was still 285 million km from Saturn—almost two astronomical units—the spacecraft’s view of Saturn was an arresting sight. Cassini’s vantage point gave it a view never observable from Earth (but rather similar to a sketch Christiaan Huygens once made) of the shadow of Saturn cast onto the rings, with the north pole of Saturn just peeking above the shadow the rings made on the planet. Titan was visible, though only as a dot a few pixels across. Things would just get better and better from here on.

Figure 3.11. An image of Jupiter taken during Cassini’s flyby in December 2000. The Great Red Spot is prominent at left; Jupiter’s volcanic moon Io appears at the right and casts a neat circular shadow on the Jupiter’s cloud deck. (NASA/JPL/University of Arizona)

By May 2004, Cassini was nearing Saturn and was within 30 million km (0.2 AU). Even from this distance, Cassini’s camera could see details on Titan’s surface, with twice the resolution that the HST had been capable of a decade before. It was, however, a very different view. Instead of being a disk, Titan was seen as a half disk. Cassini was approaching the Saturnian system from a little way in front of Saturn as it was traveling in its orbit, and so the direction of view from Cassini was almost at right angles to the Sun’s direction. Cassini took enough images to make a map better than the HST’s. These pictures required exposure times of some thirty-eight seconds, but the massive Cassini was steady as a rock, and there was no smearing of the images. The new maps showed what was by now a familiar pattern of bright and dark regions, the strongest contrast being around the equator between long, dark branched regions and the bright Xanadu. But how should they be interpreted?

Figure 3.12. Majestic planet Saturn looms, with her rings tastefully tilted, as seen from Cassini’s vantage point some 285 million km (2 AU) away in October 2002, a full twenty months before arrival. In southern midsummer, the planet casts a long shadow across the rings, and vice versa. Titan is visible at the top of the image. (NASA/JPL/Southwest Research Institute)

PHOEBE

On June 11, scientists got their first taste of the kind of close-up images they hoped would soon be pouring in. Cassini’s arrival date had been timed so the spacecraft, while on its way in through the Saturnian system, would pass close to Phoebe, the largest of Saturn’s outer moons, on June 11. Initially, the plans had been to fly by at a respectful distance of tens of thousands of kilometers, but when it emerged there was no good reason for not going closer (apart from a prudent margin of error to avoid any possible dust cloud close to the surface), Cassini was aimed to pass only 2,000 km away from this mysterious body—a thousand times closer than the distance from which either of the Voyager s had seen it.

Little Phoebe was always thought to be something of an interloper in the Saturnian system. While most of the satellites dutifully orbit in the same direction around Saturn, and very close to the ring plane, Phoebe’s distant orbit goes the other way, and at quite a tilt to the ring plane, suggesting that Phoebe did not originate in the disk of dust and gas in which the other satellites were formed like a miniature solar system. How Phoebe happened to get trapped at Saturn is a question for which nobody has good answers. Textbooks often suggested Phoebe might be a trapped asteroid.

Figure 3.13. Details started to appear on Titan from a distance of 30 million km on May 5, 2004. This image was taken by Cassini’s ISS camera in the same wavelength (940 nm) as the HST map in chapter 2. From Cassini’s viewpoint, Titan is only half illuminated. The northern high latitudes are in darkness. (NASA/JPL/Space Science Institute)

“That’s no asteroid” came back the results. The reality was, as ever, much more complicated and interesting. Although Phoebe looked slightly irregular and cratered, and some largish boulders were even visible in the sharpest pictures, it was clear that Phoebe was not built like an asteroid. The staggering pictures showed bright streaks in some of the craters, suggesting that darker material had slumped away, exposing brighter ice beneath. In fact, it seemed a lot like a quiet comet!

In terms of how easy it is for things to be flung around the solar system, the ragged edge of the Saturnian system is rather nearer the Kuiper belt beyond Neptune, with its population of icy planetesimals like Pluto and its cousins, than it is to the asteroid belt. Although some scientists talked of Phoebe being a captured Kuiper belt object (KBO), the reality may be even more exotic. It is possible that Phoebe is the last remnant of some population of objects that is now no longer there, expelled by the gravitational turmoil of the early solar system, a unique relic preserved in a Saturnocentric orbit like a wooly mammoth in permafrost. Not a KBO, not an asteroid, but something in between.

Figure 3.14. A map of Titan built up from ISS images during the approach phase, prior to Saturn Orbit Insertion (SOI). No details are visible north of about 30° north. The Huygens landing site is indicated to be near an equatorial bright/dark boundary, and the area to be observed during the T0 opportunity, just after SOI, is shown. The lack of details at southern midlatitudes is interesting, and is somehow intrinsic to either Titan’s surface or its atmosphere at this season. (NASA/JPL/Cassini Imaging Team)

Its composition, as measured by the VIMS instrument, included bound water, trapped CO₂, phyllosilicates, organics, nitriles, and cyanide compounds. The presence of all these compounds makes Phoebe one of the most compositionally diverse objects in our solar system. It also heralded the prospect that the composition of materials in the Saturnian system as a whole may be more than simple “rock and ice.”

As it sped past tiny Phoebe, Cassini was still on a trajectory around the Sun. After this encounter in the outskirts of Saturn’s extended family at a distance of 11.8 million km, it was a further nineteen days before Cassini closed in far enough for the Saturn Orbit Insertion maneuver, just 100,000 km from Saturn’s cloud tops.

Figure 3.15. Close-up image of Phoebe on June 11, 2004, from a distance of about 14,000 km. The surface is pockmarked with craters, but appears to have some bright streaks in crater walls, perhaps indicating the presence of ice under a coating of dark material. (NASA/JPL/Space Science Institute)

IN ORBIT AT LAST

When the drama of the SOI, with which we opened this chapter, was over, there was general relief all around. Only now was it truly possible to say that Cassini had arrived safely. However, some days later, it emerged that not quite everything had gone according to plan. A sensitive channel on Cassini’s magnetometer was to have been switched on to make precision measurements of Saturn’s magnetic field. It would have been a key observation because this was the closest Cassini would ever get to Saturn, and so the best chance of resolving fine structure in Saturn’s apparently symmetric magnetic field. But among the millions of instrument commands, this one had an error: the day number was wrong by one. This sort of error causes no engineering risk, so checking in the spacecraft testbed raised no warning flags. But the scientific loss was profound.

But that small setback was just a detail in the bigger picture. And the pictures themselves were not just big—they were spectacular. This was also one of the closest encounters with the rings in the whole mission. Cassini’s imaging team leader, Carolyn Porco, whose own scientific interests centered on rings, was thrilled. Reacting to images showing textbook density waves—the gravitational signature of Saturn’s moons in the dynamics of the ring particles—she declared them “absolutely mind blowing. Look at that. Ooh. . . . It’s almost everywhere you look here, you can’t miss one. They’re just all over the place.”

As Cassini shot past the rings at 15 km a second, many of the camera exposures had to be only five milliseconds long to prevent the relative motion from smearing the images. But that was plenty to reveal the splendor of the rings. Perhaps the most impressive single image was one of the Encke gap, only 300 km wide, bracketed with density waves that wrapped up at the edge of the gap to give it a beautiful, scalloped appearance.

As Cassini arced away from Saturn, again toward the south after SOI, Titan lumbered along above in its orbit in the ring plane, giving Cassini a bird’s-eye view, albeit a distant one (300,000 km away), of Titan’s south pole. Since this circumstance provided a good opportunity to reconnoiter Titan, the Cassini scientists decided to treat it as an extra, preliminary Titan flyby and named it T0.

An engine firing (a “cleanup burn”) to fine-tune the trajectory after the large orbit insertion burn was scheduled for July 3, but was canceled; the arrival had been precise enough for the cleanup not to be needed. Then, for the next several days, communications with Cassini were difficult or impossible, as expected, because Earth and Saturn entered conjunction. This event, where Saturn and Earth are on exactly opposite sides of the Sun, occurs every twelve and one-half months. Not only is the Sun directly interposed between Earth and Cassini for a short while, but for several days on either side of conjunction, the Sun (which is a strong source of radio noise) lies within the beam of Cassini’s antenna. Making sure no dramatic events like engine firings or Titan flybys occurred during conjunctions was just one of the many rules that the Cassini tour designers had to follow while weaving their tapestry of orbits around Saturn.

Figure 3.16. An image acquired just after Saturn Orbit Insertion of the Encke gap in the A ring. Faint ringlets are seen in the gap, which proves to have a scalloped edge due to the gravitational effects of one of Saturn’s moons. The spiral structures are bending and density waves. (NASA/JPL/Space Science Institute)

The first orbit Cassini completed around Saturn was its longest—a period of some four months—with the spacecraft still only loosely confined inside Saturn’s gravity “well.” Like a comet’s orbit around the Sun, this orbit was very elliptical, and without intervention would take Cassini perilously close to Saturn and its rings. Something had to be done. And so, in late August as it approached apoapsis—the farthest point in its orbit from Saturn some 150 R_s away (Saturn radii; 1 R_s is about 60,000 km)—Cassini fired its engine again for about fifty minutes to raise the periapsis (closest approach) of its orbit to a safe altitude. The burn, followed by some fine-tuning a couple of months later, also set Cassini up for its first close encounter with Titan.