This book is intended as a concise but broad overview of the engineering, science and flight history of planetary landers and atmospheric probes. Such vehicles are subject to a wide range of design and operational issues that are not experienced by ‘ordinary’ spacecraft such as Earth-orbiting satellites, or even by interplanetary flyby or orbital craft. Such issues deserve special attention, and we have attempted to bring together in one place brief discussions of many of these aspects, providing pointers to more detailed (but dispersed) coverage in the wider published literature. This volume also draws heavily on real examples of landers and probes launched (or, at least, where the launch vehicle's engines were started with that intention!).

More than 45 years have passed since the first vehicles of this type were designed. To a certain extent some past missions, of which there are over one hundred, may now be considered irrelevant from a scientific point of view, outdated from an engineering point of view and perhaps mere footnotes in the broader history of planetary exploration achievements. However, we believe they all have a place in the cultural and technical history of such endeavours, serving to illustrate the evolving technical approaches and requirements as well as lessons learned along the way. They stand as testament to the efforts of those involved in their conception and implementation.

The landers covered in this chapter have the ability to survive an initial landing impact, which may send the vehicle rolling and/or bouncing across the surface, and then commence operations having come to rest in whatever orientation is finally reached. Most achieve this by means of airbags to cushion and dampen the initial impact and subsequent rolling/bouncing motion, followed by the opening out of a system of ‘petals’ to bring the lander itself to its proper orientation for surface operations. The Ranger seismometer capsules are the exception to this; their impact damping was provided by the balsa-wood shell and liquid-bath system surrounding the experimental equipment, and the orientation being achieved by means of the natural position of the equipment within its liquid bath.

Typical payload experiments for such landers include cameras, meteorological, geological, geophysical and environmental sensors for investigation of the landing site. While some can be body-mounted on the probe, others may require deployment by means of masts, arms or a rover. In the case of the Mars Exploration Rovers, the pod landing stage itself plays no further role once the rover has rolled off.

Pod landers are particularly suited to ‘network science’, where simultaneous seismological, meteorological or other geophysical measurements are made at multiple locations. Such a network was the aim of the NetLander mission a network of four Mars landers to be carried on the CNES-led Mars Premier mission. The mission was cancelled in 2003 towards the end of Phase B of the project, however.

The Mars Pathfinder mission began as MESUR (Mars Environmental Survey), a 1991 proposal for a network of as many as 16 Mars landers to perform network science (meteorology and seismology on distributed sites) using nominally inexpensive landers. One prominent approach to reducing the unit cost of the landers was to use a semi-hard landing approach with airbags rather than a retrorocket system. The landing system proposed was sufficiently radical that a technology demonstration/flight validation was designed, originally MESUR Pathfinder, on which work formally began in 1993.

With the loss of Mars Observer and the onset of the Discovery programme in NASA, the Pathfinder concept was ‘adopted’ by the Discovery programme, and became the most widely cited example of the ‘faster, better, cheaper’ (FBC) approach (see McCurdy, 2001). NEAR technically was the first selected Discovery mission, but took rather longer to be built and reach its target. Note also that there were other FBC programmes within NASA, including the Small Explorer Earth orbiters, and the New Millenium technology validation programme. The success of some non-NASA projects like the Clementine moon orbiter, which came out of the Strategic Defense Initiative (the ‘Star Wars’ programme) also set the stage for the FBC era.

As an aside, one viewpoint of the background to the development of Pathfinder is described in Donna Shirley's book Managing Martians (1998). Andrew Mishkin's Sojourner (2004) gives a more detailed but narrower view, of the rover engineering development specifically.

Among many early concepts for a Titan probe (e.g. Murphy et al., 1981b) it is not surprising that a Galileo-like architecture was envisaged. As initially proposed in 1982, the concept of the Cassini–Huygens mission was to be a joint effort between NASA and ESA, and NASA was to supply the Galileo flight spare probe, and ESA would provide an orbiter delivery vehicle. However, in many respects the Titan probe grew in scope and complexity, in part because of the international nature of the mission.

As the joint study progressed, the roles were reversed, and ESA studied designs for an entry and descent probe (Scoon, 1985). These studies led to some quite novel ideas (e.g. Sainct and Clausen, 1993), which in all probability would not have been explored had the probe development remained in the USA.

The probe changed from an initially spherical shell (the shape adopted by the Galileo probe) to a flatter design. This also opened up novel heat shield architectures, with options such as a beryllium nose cap and a jetisonnable carbon– carbon decelerator (although in the end, neither of these concepts was adopted and a more technologically conservative heat-shield design was used – a prudent measure given the novelty of this mission for ESA).

The mass budget (Table 23.1) deserves some brief comment. In broad terms the mass breakdown is typical (e.g. with 15% of the mass devoted to power systems), although the front shield is rather conservative.

Missions to small bodies differ from those to larger worlds because the low surface gravity means that an orbiter (or rendezvous) spacecraft can approach close enough to perform a surface mission while hovering (with little or no thrust) and the speed of a landing can be very low. This blurs the distinction between orbiters and landers, and may enable orbiter spacecraft to survive landing, as shown by the landing of NEAR on asteroid Eros. Low gravity also means that a landing vehicle may risk being lost entirely on rebound from the surface, or ejected by outgassing in the case of a comet nucleus. Anchoring systems may thus be required. On the positive side, the low gravity also makes it easy to achieve mobility by jumping, and to perform ‘touch and go’ surface-sampling manoeuvres (e.g. Yano et al., 2003; Sears et al., 2004). Most small bodies are highly irregular, and their gravitational fields can be challenging environments in which to navigate. Dust thrown up from the surface (whether from natural cometary activity or the action of a spacecraft) is another hazard. Many small bodies, particularly comets, are in elliptical orbits and so experience wide variations of temperature and solar power production with time and surface location.

Phobos 1F

The Phobos project involved two large Mars orbiters, Phobos 1 and Phobos 2 (Sagdeev et al., 1988; TsUP, 1988).

Each of the missions or spacecraft in this part has been selected for description in greater detail because they have faced and overcome an unusual challenge in their design and/or mission. Collectively, the seven case studies cover: atmospheric probes and surface/sub-surface missions; worlds with and without atmospheres; low and high gravity environments, and both static and mobile elements.

This part of the book provides a basic description, key data and a drawing for all planetary atmospheric or surface vehicles launched, or attempted, from the earliest examples to 2007. Key references concerning the design, payload and results of each craft or mission are given in each case so that the reader may find more detailed information elsewhere. For the payload experiments, the names in parentheses indicate the Principal Investigators (PIs) or otherwise-titled responsible experimenters. Details of the particular experiments and the results obtained (if any) can in most cases be found by searching publications authored (or co-authored) by those named.

The many vehicles are divided into six categories, reflecting the way in which they encounter an atmosphere or surface.

• Destructive impact probes (where the mission is intended to end with the vehicle being destroyed on impact with the surface). These probes are discussed only very briefly, since they are not landers yet do play a role in planetary surface exploration.

• Atmospheric entry probes (where the vehicle's design is driven by its mission in the atmosphere).

• Pod landers (where the vehicle is designed to land initially in any orientation).

• Legged landers (where the vehicle is provided with landing gear).

• Payload delivery penetrators (where the vehicle decelerates in the sub-surface to emplace a payload).

• Small-body surface missions (where the vehicle operates in a low surface gravity environment). These can include many operations that are possible in low gravity, and various types of surface element.

The diagrams in this part of the book were drawn using information gleaned from a variety of sources.

These landers use a system of legs to cushion the landing and provide a stable platform for surface operations. With the exception of the Venera landers and the forthcoming Mars Science Laboratory (MSL) rover, all legged landers have three or four legs with footpads, and retro-thrusters perform final braking before landing. This was not required for the Veneras, whose terminal velocity at the surface was low enough (~8 m s− 1) such that the landing gear alone was able to provide sufficient damping. The landing gear was toroidal and we thus consider it as effectively a single ‘leg’. Mars Science Laboratory is due to make use of the rover's wheels as landing gear. A key feature of legged landers is that they must be the right way up for landing – beyond some tolerable limit such landers would topple over and fail. This attitude control must be performed during descent, usually by thrusters. Only for sufficiently thick atmospheres, such as that of Venus, can aerodynamic stabilisation be used.

Beyond those described here, future possible legged landers include robotic and crewed lunar landers from the US, robotic lunar landers from China and Japan, and a Mars sample-return mission.

Surveyor landers

The Surveyor landers performed soft landings on the Moon, largely as reconnaissance of the surface for the later Apollo landings. For more details see the Case Study, Chapter 21 (Figure 18.1).

A sobering thought experiment is to contemplate a world without electricity. Not only is electricity exploited as a convenient means of delivering mechanical or thermal energy to remote locations, but electricity is vital in information transmission and in sensing and control. Although the first planetary missions contemplated involved launching to the Moon a vehicle containing flash powder with which it would optically announce its arrival, and some early spacecraft used clockwork timers to sequence operations, every mission actually flown has been electrically powered.

In this chapter we first consider the overall requirements on the probe's power system, and how these requirements favour the various means adopted to meet them. The power supply and storage possibilities are then discussed, with particular reference to planetary probes. A general reference for power considerations is the book by Angrist (1982).

It is instructive to consider the electrical power requirements of various household devices to place spacecraft requirements in context. A modern PC may consume perhaps 200 W; a laptop perhaps an order of magnitude less. The Viking lander ran on 90 W. The Huygens probe's batteries supplied around 300 W for about 5 hours. The Sojourner rover had a solar array that delivered a mere 15 W.

System requirements

The total energy requirement of a mission (i.e. its integrated power requirement) is the most fundamental parameter for designing the power system.

Some years ago professional colleagues suggested that a new edition of An Introduction to Astronomical Photometry could be useful and timely. The decision to act upon this did not come, however, until the warm and conducive summer of 2003, in the stimulative environment of north-west Anatolia, once home to great forefathers of astronomy, such as Anaxagoras and Hipparchos. The former set up his school at the surely appropriately named Lampsakos, just a few miles from where the present authors are working: the latter hailed originally from what is now the Iznik district of neighbouring Bythinia. Eudoxus too, after learning his observational astronomy in Heliopolis, moved back to Mysia to found the institute at Cyzicus (today's Kapu Dagh), while Aristotle's thoughts on the heavens must have also been developing around the time of his sojourn in the Troad, after the death of Plato. In such surroundings it is difficult to resist thinking about the brightness of the stars.

But that was just the beginning. It quickly became clear that the proposed task could not be lightly undertaken. There were at least three main questions to clarify: (1) what branches of modern astronomy can be suitably associated with photometry; (2) what level of explanation can be set against the intention of an introduction; and (3) who could become involved with what aspect of the subject? An approximate size and scope were originally based on the model of the first edition.

Introductory background

‘Starspots’ are not a new notion. There was a time when starspots were offered as a general explanation of stellar variability. From that extreme, in the nineteenth century, attention to the hypothesis had dwindled away almost completely after the development of stellar spectroscopy, until, in a well-known pair of papers dealing with the cool binaries AR Lac and YY Gem in the 1950s, the photometrist Gerald Kron revived it. As events have turned out, there is now a good deal of evidence to support Kron's conjecture, for certain groups of cool variable.

The two stars that Kron referred to are good examples of somewhat different but related categories of ‘active cool star’ (see Figures 10.1 and 10.2) – stars of spectral type generally later than mid-F, i.e. associated with convective outer regions, and usually having a relatively rapid rotational speed. In the cases of AR Lac and YY Gem, both close binary systems, this rapid rotation is a consequence of tidally induced synchronism between rotation and orbital revolution. The M type dwarf components of YY Gem are typical flare stars, characterized by Balmer line emissions that on occasions become very strong, reminiscent of solar flares, but with much greater relative intensity. AR Lac also shows ‘chromospheric’ emission lines, but its particular configuration, G5 and subgiant K0 stars in a ∽2 day period binary, places it as a standard RS CVn type binary.

In Chapter 5 we considered the various oscillation modes that a star can show when perturbed in some way. Many types of star are observed to pulsate. Perhaps the best known are the class known as Cepheids (after the prototype star δ Cephei). These pulsate primarily in a radial mode, so that we observe the full amplitude of their oscillations. The period of these oscillations is directly related to the stellar radius, just as the pitch of an organ note is related to the length of the pipe producing it, the lowest notes coming from the longest pipes. Finding the star's radius in this way gives a measure of its brightness, so that simply by comparing the periods of two Cepheids we know their relative brightnesses. Thus, with careful calibration, measuring the period and the apparent brightness of a Cepheid gives its distance. As Cepheids are bright stars they can be seen in very distant galaxies and therefore can be used to produce a distance scale of great importance in astronomy.

However, most stars do not show pulsations of readily observable amplitude. This must mean that for most stars any oscillations set in motion by perturbations received during their lives, or in the process of formation, have long since been damped out in some way. The existence of these damping processes in turn implies that if a star does pulsate, an excitation mechanism for these pulsations must be operating.

In this chapter we proceed to more general effects in close binary light curves, with a wider sample of data sets. Although through most of the twentieth century close pairs formed a distinct subset of double star research, engendering its own data, purposes, methods and outcomes, in present times this somewhat artificial separation, mentioned at the beginning of the book, is being bridged. The time is in sight when photometric data can be more easily joined with astrometry for fuller analysis and information retrieval from close binary stars. To proceed with this, we need first to consider the overall geometry.

Coordinate transformation

Ambiguities are possible in bringing data on close binary systems into the conventional framework used for double stars, since, for example, the ‘longitude of periastron’ used for the radial velocities of spectroscopic binaries is normally measured inward from the plane of the sky and not the line of nodes in the equator, as in the standard 3-dimensional specification for astrometric binaries. If, as is usual for close systems, we use the local plane of the sky as the reference, the nodal angle Ω should still refer the line of intersection of orbital and sky planes to the equatorial coordinate system. This angle has no effect on radial velocities or photometric effects, though it remains a basic parameter of a binary system.

Introductory background

The eclipsing binaries and spotted stars discussed in the previous three chapters still represent only about a quarter of all variables. The largest class of variable stars are those having some inherent physical variation in luminosity, as distinct from an effect of geometry. They are often referred to as pulsating, sometimes vibrating, stars: words suggesting the physical cause of the variation. Although, as noted before, all stars would vary over a sufficiently long timescale, an appreciable intrinsic variation of luminosity accessible to human inspection implies a very short period against general stellar time frames, giving perspective to such terminology. There are examples whose light pattern repeats in measurably the same form for many cycles, with a periodicity of comparable constancy to that of eclipsing variables. Others show varying degrees of chaotic behaviour. In ‘irregular’ cases, the light level wanders up and down with no pattern or predictability. But many show quasi-periodic variations of a ‘semiregular’ nature. Some examples of the different light curves are shown in Figure 11.1.

Spectroscopy shows that the variations in apparent magnitude are linked with changes of radius. By studying the Doppler shifts of absorption lines in the regular cepheid type variables (prototype δ Cephei), it was deduced that the star oscillates inward and outward in the same period as the brightness cycle, the star being faintest not far from, but somewhat before, the time when it is smallest.

Variable stars and periodic effects

The times of observed features, usually minima or maxima, on the light curves of variable stars (single or binary), can often be accurately determined. Many classes of variable star have a repetitive, cyclic behaviour regarding such features and time, which is understood in terms of their basic physical properties. Thus, timing data are applied to measuring rotation, pulsation, or orbital periods (P) of the different kinds of variable. It follows from angular momentum considerations that any change in the velocity field and/or mass distribution within or nearby a star, or binary system component, will cause a change in rotational or orbital motion. There will then result shifts of corresponding minima (or maxima) times that have relevance to physical changes of the object in question. The ‘O – C’ method is commonly used in measuring shifts in the observed times of photometric features.

The O – C method

Defining terms: The O – C – ‘observed minus calculated’ datum (regarded as a single item) – indicates any time difference between the epoch of an observed phenomenon, presumed periodic, and its prediction, based on an ephemeris formula. Eclipses are a typical example. In this case, O – C analysis often allows identification of the possible cause of period variation. These may arise from orbital eccentricity effects, mass loss from the system or exchange between the components, magnetic effects, the presence of a third body or other reasons.

The normal unit of time for this context is heliocentric Julian days.