Perhaps more than any other kind of warfare, naval warfare incorporates a wide array of stealth technologies. The seas are so vast, the ships so small in comparison, that merely finding an enemy force has always been a major undertaking in naval warfare. One reason that so many historic battles have been fought within sight of land is that only in such places did captains have a fair chance of finding their enemies. German wolf-pack tactics were designed largely to find convoys in a vast ocean.
Now, of course, much more is possible on the sensing side. To some extent, radars and passive detectors sweep out large areas of the sea, so that ships are far more detectable than in the past. Once wide-area detectors have found a ship, much also depends on whether the sensor on board a fired weapon, such as a missile or a torpedo, can reacquire it and score a hit.
Advocates of low-observable designs seek to negate these advances in sensing, either at the wide-area or the weapon end. Over the past few decades, first aircraft then surface ships and even missiles were radically redesigned to reduce their observability, particularly by radar.
Three complementary approaches have been used:
Firstly, it is possible to reshape an airplane, ship or missile so that most radar signals are reflected away from the observing radar instead of returning to it. This approach relies heavily on powerful computers capable of simulating radar returns from a given shape. Its great advantage is that the low-observable object need not be made of any special material. On the other hand, much depends on the aspect the object presents to the radar, since there are always some aspects at which reflection is considerable. For example, an airplane may have a low radar cross-section from the side, but seen from above or below it may offer a massive radar target. Shaping may also limit performance. One great advance claimed for the F-22 fighter is that it enjoys a low radar cross-section without paying the sort of performance price which is evident in the earlier-generation F-117 and B-2.
Secondly, radar-absorbing material can be used, either as a covering or throughout the structure. Typically, such materials cover a limited frequency range. In the case of the F-117, reportedly coverings can be changed to suit the expected enemy electronic order of battle. Coverings may make even a poorly shaped platform reasonably stealthy. For example, under a programme called Outlaw Bandit, during the 1990s radar-absorbing material (Ram) was reportedly applied to vertical surfaces onboard some Spruance class destroyers and Perry class frigates, drastically reducing their radar cross-sections over particular radar bands. Other navies, including the Soviets, had their own forms of Ram, as observers noted from time to time. One incidental advantage of such treatment is that it can be invisible. During the late 1980s, for example, it was claimed that some Nato navies simply applied Ram to the sides of their frigates' hulls, dramatically improving their survivability without any outward sign. The Swedish Visby class corvette, the hull of which is made of non-reflecting composite materials - essentially carbon fibre - is probably the most extreme example of this approach.
Lastly, ships still need electronic apertures, often in the form of revolving high-gain radar antennas. One would imagine that any such antenna would automatically add enormously to the ship's radar cross-section. Hence, the third stealth measure, tuning. A radar reflector or radome can be tuned to a very narrow radar band. To signals outside that band, the radar has a very low gain; in effect, its antenna is invisible. For example, the Exocet Block II apparently gains an important degree of stealth by tuning its radome in such a way that hostile radars do not see the high-gain antenna inside. It is then vital that the band to which the antenna or radome is tuned remain secret (there is usually some means of de-tuning in peacetime to ensure that this stealth measure remains undetected until it is needed). Even so, the narrowness of the usable band may be a wartime problem, frequency hopping being a major anti-jamming technique.
Stealthy active radars, such as the Thales Scout and the Bofors Pilot, are clearly related to these measures. It may be less obvious that many conventional air search radars are now stealthier, thanks to changes in waveform. For example, adopting pulse-compression drastically reduces the peak power that a search receiver detects. Pulse compression generally entails a systematic change in frequency, which a receiver could be designed to detect, but digital control should make it easy to use random frequency changes within a pulse. Radars which use phase coding instead of frequency coding are already very difficult to counter-detect. Similarly, many air search radars, such as the SPS-49, have been modified for something close to Doppler operation, mainly to pick up close-in air targets. However, that translates to, given a fixed average power, the radar must use more numerous, hence lower-powered, pulses - which, again, will be more difficult to counter-detect. These measures are now so common that they are no longer associated with stealthiness, but have very drastically reduced the value of search receivers , particularly the simpler ones.
Netting radars and other sensors also have a stealth aspect, since a ship within a net may be able to form a usable tactical picture almost without emitting. That was certainly one advantage the US Navy hoped to gain when it adopted Raytheon's CEC, the Cooperative Engagement Capability. Current US interest in network-centric warfare also has a stealth aspect, in that so much of a ship's sensor picture will come from external sources, whose radiation cannot, by definition, reveal the ship's position. However, the US Navy has now adopted a policy under which all long-haul communication will be in Internet format, and that in turn requires continual handshakes between a receiving ship and the external network. The expectation is that the channel back to the network can be made stealthy enough not to cause problems.
A Matter of Balance
Clearly, much depends on just how much a ship's signature is to be reduced, and against which sensors. Stealth presumably encounters limits at both ends of the frequency spectrum. At long wavelengths (e.g., metric and beyond), a radar target's shape has little impact; the whole object resonates to cause reflection. Fortunately, for those developing stealthy platforms, surface and airborne metric-wave radars have been less than effective. At least a conventional radar operating at long wavelenths tends to suffer from multipath reflection from the surface, so that the lowest clear beam is well removed from the water (which is why microwave radar was so important for World War II naval applications). Much the same applies to an airborne metric-wave radar like the early British ASVs and some wartime German sets. Again, microwave radar, which really does respond to target shape, has been far more effective.
There is, however, an important caveat. In the 1980s, a British firm, Marconi, began to advertise high frequency (HF) surface-wave radar, both for ships and for shore installations. It exploited the fact that the sea surface and the ionosphere in effect formed a waveguide. Ships could be detected out to about 180 nm and aircraft to 250, and in neither case did shaping seem to have much impact. Moreover, it was difficult to imagine a coating which would be effective at very long wavelengths. All of this having been said, Marconi clearly did not enjoy much commercial success. The only known HF surface-wave radar currently in use for what amounts to coastal defence is a Raytheon system installed by Canada and used to catch poachers operating off the Grand Banks. And yet it would seem that such radars can entirely negate the sort of stealthy techniques which might well defeat airborne or conventional coastal surveillance radars, which operate in the microwave bands.
When Marconi was advertising HF radar as a solution to coastal surveillance, it also pointed out that such radars could be installed onboard ships. The US Navy financed a study of just such a radar for shipboard missile defence, its advantage being that it could see beyond the ship's horizon. At the time (1998), signal processing requirements were too great, and the demonstration project (which was not completed) envisaged acquiring data but processing it ashore afterwards. At present, BAE Systems is trying to interest the US Navy in reviving the project. One selling point is that processors are now fast enough to place the entire HF radar on board ship. Possibly the same radar would detect even stealthy surface craft.
It is not clear whether an airborne version of HF radar could be developed. It is tempting to say no, because the essence of the scheme is that the transmitter is close to the surface of the water, but then again a coastal radar site is not necessarily any lower than a low-flying airplane or missile.
At the other end of the spectrum, it seems unlikely that a ship can be shaped so precisely as to defeat very short-wave radars, whose signals may pick up minor variations in surface finish. Indeed, one question in stealth by shaping must be whether the usual hull wear (causing "tin-canning" or the "hungry-horse" effect) will tend to defeat attempts to achieve stealth simply by careful shaping.
At one end of the spectrum, there are relatively simple measures which can drastically reduce a ship's radar cross-section, particularly against microwave radars such as those used by aircraft and, at present, by missiles. The vertical parts of a conventional hull form corner reflectors with the water, so stealthy surface ships generally have their sides inclined away from the vertical. Glassed-in scuttles open internal compartments, whose bulkheads and floors form corner reflectors, to impinging radar signals, and in conventional designs large portions of the superstructure are vertical, hence form easily definable corner reflectors. Topside clutter can trap and reflect radar signals, so that even if a ship's basic design seems to show attention to stealth concepts, that performance can be ruined. That was notoriously the case with the old Soviet nuclear cruiser Kirov, which was initially described as a stealthy ship.
Today, the shapes of stealthier ships are well known. Glass portholes can have gold wire worked into them so that they reflect radar signals and continue the overall stealthy shape of the hull. A more extreme version of this measure is the chain covering of boat enclosures on board French La Fayette class frigates. The sides of the superstructure can be slanted back away from the vertical, so that no corners are formed. Bulwarks can protect interior spaces (which may be cluttered) from signals approaching at near-horizontal angles, e.g. from missile seekers (as in recent Meko frigates). This type of sloping can even apply to the upper part of the hull. A US-designed missile boat planned for Egypt showed a backwards-sloping plate at the upper part of her bow.
Clearly some protrusions above deck are unavoidable, and gun mounts seem to pose special problems. The Swedish Visby features a carefully shaped gun house and a retractable barrel. The latest British and US gun mounts (4.5- and 5-inch guns) have also been shaped for reduced radar cross-section, but the barrels are too large and space below decks too limited to adopt the Swedish technique. Instead, the barrel is wrapped in Ram. The most extreme version of such wrapping is clear in the US Advanced Gun System, the 155-mm gun planned for the new DD(X) land attack destroyer, in which stealth was clearly a major design goal.
As the French observed in the La Fayette, it was soon obvious that the ship's crew itself becomes a collection of radar reflectors, since personnel above decks are clutter. The conclusion drawn was that the crew should spend most of its time below decks, the ship operating (from their point of view) like a submarine. It is not altogether clear that such solutions are acceptable from a morale point of view. However, crews are already used to being closed up at action stations for combat, and thus such a ship might normally operate in a less-stealthy mode, to then transition to the stealthier mode when combat seems imminent.
Roll and Pitch Give-away
Very probably the key problem in any attempt to shape a ship for stealth is that, because she is waterborne, a ship cannot always present a radar with the same aspect: she rolls and pitches. For example, even if the hull is carefully sloped inwards, at times, the ship will roll away from the radar, briefly presenting something closer to a vertical surface (which creates a corner reflector with the water). Models of the winning Northrop Grumman Ship Systems' DD(X) design shows hull sides sloped away from the vertical and rounded into the deck in a revival of the old tumblehome configuration. If most of the above-water hull is curved into the deck, the hull never shows much vertical (corner-reflector) area. Another approach to this idea was the experimental Lockheed Martin Sea Shadow, a catamaran resembling a seagoing pup tent, with flat sides meeting above water. They were sloped so radically that it was most unlikely that, short of sinking, the ship would ever present a radar with a corner reflector. The price paid was that the ship has no usable deck area whatsoever.
The most extreme stealth hull forms clearly extract a high price. The DD(X) design is described as 'wave-piercing,' which means that the designers have deliberately foregone the sort of buoyancy which tends to lift conventional ships over waves. Their motive is clear; they want to minimize ship motion because any motion presents an observing radar with opportunities to pick up the ship. Similarly they will want to minimize rolling motion, and they will have to accept that waves will often break over the ship's deck. That may be more acceptable than in the past, because with the end of the Cold War, the US Navy is less likely to operate in extreme conditions. For that matter, the ship will clearly be stealthiest at one displacement, with everything above water properly curved and everything less curved safely below water. Such a displacement can easily be maintained, but at a cost in ship size due to the compensating ballast tanks needed.
Critics of such extreme measures have pointed out that abandoning many of the benefits of buoyancy, such as the lifting bow, may prove dangerous. In a conventional hull, or even in a moderately stealthy one, the sides of the ship curve outwards, so that upper decks are larger than lower ones. If the ship suffers underwater damage, as she sinks into the water her water plane area increases, and more and more water has to be taken on board to keep sinking the ship. Tumblehome reverses that equation. If the ship has massive ballast tanks and some means of blowing them, then this may be irrelevant, as once the ship has suffered damage she may simply reduce her displacement and solve the problem. Whether that will work in practice will probably not be clear until the DD(X) is built.
Stealth Has a Price
There is no question, then, but that a ship's radar signature can be reduced quite sharply, as long as a high price in construction and in operation (e.g., all crew below decks, no deck area available for adding equipment) is accepted. What is open to question is whether extreme forms of stealth, as represented by the Sea Shadow and probably by the DD(X), are worth what they cost. Current advocates of such ships can point to their ability to hide from microwave radars, which aircraft use to find their maritime targets. Will that ability survive into the future? If the answer is no, then it may be argued that stealth ought to be reconsidered. Which will be more important, the ability to evade initial detection by a missile-carrying airplane, or the ability to deal with the missile? Since the two sensor types will probably operate at different frequencies, the issue is by no means moot.
A ship in a seaway is surrounded by the radar clutter created by waves. On average, the waves have a particular radar cross-section per unit area, so in theory a ship can hide if her radar cross-section is smaller than that of the surrounding waves. In fact, as those who do sonar signal processing know, anything man-made is an island of regularity in a sea of randomness, and signal processing seeks to exploit that fact. Sonar signal processing routinely finds submarines with signatures smaller than ambient noise, precisely because the noise is random. The implication would seem to be that at some point an airborne radar can or will have enough processing power to do the same.
It also seems likely that existing HF radar can locate any stealthy ship approaching to within about 180 or 150 nm of a coast, if only the government of that coast cares to buy it. An elementary application of network-centric ideas is to use one sensor to cue something else: say, to use a fairly accurate HF radar fix to cue a missile-bearing airplane. Knowing a priori that a ship is within a small search area greatly improves the chance of finding the ship.
Of course, potential enemies may not bother to buy coastal surveillance systems such as HF radar. In that case, it is not clear whether they will buy sophisticated airborne radars either. It is possible that the failure of Marconi's attempts to sell HF radar was symptomatic of a wider lack of interest, among Third World states, in antidotes to foreign naval presence. In that case, merely staying beyond the horizon of conventional coastal radars may suffice to keep a warship largely out of harm's way. Conversely, aircraft which do detect warships well offshore prior to attacking them may have been cued by some external source of data which the ship's stealth may be unable to defeat. The USS Stark incident of 1987 may be a case in point; the Iraqi aircraft which attacked the ship was apparently vectored to within twelve miles of her.
It may be, then, that the important question is defeating anti-ship missiles, and the best criterion may be whether the ship's radar signature can be reduced below that of the decoys she can deploy. Clearly, much depends on the kinds of seekers with which missiles are equipped; most likely they will operate at frequencies higher than those of aircraft radar and the missiles will have less signal processing power onboard. There is one cautionary point: computing power keeps increasing at a very fast pace, doubling in eighteen months or less. Greater computing power makes possible not only better signal processing but also new tactics which may, incidentally, defeat stealth and other countermeasures. For example, one effect of radar netting via CEC is that a stealthy aircraft, which is visible only intermittently to any one ship of a group, becomes visible to all of them when they merge their radar data. CEC is possible because ships can accommodate powerful enough computers, but if computing power continues to rise so rapidly, at some point it will be extended to relatively inexpensive ships, or even to aircraft or missiles. The Russians already advertise missile-to-missile links in both their existing large SS-N-19 missile and in the new Yakhont bought by India. The links in question are primitive and are not even remotely comparable to CEC links, but what happens in a decade or so, when computers are more than a hundred times more powerful? In twenty years, when the factor will be well over 10,000?
Ships are designed to last a long time, and stealth is built into their configurations. One implication is that the level of stealth incorporated in a ship is unlikely to change much over time; indeed, wear and tear are likely to reduce it somewhat. A ship may well be intended to last thirty or forty years, and it may be a decade from conception to completion. A class may take a decade to build, in which case the total time elapsed between a forward-looking stealthy design and the retirement of the last ship may be half a century or more. During that time the ship's electronics will probably enjoy massive upgrades, and weapon systems will probably be replaced. Maybe it is useful to rethink whether her stealthiness will last at all, and whether the sacrifices it entails are worthwhile.
There may also be an important distinction between a ship intended to maintain her stealthiness well out to sea, like the French La Fayette or the American DD(X), and one like the new Norwegian fast attack craft, which can spend much of its time hiding inshore, where shore features may swamp radar returns. In the latter case it may well be possible to exploit shore features, and an onshore command system may help the craft do so. If so, future stealth may be exploited far more effectively by coastal defensive navies than by the oceanic navies trying to attack them.
Radar is not the only sensor able to detect surface warships. Thermal imagers are increasingly used as search sensors, on board aircraft and even in coastal defence. Compared to radar, they offer much shorter range, but on the other hand they provide easy identification of the target. Several anti-ship missiles carry infrared sensors, often complementing radar. In such a case, infrared offers several advantages. If used alone, it is passive, hence does not reveal the presence of the missile. If used in combination with a radar, it complicates any attempt by the target to decoy the missile. More generally, it seems to be more difficult to decoy an infrared sensor than a radar, because to produce sufficient total energy to overcome the ship signature, the decoy has to be very hot, yet a sophisticated (multi-colour, for example) thermal sensor can measure temperature.
Most thermal suppression measures aim at cutting the temperature, hence the intensity, of the plume the ship emits. One typical practice is to mix cool air with the exhaust. Note that a diesel produces a cooler exhaust than a gas turbine, and it can eliminate most of its thermal signature by venting underwater. On the other hand, a diesel generally has a substantial acoustic signature, which can be reduced only through elaborate measures such as rafting and hooding, and it does not offer the sort of compact high power normally obtained from a gas turbine. A future solution to the gas turbine exhaust heat problem may be to exploit some of the waste heat produced by the engine, as in the new WR-21 regenerative turbine. The advertised advantage of such an engine is usually greater fuel efficiency, but from an operational point of view, that efficiency may not be as valuable as the engine's reduced inherent thermal signature, coupled with the usual low acoustic signature of a gas turbine. In addition, plume suppression equipment is generally visible around the ships' stacks.
Ultimately it would be most attractive to adopt a powerplant generating much less heat. The only candidate on the horizon, however distant, is the fuel cell. If, a big caveat, fuel cells can be made more efficient and compact, then in theory they can replace current prime movers. Indeed, the possible adoption of fuel cells as the primary means of motivation may justify current interest in moving towards electric propulsion. From a signature point of view, a fuel cell would be attractive because it would be cold; its efficiency does not depend on the temperature difference the engine generates (as in the case of any heat engine). Its exhaust product is undetectable water. Again, whether fuel cells are ever adopted depends on many other factors; right now they just suffice as low-power propulsion for submarines (as in the German Type 212).
All of this ignores one fundamental point. A ship is generally a few degrees warmer than the surrounding sea, and an imaging detector picks up that difference. The argument in favour of reducing plume temperature is that any wide-area detector will not pick up an image. Instead, it will seek point sources of high temperature. Imaging comes into play only after the initial point detection has been made. On the other hand, a coastal defence operator may well scan the horizon looking for images, and in that case, plume suppression may buy very little. Presumably, current infrared missiles do look for a hot spot or a plume, and in their case plume suppression makes effective decoying practical.
For a long-range detector, say on board an airplane, there is at least one other important ship signature. Ships inevitably produce wakes, because to move through the sea they must push water to each side. The wake is a massive signature, extending miles abaft a large ship. Just like a radar signature, it has both detection and weapon aspects. The wake is often obvious visually, but it is also a massive radar object, detectable against the random pattern of waves. The principal way to reduce a wake is to move at low speed, but from an operational point of view that would be absurd. Among the claims for some new hull forms such as the Trimaran is that they can be tuned to reduce wakes, the wave trains from the different hull elements interfering constructively or destructively. Since the wake reflects an elemental physical process, it is difficult to imagine suppressing it altogether; the energy the ship puts into the sea in order to move through it has to go somewhere.
An important class of weapons, wake-following torpedoes, exploits the target's wake with a seeker. The main hope for countering such weapons, which are quite common, seems to be to use decoys to create false wakes (the torpedo may react to a relatively weak wake signature if it encounters that signature before meeting the real wake). That in turn requires a combination of long-range sensing and manoeuvring to avoid the torpedo when it emerges from the false wake. In any case, there seems no real hope of suppressing wakes sufficiently to eliminate the problem of the wake-follower. Nor does it seem likely that the ship's wake can be reduced to what a string of decoy explosions can create. To the extent that wake-followers are limited in their net speed up the wake, it can be argued that very high burst speed is a better countermeasure than any form of signature reduction. Such burst speed in turn would require very high installed power in a ship, and that might make the other forms of signature suppression described here much more difficult. That leads back to a fundamental question, which is how to balance different signatures in a design likely to last many decades.
Finally, there are acoustic questions, which are best saved for a separate discussion.