The salient question, then, is how to achieve design and operational flexibility and undersea superiority at a reasonable price, in acquisition and lifetime operations, maintenance, and upgrade costs.
The advent of a new class such as the New Attack Submarine (NSSN) is a major opportunity for new initiatives. Because money is tight, however, a choice must be made among different technologies, partly on the basis of expected payoffs. For example, if it seemed likely that enormous strides would be made in hydrodynamic performance as a result of a radical change in hull form or propulsion systems, it might be argued that several different hulls or powerplants deserve full-scale testing in prototype form. That expense, in turn, would limit any major initiatives in the combat system inside the submarine.
As it happens, there is good reason to believe that improvements in hydrodynamics (speed vs. horsepower), reactor efficiency, and quieting would be prohibitively expensive, at least in the near term. Also, no marginal improvement in these factors is likely to contribute directly to the mission flexibility that uncertainty demands. On the other hand, the character of the combat payload combat system, sensors, weapons, special-operations troops, etc. - inside the submarine determines how easily and to what extent it can be modified to take on new missions. For that matter, combat-system design determines the extent to which new technology can be inserted into a new submarine to improve its overall warfighting performance.
Numbers versus Innovation
Some observers have perceived a certain stodginess in U.S. submarine designs: the Los Angeles (SSN-688) class, seen as no more than a reengined Sturgeon (SSN-637), was kept in production for an unconscionably long time, while the Soviets always seemed to be introducing new and exotic submarines. The reality was rather different. For example, Admiral Alexander V. Gorbanov, Deputy Commander-in-Chief of the Russian Navy, noted that the Soviet Navy incurred great configuration management inefficiencies and logistics support costs for what in retrospect were ephemeral benefits from rapid technology introduction and numerous classes. 1
To get numbers, the U.S. Navy had to keep producing a standard design, applying incremental baseline improvements so as not to disturb production too badly. In fact, the Los Angeles class of 62 submarines is at least three distinct classes, with advances including radical improvements in sensors and combat systems and the installation of vertical-launching systems. To maintain its lead, the U.S. Navy had to recognize when a design had run out of margin for improvement, then accept a step change. That is hardly stodginess; it is deliberate policy intended to maintain sufficient force.
In early 1997, the Russians are barely hanging onto their fleet while the rest of their military disintegrates. During the Cold War, the principal advertised role of U.S. submarines was antisubmarine warfare; hence the submarineversus-submarine comparison that focused primarily on submerged speed, depth, and numbers of weapons. Very important but highly classified roles, such as intelligence collection, received little or no publicity. Yet surveillance and reconnaissance probably will develop into dominant post-Cold War roles. Existing submarines may be flexible enough for the first round of post-Cold War adaptation, but something new is needed for the long haul; the last NSSN could be in service until 2050.
We cannot know today whether some sort of greatpower conflict will return-in which case our submarines had better be superior in platform performance to the best of their future rivals-or whether the emphasis will continue to be on mission flexibility in places like littoral waters. Submarines are expensive, and the U.S. Navy seldom has had the luxury of developing parallel designs of attack submarines for different roles; one type of submarine must do virtually everything. 2
Apples to Oranges to . . . ?
The comparison with foreign submarines-which generally has meant Soviet/Russian boats-is almost always made in terms of platform characteristics: speed, diving depth, and silencing. Comparison is tricky not only because such data tend to be classified, but also because they are complex. For example, a submarine may be capable of very high speed on trials, but that speed may not be usable because of excessive noise levels. 3 It may be able to dive very deeply, but not to do that very often. 4 Silencing may degrade over a submarine's service life, particularly if maintenance practices are poor.
Platform characteristics are by far the easiest elements of submarine capability to describe, but in fact, in a submarine-to-submarine encounter, three other factors are equally vital to the outcome of the engagement: the ability to detect the other submarine (sensors), the ability to make tactical decisions based on that detection (the combat direction system), and the weaponry used to implement tactical decisions. None of the three is easy to describe, but each is critical to undersea warfare superiority.
Are we likely to be outclassed by some future supersubmarine? If we look at the sweep of technological history, we see each particular technology advancing along S-shaped (performance vs. time) curves. At the low end of the S, the technology is little understood, so enormous effort buys relatively little improvement in performance. As comprehension dawns, progress is swift up the middle steep part of the S. Eventually, most of what can be done easily has been done, and progress slows. It takes a new technological revolution to start another cycle.
Although it is difficult to say where in the S we may be, it seems fair to say that in platform characteristics we are in the upper (slow) part. We still can get improvements, but generally at a very high price. In the case of silencing, we get jumps mainly when we recognize previously unknown elements of a submarine's signature; in effect, in these cases we start nearly from scratch.
That is not to say that progress is over, but anyone seeking improvement may have to be satisfied with some fairly subtle changes. For propulsion, it appears that the really dramatic changes since about 1960 have been in reactor core lifetime. We now seem to be on the verge of a very significant improvement: the NSSN will not need refueling at any time in its 30-year life. If the submarine need never refuel-the single largest cost item in operating a submarine-then it need never be opened up for a big refit-as long as other elements of its design are amenable to upgrade without major physical changes to the hull. 5
Beyond 2001: The Computer
It may, then, be fair to say that for some time to come, it will be more profitable to pour money into the submarine's payload-its weapons and combat systems, which like computers in general are far from their ultimate capability. The nature of the weapons and combat systems often determines just how flexible a submarine can be. Also, it is in the area of total payload flexibility that a fresh approach, embodied in a new submarine, can shine.
In 1968 the Los Angeles class was designed with two important features. The platform feature - higher speed enabled the submarines to operate with and provide ASW protection to aircraft carrier battle groups, and this alone apparently justified a new submarine class. However, there also was a vital combat system innovation-the shift from analog to a digital system. Going digital and programmable meant buying vital capability and flexibility. The new digital system could handle a far more complex tactical situation than could its analog forebears, a breakthrough so important that Los Angeles-type digital systems were later retrofitted into Sturgeon-class submarines. In overall combat effectiveness, the digital combat systems had far greater impact than the increase in speed.
On the other hand, the 688s' central computer had limited capacity. Anytime a new sensor or other element (such as a vertical launcher) was added, it had to be hard-wired to the central computer complex, and added considerable weight-and submarines in general have limited weight reserves. Although the programmable digital system in a Los Angeles was more flexible than its analog predecessors, the central computer system was not too adaptable to radical changes in sensor suite or, for that matter, to new missions requiring new kinds of hardware. Change was possible but expensive.
The central computer executes a series of different functions in sequence, timesharing its central processor(s). Because the different functions are tightly connected, it is relatively difficult to rewrite software to accommodate new equipment, such as new missiles. The more powerful the computer, the less tightly the software must be written, but time-sharing still makes change relatively difficult and limits overall system performance.
In the naval world, the solution is to unburden the central computer by drawing out some of its functions, in effect distributing combat-direction functions to several separate machines. This allows the central computer to execute its remaining functions far more efficiently. For example, in the surface ship antiair combat-direction systems, unloading by itself increases the number of tracks by a factor of about eight. Distribution also makes it easier for the system to handle the failure of any individual element, which might be particularly important for a submarine spending substantial time far from home in a hostile environment.
The first fully distributed system in U.S. Navy service is BSY-2 in the three-ship Seawolf class. Although driven by the desperate need for more computer power, distribution also makes it easier to modify system software for new requirements and new equipment, such as unmanned underwater vehicles. That makes Seawolf a giant step toward the fully flexible submarine of the future that can switch functions without much physical change.
The submarine of the future must be able to incorporate and exploit rapidly changing computer technology. When the baseline Los Angeles was designed, military computers generally were more advanced than their civilian counterparts. In addition, computers were not developing at breakneck speed; so it was reasonable to imagine that a UYK-7 computer installed in 1975 still would be modern and maintainable 10 or even 15 years later. Now any such conception is ludicrous. Computer generations last 18 months or less, and the military has been forced largely to abandon special-purpose computers in favor of adapting civilian equipment. It seems unlikely that any particular chip will be available throughout the life of any given ship or system.
This reality has promoted the expanded use of commercial off-the-shelf technologies (COTS) and opensystem architectures, which are the foundations of the Navy's Acoustic Rapid Commercial Off-the-Shelf Insertion program for the improved Los Angeles, Seawolf, and NSSN fleet. The program's goal is to transform existing submarine sonar systems (AN/BSY-2, BQQ-5, and BQQ-6) to an open-system architecture that will provide the entire submarine force with a common sonar system. COTS and open-system architecture will enable rapid updates to both software and hardware on an annual basis and allow computer power growth to keep pace with the commercial industry. The envisioned multipurpose processor is to have as much computing power as the entire 688/Improved 688 fleet combined, and will allow the use of algorithms previously beyond the reach of the "legacy" systems.
The clear implication is that openness to change is essential. Without it, a foreign submarine designed a few years after the NSSN might be able to accommodate a far more effective combat system.
Civilian computers often are shunned because they cannot meet military standards (such as living amid the stray electromagnetic interference common on board ship). The NSSN's design incorporates a new modular isolated deck structure, which hosts structurally integrated enclosures to protect computerized combat systems and accommodate commercial equipment, without the millstone of military specifications around the designers' necks. The isolation element helps protect the combat systems from shock and also helps silence the submarine.
Exploitation of future, more powerful computers supporting new kinds of sensors and weapons requires more. Accordingly, the NSSN incorporates a new asynchronous data bus, comparable in concept to the Internet, that should provide the sort of capacity required. Ultimately, the data bus can be the bottleneck in the combat system. Because it is buried in the hull structure, it also has been a very difficult element to change during the submarine's lifetime. Without the new data bus, the physical modularity so carefully built into the submarine would have been far less valuable. The combination of the data bus and the modular internal structure of the submarine set it apart from previous designs, which cannot easily be rebuilt to match the improved capabilities of the next-generation submarines.
There is always a choice between building performance into the platform and building it into "smart" and "brilliant" weapons and remote platforms, such as unmanned underwater and aerial vehicles (UUV/UAVs), and other off-board systems. For example, for littoral warfare, UUVs may provide much of the shallow-water performance that might otherwise seem to require a new smaller submarine. Or-as promised by the successful tests with the Predator UAV from the Chicago (SSN-721) last summer-the submarine-UAV lashup can provide real-time intelligence and reconnaissance to support a broad spectrum of offensive operations. The rub is that to accommodate future UUV/UAVs fully, the submarine needs changes to combat and supporting systems, which are difficult to implement in most existing submarines.
UUVs are not the only example of a new capability. A fully distributed, open-architecture data bus combat system probably is uniquely adapted to the sort of changes now being envisaged, such as the use of the Army's tactical missile system to deliver brilliant antitank weapons in direct support of forces ashore.
Similarly, increased computer power can help a submarine exploit its inherent platform capabilities more fully. For example, there has been intense interest in tactical advisors that tell a submarine commander how to maneuver to reduce transient noise or to minimize exposure to an enemy's sonar or weapons. Certainly, a submarine with good platform characteristics but limited combat system computing power can be upgraded to superiority over a computer-heavy submarine with poor platform characteristics. In practice, however, such upgrades are difficult, and the difference in computing power will be likely to overshadow any difference in platforms, particularly if advanced sensors and weapons are taken into account.
Design and Operational Flexibility
The reason the U.S. Navy submarine force found the transition from 1945 to 1965 easier than many other submarine forces was because the big, rugged fleet boats built during World War II had the capacity to accommodate new kinds of sensors and weapons. Their sheer size which many other navies found ludicrous in 1940 - turned out to be a considerable asset. A less obvious asset was their extremely reliable powerplants, which needed surprisingly little postwar modification. In modern terms, these submarines had inherent design flexibility and a degree of modularity; standard hull/machinery elements could be combined easily with new features, such as new sonars and the new radar picket radars.
The greatest payoff for the NSSN will be in "payload": combat systems, sensors, and weapons. In many cases, new systems, such as unmanned underwater vehicles, can be accommodated within a standard NSSN hull. In others that may be more difficult. For example, sometime during the NSSNs' lifetime, the Ohio (SSBN-726)-class ballistic missile submarines will have to be replaced if the United States continues to embrace the need for a truly survivable nuclear deterrent. It is possible to build ballistic missiles that can be fired from a conventional torpedo or vertical-launch tube, but it seems certain that something larger will continue to be desirable. For the NSSN, design flexibility and modularity mean that it may be relatively easy to design, engineer, and build a new strategic submarine by inserting a missile section into a lengthened NSSN hull.
There are numerous new technologies competing for space and volume in the New Attack Submarine. Concentrating on flexibility and capacity in the submarine's warfighting payload systems plays to U.S. industrial strengths (computer hardware and software) and leaves open the possibility of buying much of the needed platform performance in the form of submarine-launched vehicles and a variety of off-board underwater, surface, airborne, and space-borne systems.
The fact that, initially at least, the NSSN will look much like its predecessors will merely reflect the reality that, in platform and propulsion systems design and engineering, we are high on the S-curve. It is what is inside the New Attack Submarine that will make all the difference.
1 Adm. A. V. Gorbanov, RN, "The Submarine Fleet of Russia," Naval Submarine League Review, October 1996, pp. 8-21.
2 The Los Angeles initially was justified as a carrier escort, but in fact it always was used as a multirole attack submarine. The U.S. Navy flirted with short-range coastal submarines in 1916 and again in 1938, but they were never built in any number. The main modern exception to this rule is the rise of strategic ballistic missile submarines; but the authors are convinced that there would have been no separate ballistic missile submarines had the technology (which appeared in the 1960s) existed in the 1950s to make Polaris a torpedo-tube weapon.
3 According to the Russian account of the Papa (Project 661) design in Morskoi sbornik, the ship made a record 44.7 knots, but at anything like that speed it vibrated so badly as to damage itself; and it was never fully operational.
4 The titanium used by the Russians to achieve great depths apparently creeps under pressure. That is, it distorts permanently each time the submarine goes to test depth. Thus the submarine can dive very deep to escape after attacking, but it cannot do so very frequently.
5 A Los Angeles must be refueled at about the 13-to-15-year mark, and early 688s are being retired partly because the cost of refueling is so high—$250-300 million. Anything that cut fueling dramatically would increase the percentage of time a submarine could be deployed. It may literally be cheaper to build and operate new long-life submarines than to refuel and continue to operate existing ones. The savings in refueling really are profitable only if the submarine never needs a major sensor/combat system refurbishment—if, for example, an open or modular design makes it easy to modify incrementally.
Dr. Friedman is the author of U.S. Submarines Since 1945 (Naval Institute Press, 1994), edits the U.S. Naval Institute’s Guide to World Naval Weapons Systems , and writes a monthly column, “World Naval Developments,” in Proceedings .
Dr. Truver is Executive Director of the Center for Security Strategies and Operations, TECHMATICS, Arlington, Virginia, and is a frequent contributor to Proceedings , including the annual “Navy in Review” and “Tomorrow’s Fleet” articles.