In any high-speed successor to LOS ANGELES-class SSN’s, it’s clear that the submarine design community needs to pay more attention to aircraft design concepts in solving the danger-ous snap-roll problem, which is not entirely dissimilar to dangerous roll-coupling effects experienced in early supersonic aircraft And the need for computer assistance in maintaining high-speed undeiWater maneuvering control is obvious. Henry Payne, m, has drawn our attention to these concepts with a call for action in two superb articles both in this magazine (January, 1988) and (with William P. Gruner) in Naval Institute Proceedinm; (July, 1992). May I suggest, however, the need to go even farth,er both in the exploration of potential aeronautical engineering parallels and in the consideration of roles for advanced artificial intelligence computers on submarines.
If we are to have the most effective – and cost-effective — attack subs, we need to make use of research and development findings already available to us from the aircraft industry. Aircraft and submarines both have to manage the fluid flow of the environments in which they are immersed. Although density, sound propagation speed, and other factors may differ, air and water still possess relevant similarities as support and propulsive media. Under the circumstances, we should consider a number of additional aeronautical engineering concepts that deserve brain-storming as possible performance enhancers in the areas of submarine propulsion, hull design, and tactical maneu-vering. To provide these faster, deeper diving submarines with essential c;J-I capabilities and full real-time maneuvering control under adverse combat conditions, we need to look to advanced computers for assistance in a number of monitoring and systems management areas.
Propulsion
In the area of propulsion, a 518″-depth boundary-layer bleed plate on the hull immediately foiWard of a shrouded propulsor intake might help boost propulsion efficiency by removing boundary layer turbulence and increasing the laminar flow potential of water entering the propulsive duct. At the next stage, intake (and exhaust) stator blade stages in shrouded prop designs would both improve intake flow impingement angle and minimize tell-tale exit turbulence. Regarding water mass flow entering and exiting the propulsive duct, aircraft conver-gent/divergent engine nozzle design concepts should be investi-gated as possible optimizers of flow through the propulsor duct system. Within the power section itself, increased efficiencies might arise from possible use of propeller blades with variable-pitch, variable sweep, and automatic digitally controlled mission-adaptive wing design technology (pioneered by Boeing and NASA) for shaping foil, camber, and sweep angle as a function ofoperating depth, propeller diameter, and r.p.m. Furthermore, a look at aircraft engine turbine blade convection cooling passage design could, within large sub prop blades, lead to internal fluid flow carefully engineered to exit at the tips in such a way as to minimize tip vortices and cavitation — thereby both improving propulsive efficiency and minimizing the sub’s signature and consequent vulnerability to ASW detection systems.
Hull Design
Potential for such increased propulsive efficiency is only a first step, one that needs to be combined with improvements in overall hull design if potentially synergistic effects are to benefit the whole system. For example, with sub commanders already spending far less time on deck (or even at the periscope), the functions of the sail should (as Messrs. Gruner and Payne suggest in their July, 1992 Naval Institute Proceedin~ article) be re-evaluated. With the periscope housed directly in the hull, the sail could be entirely eliminated. Benefits would include drag reduction, higher underwater speed, roll reduction while surfaced in heavy seas, elimination of snap-roll hull-sail coupling effects, and greatly reduced vortex generation and wake.
In terms of specific hull-shaping, perhaps we ought to be exploring radical dimpling of the stern quadrant – or, as on aircraft, installation of vortex generator minivanes – to detach the boundary layer for the sake of lowered overall wake drag (golf ball concept). In designing for reduction of interference drag at all fin/hull interfaces, perhaps even a look at aircraft fuselage area rule (originally pioneered by Whitcomb at NACA in the 1950’s) might be worthwhiJe. Transonic flight is clearly not involved, but might not there be some parallel benefits
accruing from minimized total cross sectional area at fin/hull (or sail/hull) interfaces? In water’s higher density fluid flow, such potential benefits might include: lower total drag; improved boundary layer control; and reduction of the interaction effects of speed, viscosity, and trailing vortices. Then, of course, designers need to look at stealth aircraft fuselage-shaping and materials technology to minimize the hull’s sonar, radar, and magnetic signatures.
Finally and more subtly, we need to consider laminar flow control at boundary layer separation points in order to reduce drag and increase the hull’s speed and stealth characteristics. Easiest to implement would be the intake boundary layer bleed plate mentioned above. Far more complex and potentially more beneficial might be suction slots of the kind researched by NACNNASA and Northrop in their Douglas WB-660/X-21 conversion in the early 1960’s; or possible adaptations of suction hole plates on the F/A-18, or now being tested on the perforat-ed pumped wing glove fitted to the two F-16XL prototypes. Clearly, such investigations would have to proceed concurrently with development of a compliant, sonar/radar-absorbent composite outer hull sheathing (on which both we and the Soviets/Russians have already done some research)- a sheath-ing in which suction slots could be machined, this same sheath-ing also chemically formulated to eliminate boundary-layer slot fouling by algae and barnacles. No small challenge, to be sure.
Tactical Maneuvering
Such design enhancements would obviously improve overall speed and maneuverability characteristics of a next-generation attack submarine, but for the quick (bordering on extreme) responses which may be required in combat, tactical maneuver-ing capabilities could be significantly increased were two major control-surface changes to be considered. First, keeping in mind snap-maneuvering air-to-air missiles, should designers not look into the possibility (after elimination of the sail) of replacing current diving planes and rudders with two cruciform sets of four control fins, each set mounted fore and aft? The mid-mounted bow diving planes would be coordinated with the all-moving vertical bow stabilizers/rudders, with four similar all-moving fins at the stem – all capable of both tandem and independent movement. Change of plane maneuvering at high speeds might thus be accomplished far more quickly, with more controllability and with fewer adverse side effects. Because such control surfaces could be smaller, they would in tum contribute to further drag reduction and speed enhancement.
Second, the external after-hull could incorporate spoilers, pop-up flap segments ringing the stem quadrant to serve as waterbrakes acting either differentially or simultaneously to contribute to change of direction and/or suddenly slower speed. Similarly, in shrouded-prop propulsion systems, aircraft engines’ clam-shell thrust-reverser concept could be explored for its emergency maneuvering potential in causing enemy surface vessels, attack subs, or even torpedoes to overshoot their quarry. In addition, both the fore and aft cruciform contrql surfaces might be designed to split into hydro-brake systems – as do control surfaces on many aircraft
Computer Monitoring and Systems Management
As Payne and Gruner suggested in their recent Proceedin~ article, it is clear that high-speed underwater maneuvering already calls for computer assistance in current sub designs. With any next-generation attack sub, however, a computerized artificial intelligence command center (AICC) should be a prerequisite – for additional reasons ranging from maintenance and navigation to damage control. Many aircraft (among them the F/A-18 and the prototype YF-22A) have been designed with automated, computer-controlled maintenance analysis capability
— including that of the computer system itself. Shouldn’t our most advanced submarine systems have the same potential?
Furthermore, in conjunction with the more advanced sensor systems coming on line, the AICC could provide ocean-floor mapping and contour-matching navigation and avoidance capabilities similar to those in the Tomahawk cruise missile’s guidance system, at the same time serving as the center of the tactical and strategic data system link with other friendly submarine, surface, air, or space combatants or sensors.
In approaching its quarry, an advanced attack sub could have the potential to shield itself from premature detection by utilizing active low-volume sonar stealth masking in the form of AI-generated matching of ambient background noise. Further enhancement of this cloak of near-invisibility could occur through AI computer-generated white noise masking in wave- lengths reciprocal to those of standard tell-tale noise sources on most submarines (coolant pumps, bearing vibration, cavitation, etc.). All the while, the AICC’s own active target-seeking system would be on full acquisition and homing alert, using passive sonar target/threat analysis {via sound and other radiation pattern anomalies measured against ambient back-ground sound, magnetic field, infrared signature, et al.).
Once contact has been made with a target or with threaten-ing ASW forces, the AICC could be put in control of the sub’s advanced anti-ship, anti-aircraft, anti-satellite active counter-measure systems. As the moment of engagement approached, it would provide firing solutions for torpedoes and cruise missiles {anti-ship or land attack) and would serve as controller of any defensive RPV decoys. And should the sub itself come under attack in this process, the AICC could, through the possible use of the unique properties of organic metal {and/or other) sensors, initiate activation of semi-automated damage control systems (see my “AI: What’s Our Obligation?”,~ TRUM. Spring 1988, p. 10).
It is clear that for such an Al-assisted system eventually to operate with maximum potential in all these areas, subs would need to be designed for modular substitution of subsystems incorporating normal evolutionary improvements in the spectral range and sensitivity threshold of all SSN sensors. It would also need to be able to absorb inevitable advances in basic data-bus system technologies in the areas of organic metals, lasers, and optronics.
Does it make sense, as the SEAWOLF program is being severely curtailed or even about to be canceled entirely, to consider building an attack submarine with even more advanced capabilities? Of course it does. It always will. But here’s where persistent questions of national and economic policy understandably arise. We’ve heard them before. If we build an attack sub with such capabilities, how many can we afford? And with the apparent demise of Soviet and Warsaw Pact threats, which is our most technologically advanced potential enemy? Is a simpler CENTURION based on a less forward-looking design philosophy sufficient for our needs? On the other hand, how politically stable is Russia — and what radical group might take over if economic reforms are not given a chance to work? Whatever the situation, we need to design and work smarter, not just harder. For 1990’s military aircraft procurement the Defense Department, turning away from the 1960’s TFX practice of awarding contracts on the basis of computer competitions and paper proposals, has returned to a competitive prototype, fly-before-you-buy, system reminiscent of the 1950’s. And even the Soviets used their 6-boat titanium hulled ALFA class high-speed attack subs as a developmental tool, as they did with the ill-fated single-sub MIKE class KOMSOMOLETS prototype – not to mention the reported BELUGA experimen-tal alternative powerplant prototype. In this light, should we use our computer-assisted design and manufacturing capabilities to build one prototype attack sub incorporating advances outlined above, test it, and hold the supercomputer-generated production software in storage until there is a real need to produce additional numbers of such an advanced sub?
But in the meantime, with no subs to produce, what happens to General Dynamics’ and Newport News’ construction teams and highly specialized subcontractor supply base? What happens if there’s an unexpected short war– but it takes several years to build a submarine?
Clearly, we’re at a watershed. Some vital policy questions need to be answered before we produce large numbers of new attack subs – SEAWOLFs, CENTURIONsf or an advanced prototype incorporating such aerocngineering concepts as those outlined above. The issues are, therefore, ones not just of hull design, advanced acroengineering cross-fertilization, and artificial intelligence systems management, but also ones of production base preservation and control of soaring budget deficits and a ballooning – almost crippling — national debt. For such fundamental, urgent, and conflicting issues of basic national and economic policy there are no easy answers — but we, and the new Clinton administration, must somehow find an acceptable compromise.
