Read Microsoft Word - ASNE Development of a SeaTrain Concept- Final_v1 1 oct 2009 _2_.docx text version

ASNE: High Performance Marine Vehicles Symposium 9-10 November 2009

Dr. R. Scher, Dr. T. C. Fu, J. C. Ryan, Dr. M. Irvine, W. Hu, Dr. C. Jiang, and P. Clark

The Development of a Surface Effect SeaTrain Concept


As part of an effort initiated by the U.S. Office of Naval Research under their High Speed Sea Lift (HSSL) to Austere Ports program directed by Dr. Patrick Purtell, Broad Agency Announcement (BAA) 05-007, 2005, an articulated Surface Effect Ship (SES) Sea Lift concept was investigated, including necessary associated science and technology advancements. During the three year three Phase effort, Alion Science and Technology designed and constructed various size models including the Phase III 20th-scale four car articulated SES remote controlled SeaTrain model. The relatively large model, 16.6 m, was chosen to reduce scale effects associated with the cushion dynamics, seals, skirts, and articulation details. Model tests were done in both an indoor towing tank at NSWC Carderock and outdoors on the lower Potomac River. The results of these tests show very good seakeeping and hinge load characteristics plus breakthrough powering performance.

inherent in the physics of vehicles deriving lift from hydrostatic or hydrodynamic forces, or is merely an artifact of insufficient engineering exploration and development. In spite of this, we should agree that weight is at the heart of the problem. The intervening variables are familiar to us as ship designers. They include the structural weight of the vehicle, imposed by the need to carry loads, especially in the severe conditions that may exist for ocean-going vehicles encountering waves, and the power-to-weight, efficiency, and specific fuel consumption of the propulsion machinery (and the lift machinery, t , a pw r "omo ltrdco o i "o e d fr fi pout n of e f i happens to be involved). The weight of the structure, machinery, and fuel are added to the weight of the cargo and then we lament the unaffordability of demanding high speed over long stage lengths at sea, yet the demand for high speed remains. When other constraints are added, for example, length and draft limitations for access to small or undeveloped ports, or for beaching, the technical challenge is especially severe, and one might even be tempted to say impracticable. To achieve a substantial increase in speed over a trans-oceanic stage length, but without incurring unacceptable increases in ship size, power, fuel consumption, and draft, it is clear that some major innovations would have to occur in design concept, design criteria, or operating methods, and life-cycle management, or perhaps in all of these areas. At the risk of over-simplification, the physical facts that govern the achievement of higher speeds for surface vessels put critical emphasis on a few aspects of the solution: (1) Operating in a favorable speed regime, that is, either below or above a critical


Several notional logistic architectures for future surface transport have envisioned a doubling of sea speed, from the 20 - 24 knot range typical of present sealift ships, to 35 - 40 knots or more, over trans-oceanic range. It is ntupin t t p t fbspate o srri h a l o "et r i " sg a o cc lift/drag ratio or transport factor versus speed for various vehicle types (Kennell 2001) reveals a domain of speeds in which economical performance is relatively difficult to achieve, implying that even the best vehicle designed for such a speed regime may be too fast to be an economical surface ship, and not fast enough to be efficient aircraft. It is a matter of some disagreement, however, whether that gap (if it exists) is somehow

ASNE: High Performance Marine Vehicles Symposium 9-10 November 2009

range of Froude number, and with suitable hull form and proportions to avoid excessive wave-making resistance. (2) Incorporating a suitable form and proportions, or other measures, to avoid excessive viscous drag (and air resistance, too, in cases where that constitutes a significant fraction of the total). (3) Reducing to acceptable levels the local and global structural loads that are imposed by wave encounters at higher speeds. The most suitable form and proportions for reducing wave drag and for reducing viscous drag are not necessarily compatible. Further, reducing the occurrence and peak magnitudes of loads in waves presents a technical challenge. The history of high-speed ship design suggests that there have been two very distinct but classical approaches to higher speed for ocean-going surface ships (without incurring completely infeasible or unaffordable machinery and fuel weight penalties): Slenderness: (interpreted as a high ratio of length to cube-root of displaced volume) to achieve low wave-making resistance. Dynamic lift: supplementing or (as a limiting case) supplanting buoyancy, essentially lifting the body out of the water, concurrently reducing wetted surface and frictional resistance. Both of these approaches may, but do not necessarily, involve the ability to pass through the critical range of Froude number (even at the expense of the greater thrust that may be required to do so), and then operating with lower wave resistance in a "ue ri l spr ic " -c ta speed regime. In principle, both slenderness and dynamic lift may be combined in a single vehicle. The important point is that weight is an antagonist to the achievement of either. In addition, operational or economic constraints on vessel length often limit the ability of a designer to exploit slender proportions. The ONR Broad Agency Announcement etl " n td Architectural Concepts and ie

Hydrodynamics Technologies for High Speed SaftA s rP r" a case in point. elto ut e ot is i e s The principal ship performance objectives of the BAA are given in TABLE 1. The length constraint is particularly onerous, in view of the speed and range objectives. The fact that the original length constraint was relaxed by the sponsor during the course of the resulting studies shows how serious a constraint of this kind can be. In addition to these constraints and objectives the BAA also explicitly m n oe a t gtd p cm n it et nd " re i l e etnh i a " sa e neighborhood of 12,000 tons. This displacement was not construed as a strict limit (as the length constraint was), but rather it was interpreted as a further indication that cnet nlb si sl i s e not ovn oa"i h " o t n w r i g p uo e what the sponsor intended.

TABLE 1 BAA Performance Objectives Maximum length (over-all) Maximum draft (port entry) Payload Sustained transit speed Performance environment 4,000 tons 43 kt Through sea state 4 (significant wave height 2.5 m) Range (unrefueled at transit speed) 5,000 nautical mi Original BAA: 137 m later increased to: 170 m 6.5 m

Three independent teams were awarded Phase I studies under this ONR BAA. Heading one of these teams, Alion Science and Technology (at the time the studies began, John J. McMullen Associates, JJMA) began with parametric investigations spanning a very wide range of high-speed vehicle concepts ranging from the conventional to the highly speculative (Scher, McKesson, and Ryan, 2006). These included: Catamarans of various hull forms, slender displacement, semi-planing, and foilassisted.

ASNE: High Performance Marine Vehicles Symposium 9-10 November 2009

Slender monohu ad s l n "tabilized l monohull" concepts (for example, unequal trimaran / pentamaran variants in which the center hull contributes by far the greatest portion of the displacement or lift). [Among these concepts, one variant w s s ae " t r D uh ri a a cld Mo e agt "n l h­ e which a large slender monohull structural separates into two sections, near amidships, so that each section can enter port (under auxiliary propulsion) within the length constraint.] Equal (or near-equal) hulled trimarans, primarily to allow larger displacements on the severely limited length and draft. [Among these concepts, one variant gradually evolved into the staggered "dp v l g "r a n T e et aat ee t tm r : h cn r i nh i a e hull is staggered for sea transit, so the over-all length of the trimaran exceeds the austere-port length constraint. For port entry, the center hull separates from the "a m r sco"to i hl ad ct a n et n ( wn u s n a a i w g l cross-structure) or retracts under the crossstructure.] " e ad h k" ocpae avsi H n n ci s cnetlr t e, c tn i n which several physically unconnected hulls are operated in close formations, with alignment and separations intended to reduce the total power requirement, primarily by wave cancellation effects. Surface Effect Ship (SES) concepts. These included conventionally proportioned SES variants, segmented cushion (differential cushion pressure) concepts, and a speculative highlength/beam articulated cushion-borne vehicle concept known at the time simply a "eTa . s Sa r n i" The principal conclusion to be drawn from Phase I is this: given the current state-of-theart of marine power systems, propulsors, and sut e,h ds n fr 4 ko "e i " t c r si ei so a 3 ntsr c r us p g ve speed over a range of 5,000 nautical miles, with a disposable payload of 4,000 tons, do not converge if strictly limited to the length and draft constraints of the BAA. Fuel weights on the order of twice the cargo payload, and installed power requirements of

250,000 hp to 400,000 hp, unaffordable to say the least, were not uncommon. H w vr"dp v l g " ocp apa o ee aat ee t cnet per , i nh s to offer the possibility of power and fuel weight savings of nearly a factor of two. Consequently, if engineering concepts and m cai so e i"dp v l g " a ehn m t pr taat ee t cn s m i nh be successfully developed, the magnitude of improvement is sufficient to achieve the performance objectives of the BAA. Two adaptive length concepts were considered most promising: the staggered trimaran with a retractable center hull, subsequently developed and model tested by a team headed by the University of Michigan (Maki et al, 2008); and the Surface Effect SeaTrain, further developed and tested by the Alion Science and Technology team. It should be noted that a number of dissimilar concepts have all come to be called SeaTrain, or Sea Train. The cm o i r i ti a o t " r n o m n n e e sn l fh Ta " gdn l e i concepts are modularity and adaptive length; that is the ability to transit at high speed in a hdoya i l "l drcni r i and yrdnm c l s ne of uao ay e " g tn then reconfigure into separate sections for other phases of an operation.


The Surface Effect SeaTrain was first proposed by Mr. Larry Keck, of Keck Technologies LLC. The concept is a Surface Effect Ship in which the modules are coupled, permitting flexure of the train, but essentially n "l k ( ai n r av sre With o s c"t ts o e t e ug) a h , li . suitable design of the matching ends of the modules, the air cushion extends continuously from the bow seal (located near the bow of the fr a ui o " a 1 a icm t b ow r n ,rC r , sta eo e d t " called) to the stern seal (located near the stern of the aft unit). Initially, the description of the coupling between adjacent cars was generalized to include various degrees of freedom. However, the present realization of the concept, as described below, uses couplers similar in principle to those of articulated tug barges, to permit flexure in relative pitch only, while remaining stiff in the horizontal plane.

ASNE: High Performance Marine Vehicles Symposium 9-10 November 2009

Preliminary evaluation of the speed-powering characteristics of a Surface Effect SeaTrain concept was undertaken at JJMA, in cooperation with Keck Technologies, in September of 2004. At that time, potential applications for both commercial and military purposes were considered, but at sizes considerably smaller than those which later emerged from the HSSL studies. In particular, preliminary calculations were performed for a notional landing-craft application, sized for transport in an LSD or LPD well deck, with t i i da"a " i daapoi a l h n v ulcr s e tprx ty e di s z m e 30 m coupled length x 14 m breadth, with a train consisting of three or four units. Powering calculations for 40 to 50 knot speeds were considered promising, especially when compared with an equivalent well-deck occupancy of LCACs. The SES configuration permits higher practical cushion pressure, and thus heavier cargo loading, than a fully-skirted air cushion vehicle (ACV), although the SES is, of course, not fully amphibious. Studies of a Surface Effect SeaTrain with over-all dimensions of 122 m x 14 m, and with a rather heavy displacement of 2,000 tons (equivalent to approximately 10 or 11 fully loaded LCACs) indicated total drag to weight ratios of about 0.04 could be obtained at speeds from 20 to 30 knot, and about 0.05 at 48 knot. These are very attractive numbers from a powering perspective, especially when combined with typical propulsive efficiencies that can be obtained from waterjet propulsion, versus airscrews. Notably, a slender SeaTrain has a much lower hump drag than conventionally-proportioned SES or ACV craft. Further, from a practical standpoint, in spite of the non-amphibious limits of the SeaTrain, shallow draft and articulation were considered advantageous for beaching: high ground loadings on the bow of Car 1, and longitudinal bending moments on the entire coupled train, are avoided due to the articulation. In parallel with studies of possible landingcraft applications, commercial variants (intended for coastal and short-sea container feeder and trailer-ship services) were also considered, with breadths up to 23.1 m and

train lengths up to 222 m (St Lawrence Seaway lock constraints), and displacements of 3,000 to 4,000 tons, with similarly encouraging results for speeds from 30 to 40 kt. It was suggested, too, that a slender SES, having exceptionally low wave-making drag, could be anticipated to produce very low wave wakes and pressure drawdown effects, important environmental and property-damage considerations for high-speed operation in confined waters.


In response to the ONR HSSL BAA, which appeared in mid-2005, the Surface Effect SeaTrain was initially proposed by the Alion (JJMA) team as one of the concepts to be investigated during Phase I. It was realized, of course, that the size required to meet the BAA objectives would be considerably larger than had been considered previously, and that the engineering challenges would be accordingly magnified. However, among the Phase I concepts outlined above, the Surface Effect SeaTrain concept was judged to have several unique characteristics of value for HSSL. The hydrodynamic advantages are quite clear on theoretical grounds. For example, extreme slenderness in a conventional displacement hull is unavoidably accompanied by an increase in wetted surface per unit displacement; consequently, the trade-off between reduced wave-making resistance and increased frictional resistance determines the point at which further increases in slenderness actually increase the total resistance for a given speed. For an SES, however, the hull wetted surface and the magnitude of the frictional drag are reduced when operating oncushion. Thus increasing slenderness of the Surface Effect Seatrain is unusually beneficial, in terms of total resistance. The proportions of the vessel are extremely slender by all measures: an L/B of approximately 10.4 for vessel over-all dimensions, 12.8 for the cushion, and approximately 90 for each sidehull. The volumetric coefficient (W / L3) g is 0.00037.

ASNE: High Performance Marine Vehicles Symposium 9-10 November 2009

The ability of the Surface Effect SeaTrain to deflect in the vertical plane permits the adoption of an unprecedentedly high L/B ratio for the cushion and for the over-all vehicle, without requiring a corresponding increase in freeboard, wet deck clearance, and over-all height. This fact contributes further to structural weight savings, windage and air resistance, and stability. The bow and stern seals are completely conventional, no larger in breadth or height than those of a much smaller conventional SES (essentially an SES the size o oe fh "a " The bow and stern seals f n o t cr ) e s. being proportionately narrow, and with the ability of the side hulls to contour larger waves, cushion-air losses are accordingly smaller than would otherwise be the case, and thus the required size and power consumption of lift fans, per ton of cargo, is reduced. Perhaps most importantly, as a result of its length, a Surface Effect SeaTrain of the scale required for HSSL can be designed to transit at a sub-critical Froude number, at considerably lower total resistance per ton than conventionally proportioned surface effect ships. For example, the HSSL Surface Effect SeaTrain shown in Figure 1 is 332 m LOA, 32 m beam, draws 1.8 m on cushion and 4.9 m off cushion, and weighs 13,800 tonne full-load departure. Thus, a speed of 43 knots represents a Froude number of only 0.39, not ee "u -sed for this length of vehicle. vn hl pe" l

Effect SeaTrain represented technically unexplored territory. This was especially so with respect to determining the maximum loads and articulation angles required for the couplers, the possible means of containing cushion pressure in way of the articulations, and at least the suspicion that some unexpected behavior of a high-speed articulated marine vehicle could exist. Perceived development risks in the hydromechanics area required some very fundamental investigations. Initially, for example, it was by no means universally agreed that an articulated surface effect ship could even be successfully propelled by thrust applied only on the aft car. On the basis of earlier (unrelated, but articulated) marine vehicle concepts, it was suspected that certain degrees of freedom in the hinge couplings, notably relative sway and yaw, of course, could produce maneuvering and coursekeeping difficulties, or even dynamic instability of the couplers themselves. For these reasons, Alion Science and Technology funded the construction of an Internal Research and Development (IRAD) proof-ofconcept model, constructed by Island Engineering, Piney Point, MD, during the summer of 2006. The model, shown in FIGURE 2, was 11 m long, 1.22 m beam, and with a mass of about 550 kg. It was equipped with four steerable outboard motors (4 hp each), two in the forward section (Car 1) and two in the aft section (Car 4), with the lower units extending through the wet deck and the propellers fully submerged below the air cushion. Electric power for two lift fans (400 W each, one forward and one aft) and for steering and remote control was provided by a 2 kW generator (the red object located on Car 3 visible in FIGURE 2). The model was equipped with hinge-pin couplers mounted on variable pneumatic springs. These permitted relative heave and relative roll between cars, with variable stiffness, in addition to freely hinged relative pitch. The air cushion was contained by flexible foam sections of the side hulls and wet deck (the white bands in FIGURE 2).

FIGURE 1 Alion HSSL SES SeaTrain concept

Investigations during HSSL Phase I indicated that such a vehicle could operate at values of total drag to weight approaching 0.02 at 43 knot, and could be propelled by gas turbines and waterjets totaling approximately 140,000 hp installed, plus diesel powered lift fans of about 36,000 hp total. Accordingly, at the conclusion of Phase I of the study, the Alion team recommended that the Surface Effect SeaTrain concept be pursued into Phase II, including model testing. In part, the motive for this recommendation was that the Surface

ASNE: High Performance Marine Vehicles Symposium 9-10 November 2009

FIGURE 2 Alion IRAD SeaTrain model

decisions for a second model, with funding from the ONR HSSL program, Phase II. This was a 1/36-scale model representing the HSSL SeaTrain design, built and tested in the towing tank and seakeeping basin at MARINTEK, Trondheim, Norway, in late 2006, FIGURE 3. The model was approximately 9.3 m long, 0.89 m beam, and with a mass of only about 290 kg. It was propelled by two stock-model waterjets, located in the side-hulls of Car 4, with electric power for propulsion and lift delivered via an umbilical from the carriage. The couplers were articulated in relative pitch only.

Observations showed that relative heave and roll freedom were not necessary, nor even desirable. A car (or one side of a car) which bcm " w o "i " i r pctt ea e l " rh h wt e etoh o g h s e adjacent car, due to relative heave or roll, often tended to remain misaligned or t"ut o hn " between low and high misalignment in waves. More predictable behavior was observed when the pneumatic springs were pre-charged to increase their stiffness, and thus articulation was essentially limited to relative pitch only. Because of the low shaft lines of the outboard motors, the application of thrust at one end or the other could produce a significant effect on the trim of Cars 1 and 4, and consequently the hinge angles and thus the over-a "hp" f l sae o l the train, viewed in profile. It became cs m r t sek fh t i a " hpd ut a o pa o t r n s M-sae" o y ea o " sae" tpe, ai w the bow rW- hpd asedt ts ith h , and stern of the train depressed, or elevated. The shape, in turn, produced visible and dramatic effects on spray formation, and on speed. Nevertheless, it proved completely possible to propel the model using either the forward or aft propellers alone, or using all simultaneously. The turning behavior using vectored thrust of the aft outboard motors was quite conventional. The selection of a single degree-of-freedom coupler concept (hinged for relative pitch only) and stern propulsion only, were based on experience with this uninstrumented proof-of-concept model.

FIGURE 3 Alion HSSL SeaTrain 1/36-scale MARINTEK model

Roll, pitch, accelerations, coupler forces and hinge angles, cushion pressures, and waterjet rpm and torque were measured (Tvete and Jullumstro 2007). Experimental results confirmed theoretical estimates of resistance and powering, as expected, and estimates of maximum hinge forces and angles (Jiang and Musatow 2007) based on numerical simulations were in good agreement with statistics based on the model measurements. Some design differences between these first two SeaTrain models are worth pointing out. The IRAD model side-hulls had fairly bluff "l i " o swtt bwsactoh p n g bw , i h o elutt an h e e baseline, similar to many SES designs of conventional proportions. The 1/36-scale HSSL SeaTrain model, by contrast, had very fine entry waterlines and the bow seal cut at the on-cushion draft. This model generated little or no spray, and ran stable, straight and level (zero hinge angle at the first connector), and was never fitted with spray rails. The IRAD model initially showed a slight pitch instability and bow-down trim of Car 1,


Observations made with the Alion IRAD proof-of-concept model contributed to design

ASNE: High Performance Marine Vehicles Symposium 9-10 November 2009

associated with non-scalable spray formation, but the addition of spray rails eliminated this behavior. However, Car 1 of the 1/36-scale HSSL model was apparently more lightly damped in pitch than the earlier IRAD model, and tended to develop higher amplitude relative motions in head and bow seas, producing an unusuall k g o o t t ea e nw a " ae o i m t n h bcm ko n s w von i a co p g T e hs r aosi bten hm i . h paee t nh e e n" li p w heave and pitch motions of Car 1 is, of course, strongly influenced by the coupling forces at the hinge. For this reason alone, motions should be expected to differ from those of a rigid SES the size of Car 1 running alone. Nonetheless, this observation motivated an investigation of active ride-control options for Car 1. Speed loss in waves was also substantial. Much of this was attributed to waterjet air ingestion ­ stock waterjet the inlets were located close to the outboard chines of the side-hulls. However, the limited length of individual runs produces a built-in bias on speed loss and recovery. A i u d et r a doh "hp" fh n s e i cy e t tt sae o t s r l le e e t i asedsh nm e o "a . r n tpe it u br f cr" a e s However, as it has turned out, from a purely powering standpoint this is not a sensitive hydrodynamic decision once the over-all L/B of the train has been established. The number and length of cars is, of course, a key structural issue, involving both the longitudinal bending moments acting on an individual car, and even more importantly for train that is essentially rigid in the horizontal plane, transverse bending moments. The number of cars is also an operational issue for adaptive length, to comply with length constraints for port entry or other operations. For example, the length of each independently mobile sub-section of the train was selected for HSSL to comply with the 170 m length constraint, each sub-section of the train, in fact, consisted of two cars, because this saved structural weight by reducing the longitudinal bending moment to a point where required scantlings matched those required for local loads and transverse bending. However, in retrospect, there is nothing sacred about an even or odd number of cars. In fact, if the

HSSL BAA had retained it original austere port constraint of 137 m, the team might have been tempted to use only three cars, each car independently mobile, off cushion, and under the length limit for port operations. Such a configuration would have involved one less coupler set, but one additional auxiliary propulsors set and an additional auxiliary control station, to permit port operations by three independent sub-sections rather than two.


In view of the typical seakeeping-basin environment, high over-all cost, limited model size, and limited duration of individual runs, it was judged that the most important remaining technical questions for SeaTrain could be effectively resolved by outdoor testing with a larger model. The main technical issues included: (1) coupler loads on all headings, including responses to slamming; (2) complex dynamic responses in short-crested seas; (3) ride control; (4) speed loss in waves, and methods of mitigation. On the basis of seakeeping model work on a U.S. Navy program, it was judged that wave environments in the neighborhood of Piney Point or Patuxent River, Maryland would be appropriate for a model scale of about 1/20. For HSSL SeaTrain this scale results in a model over 16 m long, weighing about 1.73 tonne, and requiring approximately 10 hp, total. A model endurance of at least one or two hours without refueling or battery recharging was considered highly desirable.

Indoor Testing Constraints

The development of a large and powerful outdoor model is very similar to the design of an actual small boat. Without further test objectives or constraints, a natural choice for a craft of this size and power would have been gasoline engines for propulsion and either gasoline or electric-motor driven lift fans. The latter, of course, would require gasolinepowered generators for electric power on board. But in fact electric power would be required in any case: for instrumentation, remote control, steering and reverse actuators,

ASNE: High Performance Marine Vehicles Symposium 9-10 November 2009 10

ride control (e.g., active variable-geometry lift geometry fan and vent box), boost fans for bag-type type stern and intermediate (ride control) seals, and auxiliary systems such as lights and bilge pumps. However, it was also desired to perform instrument calibrations, ride control ss m" n-us ad e f ao o yt t e p, n vri t n f e u " ic i resistance and powering requirements and coupler loads (head seas only) in the towing ) tank at NSWCCD. Gasoline power is not permitted indoors. Accordingly, to accommodate the needs of both indoor and eeds outdoor testing, considerable effort was devoted to systems using electric power for all onboard. For outdoor tests, power would be supplied by on-board ac generators, gasoline c gasolinepowered. Indoors the generators would be secured and ac power would be drawn from the towing carriage.

FIGURE 4 HSSL SeaTrain nomenclature

FIGURE 5 Typical car-box structural

Model Design and Construction STRUCTURE

T e oecnis forcr"C r ad h m dlos to fu "a . a1 n s s Car 4 are the bow and the stern of the train, respectively, while Cars 2 and 3, completely identical in construction, for the midbody. E c cm le cros to a cr o ah o p t acnis f "abx ed s sco,w i cn i t a angeable et n h h otn h r i " c as e r volume for equipment in each car. Attached to the car box, by bolting to the end bulkhead(s), are the matching sections of the aj etcnet m dl s"E c d cn"onc r ou ( . ah a o e) connector module in turn consists of a " i w lsco" n a t ut o sco. wn a et n ad " rsbx et n g l i h i " Further details and arrangements of the connector modules are discussed below. The nomenclature of a complete car assembly is illustrated in FIGURE 4. FIGURE 5 shows a typical car box section.

The connector modules are unique in that they must serve both as structures and as instrumented pieces, and must also contain cushion pressure while being able to rotate freely. Each of the three connector modules consists of a thrust box and a wingwall ngwall section, as shown in FIGURE 6. The instrumented pins are removable, but it was determined that once calibrated they should not be routinely dismounted. Instead, the bolts attaching the midship connector module to the car box of Car 3 are removed for transport and ar reinstalled before launching.

FIGURE 6 Connector module wingwall and thrust box sections

The connectors accommodate a hinge angle of about ±28 deg before hard contact occurs between structures. A small gap is left between the matching surfaces of the wingwall and thrust box sections. The concentric large cylindrical surfaces in way of the side hulls have a nominal gap of 10 mm, with a similar gap across the width of the thrust box. The gaps between the port and ween starboard sides of the thrust box and the matching inboard surfaces of the wingwalls are somewhat larger, 15 mm (this lateral clearance was needed to facilitate a demonstration of separation and reconnection afloat). The gaps are sealed (in the full-scale

ASNE: High Performance Marine Vehicles Symposium 9-10 November 2009

HSSL design concept) by inflatable or Pshaped extruded materials so as to provide sliding compliant seals. For the model, however, the gaps were initially filled with foam inserts, but this material proved to have excessive compression and friction, and the model did not articulate freely. The foam was removed and replaced by lowpile carpeting, as shown in FIGURE 6. The carpet-seals, ungreased, ultimately allowed excessive air leakage and prevented the model from maintaining its full design cushion pressure. In retrospect, inflatable seals, faced with low friction plastic, may have provided a better seal without increasing friction. The thrust box and wingwall sections of each connector module mate to each other with two instrumented pins (couplers), port and starboard, providing the hinge function of the connector and allowing each car to articulate in relative pitch with the adjoining car. Each pin is a 1-inch diameter stainless steel bar, with a 3/8 inch central bore with strain gauges cemented to the interior surface. The design working load on each pin is 5.5 kN (about 1200 lb). As shown in FIGURE 7, each pin is bolted to a foundation fixed in the wingwall section. The body of the pin rotates in a bearing mounted on the lateral surface of the thrust box section. The inboard end of the pin bears on a button load cell, which is in turn supported on a pre-loading mechanism that can be manually adjusted to eliminate athwartship slack in the coupler. Thus, longitudinal and vertical forces in each pin are measured by the strain gauges inside the pin, while lateral forces between sections (axial forces in the pins) are measured by the button load cells.

The design of the full-scale SeaTrain propulsors was not an issue for this model. Therefore, primarily to control costs, two inexpensive commercial units (Berkeley WJ5, no longer in production) were used, FIGURE 8. These jets are rated to absorb a maximum 15 kW (20 hp) each, at 6000 rpm. In this model they actually ran up to about 4200 rpm in seakeeping tests, absorbing about 5 kW (7 hp) each. On the highest-speed calm water runs, 4650 rpm was obtained, corresponding to 7 kW (9 hp).

FIGURE 8 Waterjet installed in the model

FIGURE 7 Instrumented pin


The size and required power of this model are byn t cpb i o "ob so" i r t eod h aait f hby hp a c f e ly ra electric motors, and obviously not large enough for ac motor controls. Over-all weight constraints and the desire for outdoor test duration u d gi t 2 9 V C"o -cr rl aa s7 /6 D gl a " e n f t propulsion with batteries. Accordingly, a system was designed and built using three ac generators (two 6.50 kW + one 3 kW), rectifiers and solid state motor controllers for dc propulsion, and two motors per waterjet. These motors are mounted as an extremely compact assembly, narrow enough to fit in the side-hull dimensions, driving a single belt down to the waterjet shaft-line, as shown in FIGURE 9. The system was given an initial shake-down on the model, outdoors. Although all components of the system functioned, the motors, even with watercooling, were not capable of developing their claimed rating for more than a few seconds. In spite of several efforts to improve or replace these motors, the electric propulsion concept ultimately had to be abandoned in favor of a gasoline-engine driving both waterjets, prior to outdoor testing.

ASNE: High Performance Marine Vehicles Symposium 9-10 November 2009 10

FIGURE 9 Electric propulsion assembly design and actual unit

As it became obvious that the electric motors would not be able to develop sufficient power, and as the weather-window for outdoor testing window was rapidly going by, a single Honda V-twin industrial gasoline engine, rated 18-hp at 3600 hp rpm, was installed in Car 4; FIGURE 10 ; 10.

FIGURE 11 Variable geometry fan assembly

FIGURE 10 V-twin gasoline engine and belt twin transmission layout

The cushion under Car 1 is separated from the n remainder of the air cushion by a five-lobe bag lobe seal, shown in FIGURE 12. This seal is . rigged with internal straps to hold the lobes in their triangular stacked arrangement and maintain this shape against small overpressures from either forward or aft. The sures bottom lobe rides just above the cushion water surface at the design speed. The location of the intermediate seal is of key importance in the ride control system. For the tests, the seal was actually positioned on Car 2, just aft of ar the connector module.


Two industrial electric fans, 1.5 hp each, are installed on Car 1 and Car 3. For ride control, the forward lift fan was fitted with a variable geometry (VG) system designed and bui by built Island Engineering, illustrated in FIGURE 11. A linear servo actuator, controlled initially by a vertical accelerometer on Car 1, adjusts the axial location of an internal cone, inside the fan cage, in effect varying the wheel width and air flow of the constant rpm fan. The vent box, a rectangular butterfly valve chainactuated by a separate servomotor, is also located on Car 1, discharging to the sides of the superstructure, port and starboard.

FIGURE 12 Intermediate ride control seal.

Bow and stern seals are completely conventional: the bow seal with 10 open openbacked fingers, each independently mounted t t w tek s g p sc x ui "ao h edc ui a l t et s n si e n ai r o l t c" ei ,o es o r l e ett r k dv efrae fe a m n h a c pc ;e stern seal a two-lobed bag seal with water lobed drain openings in the lower lobe. The stern

ASNE: High Performance Marine Vehicles Symposium 9-10 November 2009

and intermediate seals are inflated 1 to 5 percent above cushion pressure, using small " uf "as dp d rmds-top computer m fn f aat f i n e o ek cooling fans. These fans proved susceptible to corrosion in the marine environment and a number of them failed during the course of testing. All seals are constructed of Fairprene, a light rubberized fabric.


An unusually large instrumentation suite was designed and installed by NSWCCD. Several on-board video camera-recorders (and lights) are positioned to view the water surface and seal behavior inside the cushion. Data is sampled at 10 Hz and recorded on a computer on board. The model remote control system was designed and built by Island Engineering, and operated completely independently of the experimental instrumentation. The remote control system provides both manual-mode and autopilot steering (commanded heading only); commanded throttle setting(s) for the propulsion motors or engine; and reverse bucket position (full up or full down only) on port and starboard jets. With the final gasoline-powered propulsion, a remote control for the magnetic clutch is also incorporated. A kill-switch on the propulsion system, operated automatically on loss of network connection, or on manual command from the operator Graphical User Interface (GUI), is included. Ride control system settings (VG fan and vent box controller zero settings and gains) can also be adjusted from the remote GUI. The SeaTrain model is extensively i t m n do u y vl tec cr n r et tfl ea a ah a s su e l ue ' motions and accelerations; cushion pressure; connector pin loads; waterjet rpm and torque; and ride control. Data from the instrumentation is transmitted wirelessly to a laptop data acquisition computer. The waterjet rpm, steering bucket angle, and ride control system are controlled by a separate onboard computer that can be commanded wirelessly from a separate model control computer.

Car 1 is instrumented with a Crossbow NAV440 Inertial Navigation system that measures roll, pitch, heading, roll rate, pitch rate, yaw rate, surge acceleration, sway acceleration, heave acceleration, longitude, latitude, altitude, speed-over-ground, and track angle; an acoustic wave height sensor mounted on the bow to measure relative wave height; a position sensor to measure the VG lift fan cone position; an inductive pickup to record the VG lift fan RPM; a linear potentiometer to measure the VG lift fan vent valve position; and a differential pressure gage to measure the cushion pressure under Car 1. Car 2 and Car 3 are instrumented with a 3DMGX1 Gyro Enhanced Orientation Sensor that measures the surge, sway and heave accelerations, and roll, pitch, and yaw rates. Care 2 also has a differential pressure gage to measure the pressure in the intermediate bag seal; and a differential pressure gage to measure the cushion pressure under Car 2. While Car 3 has an inductive pickup to record the non-VG liftfan RPM; and a differential pressure gage to measure the cushion pressure under Car 3. Car 4 is instrumented with two 10.16 cm variable reluctance transducer block gages to measure the resistance and sway forces; a 3DM-GX1 Gyro Enhanced Orientation Sensor; an acoustic wave height sensor mounted on the stern to measure relative wave height; port and starboard torque meters that measure the RPM and torque for each waterjet respectively, a potentiometer to measure the waterjet (rudder) angle; a differential pressure gage to measure the pressure in the stern bag seal; and a differential pressure gage to measure the cushion pressure under Car 4. Each thrust box/wingwall connector sections is instrumented with port and starboard connector pins that measure the surge, sway, and heave forces on each respective pin; a potentiometer to measure the relative pitch of the thrust box to the wingwall (which is used as the relative angle between two connected cars); and a portside mounted acoustic wave sensor to measure the relative wave height.

ASNE: High Performance Marine Vehicles Symposium 9-10 November 2009

The carriage is instrumented with two acoustic wave height sensors attached to a beam on the west end of the carriage to measure the incident wave-field, and an encoder to measure the carriage speed.


The Surface Effect SeaTrain presents an unusual problem in model weight control. The initial target condition for the model was to be at zero hinge angles and zero vertical loads in the pins, at the design draft (92 mm) on cushion at the design cushion pressure (683 Pa), and with full load weight of the total train. Even with each car at the proper weight, the trim of the car affects the adjoining car, via forces at the hinge. So, to obtain the desired draft and trim of each car, the weight, longitudinal center of gravity, and lift (the sum of buoyant and cushion pressure forces) must be balanced. The target weights and longitudinal centers for each car, satisfying these conditions, are shown in TABLE 2. It is convenient to use the axis of a connector hinge a a oi n o ec cr l g ui l s n r ifrah a so i d a g ' nt n coordinate. Arbitrarily, the origin for the entire model and the origin for Car 1 and Car 2 coincide at the center of the hinge of the first (forward) connector. The origin for Car 3 is the hinge of the second connector, and for Car 4 the hinge of the third connector. Coordinates are all taken positive aft of the first hinge. Thus, the LCG for Car 1 is negative (forward of the hinge).

TABLE 2 Design car lifts and longitudinal centers.

Model Section Car 1 Car 2 Car 3 Car 4 Draft (mm) 92 Total Lift (kgf) 424.5 435.9 683 435.9 426.7 Total SeaTrain = 1723.1 Cushion Pressure (Pa) LCTL (m) -1.926 2.081 2.081 1.987

is rigidly attached only to a wingwall section, forward. The weight estimate for the model t ce ec cr w i tn l g ui l r kd ah a s e h ad o i d a a ' g nt n center as the design developed, and compared each car with the target lift and center of lift. The difference in weight represented the amount of available ballast on each car. By tracking the design in this way it could be shown whether there was enough longitudinal moment available (using ballast) to place the car LCG at the required longitudinal center of lift. The model construction weight estimate was updated monthly, or when major milestones occurred ­ such as part assembly completion, equipment arrival or a design modification.


Significant results of the Phase III model tests, towing tank and outdoors, can be summarized as follows: With regard to hydrodynamic efficiency, as expected, SES resistance predictions using classic methods (Doctors and Sharma, 1972) are reliable for the high L/B configurations envisioned for SeaTrain. Changes in resistance due to deflection angles at the hinges (that is, due t t "hp" fh t iipoi)r o h sae o t r nn rfea e ea l e extremely slight, and drag due to local flow disturbances at discontinuities is small, provided that matching connector surfaces are reasonably closely joined. Stern seal drag is less important than for cnet nl poot nd E ' ovn oay rproe S Ss i l i ; accordingly there is less sensitivity to stern seal pressure variations (Irvine et al, 2009). Design details to prevent excessive cushion air losses at connectors are important. This has significant implications for the practical design of a full-scale connector. Typical shipbuilding tolerances and the matching surfaces of connectors will almost certainly demand an inflatable seal of some kind to allow full contact without excessive normal force and consequent wear (Scher et al, 2009).

For the purposes of determining mass properties of an individual car, a car box and the appropriate sections of the connector module(s) must be considered as a single rigid body. That is, Car 1 is rigidly attached to only a thrust box section, aft; Cars 2 and 3 are each rigidly attached to a wingwall section (forward) and a thrust box section (aft); Car 4

ASNE: High Performance Marine Vehicles Symposium 9-10 November 2009

Maneuvering, course-keeping, and windage effects on handling of the SeaTrain closely resemble those of a conventional vessel with long and slender proportions. With regard to coupler design, hinge pin forces were in reasonable agreement with numerical predictions (Jiang et al, 2008), and do not represent unachievable levels for practical design. As anticipated, midship coupler forces are driven primarily by lateral bending moments, resulting in the largest longitudinal coupler pin forces in oblique seas, particularly oblique bow seas at high speed (Ryan et al, 2009). The largest coupler forces between Cars 1 and 2 are caused by wave impacts in head and bow seas. Maximum vertical components in the forward coupler are of similar magnitude to the horizontal forces experienced at the amidships coupler. Roll appears to be effectively damped. It is possible that synchronous rolling is less likely to develop because the roll gyradius of the train is not constant, but is affected by variation of hinge angles. In following and quartering seas, in higher waves, at speeds where wave encounter frequency became low, persistent large connector (hinge) angle at the Car 1 to Car 2 connector was occasionally experienced. This condition resulted in a few events of large bow down trim of Car 1, elevation (so-cld j k g o t C r ­ 2 ae " ci " fh a 1 Car l a n e hinge, with deep immersion at the bow of Car 1. The resulting loads on the connectors were not exceptionally large in these events, and Car 1 recovered after a loss of speed. Extreme hinge angles at the forward hinge approached, but did not reach, hard contact at the keel (in the model, a limit of approximately 24 deg relative pitch). Additional freeboard and reserve buoyancy at the forward end of Car 1 would be an advantage.

In other respects, no remarkable responses of the SeaTrain could be attributed to articulation. Separation and reconnection of the couplers in protected waters, with a modest relative motion environment appears to be completely feasible. The concentric cylindrical surfaces of the connectors eliminate vertical relative motions once contact is achieved, and at this point the coupler rams (of a full-size design) would be in the correct alignment for engagement.


Some of the chief technical findings of the three-year effort reviewed in this paper include: The SES SeaTrain hull form achieves lift to drag ratios that are considerably above typical values. Required powering at 43 knots in calm water is approximately half of that for a conventional ship of similar payload capability. An associated feature of this hull form is its very low structural weight compared to payload ­ providing significantly higher payload to weight ratios for high speed ships. . The multi-car, hinged model shows no unusual behavior in a seaway, even up to Sea State 8. Connector loads are within the state of the art for articulated tug-barge technology. Speed loss in waves is only around 25% in Sea State 5. A good result for a hull form wt a e " a sed i vr f t pe-power curve. h y l" The developed predictive computer tools correlated well with most motions and loads test results. These tools allow for reasonable margins to be used for actual ship designs. The work done to date provides evidence that the SES SeaTrain concept is both feasible and practical.

ASNE: High Performance Marine Vehicles Symposium 9-10 November 2009

While the fundamentals of Surface Effect Ship systems and the SeaTrain are now generally considered within the state-of-the-art, there are many areas of both technology and design that needs further exploration before a full scale SeaTrain could be undertaken, including: Inter-car seal material and configuration development to reduce air losses. Ride control system effectiveness including design, tuning, and intermediate seal location. The need for additional buoyancy forward t r ueh t dny o " o n" bw o e c t e ec frr t g (o d e n oi digging into the back of a wave when going with the seas). Large-scale experimentation with the connectors using actual hardware. Additional structural experimentation to confirm bending moment loads on individual cars. Bow slamming loads are another area worthy of additional structural testing. Reduction of waterjet air ingestion in larger sea states. Design features such as: inter-car fuel system; personnel access; loading and unloading systems; reconnection systems and concept of operations. Finally, the benefits and challenges of model testing have confirmed the utility of doing both indoor and outdoor testing. At the larger scale needed for realistic configuration of the model components, such as connector assemblies, model tank testing provides excellent powering and controlled head seas conditions. In comparison, outdoor testing enables higher sea states and more realistic uncontrolled seaway conditions at the expense of increased measurement tolerances. This dual testing approach, while requiring a constrained model design, does provide a very complete picture of the behavior of a radically new concept.

1. K ne, .O t N t e fh enl C " nh a ro t l e u e Transport Factor Component TFship, " Marine Technology, 2001. 2. Scher, R., McKesson, C., and Ryan, J.C., " ONR High Speed Sea Lift Phase I Final R pr" l n c ne n E g er g eot Ai Si c ad ni e n , o e n i Report, Contract N00014-05-C-0347, Feb 2006 3. Maki, K. J., L.J. Doctors, R. Scher, W.M. Wilson, S.H. Rhee, A.W. Troesch, R.F. B c,C net l ei ad ek" ocp aD s n n u g Hydrodynamic Analysis of a High-Speed Sealift Trimaran, SNAME Transactions, 2008. 4. Tvete, M. and Jullumstro, E., "HSSL SES Concept Testing," MARINTEK Report MT53 F07-028 530301.00.01, 2007. 5. J n, . Mua w M. O RHg i gC, a & st , , N i o " h Speed Sealift Phase II Hydrodynamic Performance Assessment: Simulations vs P yi l dl et eu s Ai hs aMoeT sR sl, l n c t" o Science and Engineering Report, 2007. 6. D c r LJ ad hr aSD,T e ot s .,n S a , .. h o, . m " Wave Resistance of an Air-Cushion Vehicle in Steady and Accelerated Mo o,Journal of Ship Research, Vol tn i " 16, pp 248­ 260, 1972. 7. Jiang, C., McKesson, C., Musatow, M. and Scher, R., Time-domain Nonlinear Surface-Effect Ship Motion Simulation, Twenty-Seventh Symposium on Naval Hydrodynamics, Seoul, South Korea, October 2008. 8. Irvine, M., Fu, T.C., Rice, J.R., Clark, P., J n, . ce R, R a,.. T w i gC, hr . ynJ , o a S , & C " Tank Test Results of a Surface Effect SeaTrain C net F S 20. t n, ocp , A T 09Ah s " e Greece, 2009. 9. Scher, R, Clark, P., Fu, T.C., Rice, J.R., & R a,. T e ee p et f S r c ynJ " h D vl m n o a uf e , o a Effect SeaTrain Concept Model for H doya iT sn"F S 20. yrdnm c et g, A T 09 i Athens, Greece, 2009. 10. Ryan, J.C., Irvine, M., Rice, J.R., Scher, R., Clark, P., Hu, W., Quick, M., Jiang, C. &Mua w M.O t o T sR sl o st , , u or et eu s f o " d t


ASNE: High Performance Marine Vehicles Symposium 9-10 November 2009

a uf e f cSa r n ocp , S r c Ef t eTa C net a e i " FAST2009. Athens, Greece, 2009.


The authors wish to acknowledge Dr. Patrick Purtell, of the Office of Naval Research, for his consistent support through the entire HSSL program; Mr. Larry Keck, the originator of the SeaTrain concept, without whom none of this would have happened; and all those whose hard work went into seeing the model construction and testing to fruition, including James R. Rice, Don Walker, Conner Bruns, Martin Sheehan and Island Engineering.

from Florida Tech in Ocean Engineering. He also has a B.S. degree in Naval Architecture and Marine Engineering from Shanghai Jiaotong University. Dr. Changben Jiang is a Senior Principal Naval Architect at Alion Science and Technology. He has a Ph.D. in Naval Architecture and M.S. in Mechanical Engineering from the University of Michigan. He also has a B.S. in Naval Architecture and M.S. in Ocean Engineering from Shanghai Jiao Tong University. Ms. Pamela Clark is a Program Manager in the Naval Architecture Division of Alion Science & Technology. She earned her B ce r dge f mB s n n esy ahl ' erer ot U i r t os o o v i.

Dr. Robert Scher is the principal author and he is a Sr. Principal Naval Architect at Alion Science and Technology. He earned his B.A. in Physics from Johns Hopkins University and holds a M.S.E and Ph.D. in Naval Architecture and Marine Engineering from the University of Michigan. Dr. Thomas C. Fu is the Head of the Resistance and Propulsion Division of the Naval Surface Warfare Center Carderock Dv i . e b i d iMat 'dge i is nH ot n h io a e s s r eren es Physical Oceanography from the Scripps Institution of Oceanography-University of California, San Diego and his Ph.D. in Mechanical Engineering from Johns Hopkins University. J. C. (Kit) Ryan is the Chief Naval Architect at Alion Science and Technology. He has a B.S. in Naval Architecture and Marine Engineering from Webb Institute and a M.S. in Naval Architecture from MIT. Dr. Marty Irvine works in the Resistance and Propulsion Division of the Naval Surface Warfare Center Carderock Division. He has a B.S. and M.S. in Ocean Engineering from Virginia Polytechnic Institute and State University, and a Ph.D. in Mechanical Engineering from The University of Iowa. Mr. Weimin Hu is a Principal Naval Architect at Alion Science and Technology. He holds a M.S. degree from Virginia Tech in Aerospace Engineering and a M.S. degree


Microsoft Word - ASNE Development of a SeaTrain Concept- Final_v1 1 oct 2009 _2_.docx

15 pages

Find more like this

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate


You might also be interested in

AnRp93 Braiding Machine Design to Increase Production Rate
demo brochure 2008 for PDF_website:Layout 1.qxd
AP Physics Review Fall Semester
Microsoft Word - FPAS Handbook Part B_Syllabuses 10-12- A4 Paper2.doc