NAME 4177 - The Practical Design of Advanced Marine Vehicles - Chapter 18
SWBS 119 - Design of Air Cushion Skirts
This is another section where an entire textbook in its own right is needed. In this section I shall be able to do little more than acquaint the reader with the issues involved in the design of a skirt system, and educate the student on the basics of inflatable structures.
An ACV skirt system is a complex engineered structure. Figure 204 illustrates the components of one such system, and I reproduce it here merely to underscore the complexity of the system.
Figure 204 - An ACV skirt system
Very early Air Cushion Vehicle concept development began with Sir Christopher Cockrell’s first ACV in the late 1950’s which had a lift fan, an air propeller and no flexible seals to contain the air cushion which supports the weight of the vessel. Thus there was a very low daylight gap between the ACV’s hard structure and the land or water below it. The evolution of the flexible skirt concept soon followed which enabled the acv to keep its very low daylight gap to minimize the power required for the cushion and to permit the vessel’s hard structure to be above nominal perturbations in the water or land surface. It was later in the 1960s that the bag and finger skirt appeared for the ACV. The bag was needed to provide a more reactive surface on the seals, enabling it to better react to waves and other surfaces while the acv is pitching or rolling. All operational ACV’s now use bag and finger seal or skirt systems.
The air cushion vehicle must use air propulsion since it is amphibious and can not have water propulsion to take advantage of the factor of 800 in fluid density between and water. (There was an ACV ferry which used water propulsion with a skeg and water prop, but it was limited to in-water operations only.)
In the early 1960’s, Allen Ford patented his Captured Air Bubble Vehicle while working at the Naval Air Development Center in Warminster, PA. The CAB vehicle had movable seals at the bow and stern and rigid sidewalls on each side to better contain the air cushion and to support water propulsion. A 10-ton, 50 ft. long test craft was designed and built in the shipyard in Philadelphia. The XR-1 testcraft had seals that were comprised of two athwart-ship panels that were held down by the pressure in the CAB’s cushion and their lower movement was limited by cables or down stops.
The XR-1 and its 50% increased beam modification, known as the XR-1A,became the very early workhorse for testing out new CAB (later to be called a Surface Effect Ship) seal systems, dynamic controls for its lift system to improve motions and new waterjet systems. Most of the seal development on the XR-1C, -1D and -1E were planning type seals but with much improved flexibility and dynamics relative to the rigid planning seals. This type of seal was eventually designed, built and installed on the SES-100A1 testcraft and was the original seal design for the Rohr 3KSES. But local structural loads at the fasteners of the down stop cables and the fiberglass planning sections became difficult to solve in the time-driven program, so the final design changed to a bag and finger bow seal, similar to the seal on the SES-100B testcraft and other lesser-size SES vessels being built at that time. Again, the fingers were to help in containing the cushion air while operating in seas and the bag provided pitch-up forces and moments in addition to the pitch moments provided by the increased cushion length due to the slope of the fingers, for the craft in higher seas.
Purpose and Types of Skirts
As with so much of the AMV design, we begin with the teleology of a skirt: What is the purpose of the skirt on a powered-lift craft?
First, let it be well understood that skirts are not necessary. Sir Christopher Cokerel’s first hovercraft did not have a skirt. But the skirt was quickly invented and retrofit to the SR.N-1. During the early development of the SES, we learned about the physical phenomena that impact the properties that we want in bow and stern seals. The seals should:
- be adequately flexible to respond to various shaped waves to minimize airflow leakage, in various headings to the sea;
- be able to conform to the slope of the cushion pressure generated wave which varies with Froude Number. This has a strong effect on low speed resistance.;
- at lower sea conditions, an SES pitches very little but in higher seas, it gets its pitch restoring moments from the bow seal and the bow seal must be able to restore its shape to its original condition so that the lift from the cushion is restored rapidly;
- to minimize the frictional resistance of the sidewalls of an SES, the seals optimally are configured such that they function in their inflated mode with its lowest point is at or very near the keel of the sidewalls at the bow and stern. The cushion generates a wave pattern inside the cushion and the slope of this pressure generated wave varies with Froude Number, attaining a maximum slope at the vessels hump speed at a nominal Froude Number of 0.6; the magnitude of the slope varies with cushion length-to-beam ratio and cushion pressure.
- the cushion generated wave which approaches the stern seal is very 3-dimensional and its shape varies with Froude Number, requiring a degree of athwart ships flexibility.
- the seals must be able to eliminate any suction forces. The stern seal is much more prone to this issue than the bow seal;
- as the seals move vertically due to wave action, they must be able to survive the structural snatch-loads which are felt as the seal rapidly moves down after the wave passes;
- the lift system mechanism which strongly assists in driving the seals back to their lower position after they have been moved by the waves, must force the seals down very rapidly;
- since the length of the seals affect the amount of sidewall wetted area and frictional resistance, the seals must be durable relative to material wear;
- the seals concept and design must be able to be manufactured and repaired, preferable at sea.
Simplifying this, the ideal skirt system will:
- Retain the air bubble – reduce air leakage
- Have no drag
- Conform to waves without exciting ship motion - weightless / massless
- Assist with pitch stability
To accomplish the first of these - Retain the air bubble – the skirt must:
- Resist cushion pressure
- Retain desired geometry
- Be impermeable
As if that is not enough, consider the goal of being dragless. To accomplish this we want a perfect geometry of water contact, so that there is no wetted surface of skirt. In a static hovering condition this might be possible, but what about in waves? In waves we want the skirt to deflect out of the way instantaneously. This requires the skirt to be inertialess, or massless.
Finally, we want the skirt to assist in providing pitch stability for the craft. This means that the bow skirt will have a forward slope to it, so that when the bow pitches down there is some forward shift in the center of pressure, resulting in a pitch restoring moment.
To address these multiple goals, many types of skirt have been invented and tried. My list includes No skirt, Air Curtains, Water Curtain, Pericells, Fingers, Bag and fingers, Stayed bags, and Transversely stiffened membranes. But despite this broad range, these can be collapsed into three major types, which I shall address in turn
- ‘Virtual’ Skirts
- Rigid Skirts
- Inflatable Skirts
I class both “Peripheral Jets” and “Water Curtains” as ‘Virtual skirts’ because there is no physical structure retaining the cushion, instead it is retained by an inertial barrier formed by a mass of fluid – either air or water.
A peripheral jet system consists of a thin slot around the perimeter of the craft, and a high pressure jet of air blowing through this slot toward the ground. The momentum of the air jet is sufficient to retain a positive pressure inside the perimeter, in the cushion area of the craft.
The governing relationships, that give us the required flow and pressure from the jet, are given by Yun & Bliault (Reference 16) in the page that I reproduce as Figure 209. Note that the peripheral jet also supplies the air to the cushion – there is no separate lift fan system in addition to the jet fans.
Figure 205 - Yun & Bliault presentation of the governing relations for a peripheral jet
A homework assignment will be given in which the student will use these relations to find the lift power for a small number of hypothetical hovercraft. As will be seen, the problem with the peripheral jet method is that it requires a lot of power: The air jet must be given enough momentum to retain the cushion, which requires a substantial jet pressure and flow rate.
The Water Curtain concept is similar to that of the peripheral jet. But whereas the peripheral jet combines both cushion retention and cushion creation into a single air flow stream, the water curtain does require a separate air cushion fan system. It then uses the water curtain only to retain the cushion.
The idea of the water curtain is to use a mass of falling water to produce the pressure barrier that retains the cushion. The innovation is that water will have no drag when it touches the ocean, because it will ‘disappear’ into the ocean. It will also conform perfectly to waves.
The problem of course is that to create the water curtain we must lift seawater up from the surface and then eject it downwards with enough momentum to seal the cushion. It turns out that the energy required to do this is much larger than the savings due to elimination of seal drag.
There are no water-curtain craft in existence that I know of.
The peripheral jet and water curtain ideas are two ideas that seem to get re-invented once each generation. Rigid skirts are another similar case. Many people have thought of using a rigid structure to retain the air cushion, and then articulating that structure on a system of hinges and springs to give it the desired dynamic performance. Obviously a rigid skirt is a good solution to the permeability goal, and it requires no power for the skirt itself.
Early rigid skirts consisted of simple hinged plywood panels fitted at the bow and stern of an SES. The first generation of this simply hinged the panel at the top with a door hinge. The problem is that the cushion pressure acting behind this panel results in a large force, and simply makes the panel into a plow, eliminating the resistance advantages of the SES.
To solve this, the inventors switched to a ‘balanced’ type design, where the panel was hinged about a mid-chord, and not at the edge. This results in a panel with good conformance to the 2D surface. The problem now is that any athwartships ‘shape’ to the wave is met with a single monolithic panel, which is then plowed through the wave. So the clever inventors conceived of segmenting the panel athwartships into a system of several rigid fingers, that look something like piano keys. These must of course be of balanced design, as well.
As the individual keys move, they must have some means between them so that air doesn’t escape between adjacent fingers. This requires some kind of side panels to close the gap, and these side panels will rub on each other. The friction thus introduced will reduce the conformability of the fingers, reducing their effectiveness and increasing their drag.
In practice, nobody has yet overcome these solutions with a system that is superior to the inflatable fabric skirt.
Inflatable Fabric Skirts
I classify fabric skirts into six basic families, each of which will be discussed below. There are:
- Simple curtain
- Transversely Stiffened Membrane
- Pericell / Jupe
- Bag and Finger
A flexible skirt helps reduce the air flow required to support the craft. Making this skirt of fabric will help reduce the weight of the skirt and may reduce it’s tendency to plow and drag, because of its flexibility.
The simplest type of flexible fabric skirt would be a simple curtain hanging down from the wet deck. This skirt would have to be tensioned at the bottom in order to hold down or it will simply blow up under the influence of the cushion pressure. Some of the hold-down effect can be attained by making the skirt go around the full perimeter of the craft and making it somewhat conical – tapering downward. The sloped sides of the cone and the inherent geometry of the cone will help to keep the skirt in place.
Unfortunately, the same forces that keep a curtain skirt in place also stiffen the skirt and make it more likely to drag by plowing.
Transversely Stiffened Membrane
Imagine an SES curtain skirt that is about the size of the door on a two-car garage. Imagine that it is secured along the entire top edge, and also at the two sides, but not at the bottom. Now imagine it subjected to 1 psi of pressure on one side. Obviously it is going to bulge outward, and will no longer be a simple 2D shape.
To alleviate this bulging some practitioners have experimented with transversely stiffened curtain skirts. In this case long thin flexible battens are included in the skirt, spanning the full width from side to side. These battens help reduce the transverse bending of the fabric. They may also be tethered to the ship structure for further geometry control.
Very few TSM skirts have been built, and little is known about the potential of this system.
At some sort of ‘opposite extreme’ from a curtain would be to surround the full perimeter with an inflatable “horse collar” or “inner tube” all the way around. This skirt will work. There will be a tradeoff between pitch stability and plowing / drag – a higher pressure inflation will make it stiffer, yield more pitch stiffness, but also result in higher drag. Indeed at the limit – infinite inflation pressure – this becomes simply an analog of a rigid non-compliant skirt.
In actual practice skirt inflation pressures are far below infinity, but they must still be somewhat higher than the cushion pressure. Bag skirt systems are common (indeed, ubiquitous) as stern seals in SES. In this application they are usually inflated to 5% - 15% above the cushion pressure. This produces a very soft bag which is easily deflected by incident waves.
Pericell / Jupe
The next type of skirt is to use a series of smaller conical structures. These are called “jupes” which is simply the French word for “skirt.” Each pericell or jupe looks something like the garment called a “tulip skirt.” A series of these jupes surrounds the cushion – sometimes in combination with a common bag section, as illustrated in Figure 206.
The pericell yields good vertical stiffness if the cells are conical in shape. The drawback to a pericell is that the portion of the hem of the skirt that is concave-forward is shaped to scoop water when in motion, which can cause drag, skirt damage, or other undersirable behavior. This can be mitigated by slanting the tips of the cones somewhat so that the forward facing edge is slightly higher than the aft-facing edge.
Figure 206 - A Pericell and Bag (or Jupe and Bag) skirt system
Somewhere between a curtain and a pericell lies the concept of the finger skirt. A fabric ‘finger’ is a half-cylinder of fabric, suspended from the wetdeck at an angle of about 45 degrees from the vertical. The half-cylinder has its convex face outward, concave toward the cushion pressure.
The finger skirt may be considered to be a derivative case of the curtain skirt, where a single large curtain is replaced by a series of multiple curtains. This philosophy is illustrated in Figure 207.
One may also imagine a finger skirt as consisting of only the outboard halves of a series of pericells.
Figure 207 - The finger skirt (right) explained as a derivative case of a single curtain skirt.
Bag and Finger
In a figure above I illustrated a bag-and-pericell skirt system. I have also said that a finger skirt may be considered equivalent to half a pericell. In that case it is unsurprising to introduce the bag-and-finger combination, illustrated in Figure 208.
Figure 208 - A bag-and-finger skirt system
Basics of Inflatable Structures
To further understand the skirts on Powered-Lift AMVs we must learn a few basics about inflatable structures. There are two simple facts that are all-important:
- The force balance on a uniform membrane will always result in a circle (or segment)
- The stress in an inflated segment is directly proportional to the radius
These two facts can be clearly seen if you imagine a fabric having zero stiffness. If it is subjected to a uniform load like a cantilever beam it will of course deflect. With zero stiffness the resultant forces at the endpoints can only be in the direction of the fabric – in pure tension.
Consider the case shown in Figure 209. Here the diameter of the circle is equal to the space between the supports. The total force acting on the restraints must be the integral of the pressure over the girth of the bag, resolved into X and Y components. It is easy to see that the Y components will cancel out due to symmetry, and the X component will be equal to P x D. PD is thus the total reaction, which is the sum of the two endpoint force R1 and R2. Thus R1 = R2 = P x Radius. Now what is the tension in the fabric? Is it not simply the reaction force R? Thus the tension in the fabric is t = R1 = R2 = PRadius.
Figure 209 - Basics of inflatable structures
Basic Design of SES Skirts
The most common SES skirt system today consists of full-depth fingers forward, and a multi-lobed bag aft. Since this system is common, and since an AMV-acquainted naval architect might be called upon to develop a concept design fairly quickly without recourse to consultants, I shall provide an overview of how to design this type of skirt system.
SES Bow Finger Skirts
Beginning with the bow skirt system: This system consists of fingers extending the full depth (or height) from the wet deck to the designed cushion depression. The key features of such a system are:
- Semi-cylindrical fingers
- Angled to the waterline
- Restrained at tips
The fingers are half cylinders facing convex-forward. The diameter of these cylinders is determined by the strength of the skirt fabric, since as we have seen, the tension in the fabric will depend simply upon the cushion pressure times the finger radius.
As a design parameter, SES in the 40-80m size range have cushion pressures of 0.75 – 1.5 meters of water, and have finger diameters of about 1 meter. Thus for a ‘starting point’ we may take a design value of hoop stress from these values, and scale to any particular project’s cushion pressure to estimate that project’s finger diameter, assuming the same hoop stress as the design value.
The fingers do not descend vertically – they form some angle with the vertical. This angle is a major source of the pitch stability of an SES, and is also important to the drag of the skirts. The common design angle is 45 degrees. I have seen and tested angles from 30 to 60 degrees, and there is nothing to recommend them at this time. A flatter, more horizontal, skirt angle will increase the pitch stiffness because it yields more shift of the center of pressure, but it will likely increase the wetting of the skirt and thus skirt friction. It probably reduces skirt wavemaking drag because it forms a more gentle entry angle for the cushion, viewed in profile. A more vertical angle reduces pitch stiffness, but may also increase drag because it presents a more blunt entrance angle to the cushion pressure.
Imagine if the fingers were simple half-cylinders, attached at the wet deck, and angled 45 degrees from the vertical. Now subject them to cushion pressure on the aft face. Clearly they will buckle and fold forward unless they are restrained in some manner. The restraint must hold the finger tip aft against the force of the cushion pressure, and it must not yield any effective force acting up the long vertical axis of the cylinder, as this would simply cause the finger to crumple vertically in buckling. Thus we see that the restraint could be as simple as a pair of ropes attached to the lower aft corners of the skirt and lead to the craft structure, provided that these ropes form an angle of at least 90 degrees with the finger axis.
In practice, ropes are not used for this purpose because of the next required feature: The fingers must seal against their neighbors. If the fingers were simple half-cylinders, then for any non-zero deflection they would open a gap between themselves and their neighboring finger, and cushion air would leak out of this gap. Therefore the half-cylinder of finger has straight-line extensions aftward. These flat panels of fabric bear against the neighboring fingers (or the craft sidewalls) even when the finger has moved appreciably.
In practice, the fabric extensions are carried all the way aft to serve as the restraint “ropes” mentioned previously. Of course, this fabric exists at every point along the edge of the finger, not merely at the tips, so the restraint force is applied distributed over the length of the finger, which avoids load concentrations.
The resulting finger geometry is depicted in Figure 210.
Figure 210 - Drawings of generic SES bow-finger geometry
SES Stern Bag Skirts
An SES stern bag is a simpler geometry. A two-lobed bag is shown in Figure 211. The stern bag geometry is dominated by the ratios of pressures inside and outside the chambers of the bags. Consider Figure 212, which shows a simple single-lobe configuration. The key to this geometry is that the tension in the fabric must be the same at every point along the perimeter – there is no mechanism for increasing or decreasing tension except at the end points. The controlling points then, which are under the control of the naval architect, are the two endpoints and the point of tangency with the sea surface (labeled “t”.) I have simplified this geometry such that the aft endpoint “A” is vertically above point “t”. This makes the aft lobe of the bag a complete semi-circle. The student can generalize this geometry to other cases, and indeed Larry Doctors has provided a complete generalization in his work, see earlier in this text.
There are two different pressures acting on the bag in Figure 212. The aft face of the bag sees a pressure which is P-aft = Bag Pressure minus Atmospheric Pressure. This is higher than the pressure seen on the front curve of the bag, which is P-forward = Bag Pressure minus Cushion Pressure.
Now, since the tension at point “t” must be the same on the forward and aft parts of this point, this means that P-aft times Radius-aft must equal P-forward times Radius-forward.
With the height of the cushion known, and the pressures known, the designer can then find the resulting two radii, and from these can calculate the amount of fabric (girth) required to make the stern seal. Typical stern bag pressures are 5% - 20% above cushion pressure, with the philosophy being “the lower the pressure the better.”
This geometry is sufficient to design a practical bag skirt - by applying these relationships one establishes the required girth (arc length) of the lobes of the bag, and thus the skirt.
There is one additional parameter to be considered, however: When the bag is washed aft by a wave, it may be carried aft of the transom of the sidehull. When this occurs the ends of the bag will be exposed to atmosphere and the bag will deflate. This in turn leads to loss of cushion pressure, etc.
The only sure way to prevent this is to ensure that the cushion walls extend far enough aft that the bag will still be contained by these walls, even when it is flattened against the wet deck. In terms of the figure below, this would mean that the transom should be located at a point "x" aft of "A", such that the distance |xA|+|xF| = arc-length "A-t" + arc-length "F-t".
Figure 211 - A two-lobed SES bag-type stern seal
Figure 212 - Definition sketch for a simplified case of the geometric balance of a stern bag seal
Note that a bag-type stern seal does not have end caps – the edges of the fabric simply slide along the rigid craft structure. This means that, especially when un-inflated, the stern bag can fill with sea water. To drain this water a series of small holes included, lying along the line of tangency “t”. To prevent the holes from catching the water at high speed, and thus tearing the fabric bag, a simple flap of cloth on the outside (attached forward and loose at the aft end) covers them. This flap of cloth is called a “feather” – despite not looking like one at all!
It is appropriate here to comment on the hydrodynamics of the stern bag. In an ideal stern bag there is a small daylight gap between the stern bag and the sea surface at point “t”. This gap is an interesting mess of stuff to analyze. First, the inflation of the seal wants to press the bottom edge against the water surface while the air in the cushion wants to lift the bottom edge and get out. In addition, there is a venturi effect as the cushion air jets through the gap, and this causes a suction that pulls the bag down to the surface.
An interesting flutter can be created: when the venturi pulls the bag down to the surface it closes the gap. When the gap closes, the flow stops, and the venturi suction goes away. Absent this suction the static forces reassert themselves and the bag pulls up from the surface an inch. This of course causes cushion air to flow once more, recreating the venturi, and pulling the bag back down to close the gap. In practice, the result of this is a resonant mode that is exactly like a “whoopee cushion” or clarinet reed.
Balancing those two forces plus the venturi effect is indeed the métier of a seal design specialist. In addition to the seal pressure and cushion pressure affecting the shape of the inflated seal, the venturi under its gap has to be tweaked with the correct approach angle and number of drain holes that 'bleed' air in to the venturi and, sometimes, even a trip or step along that edge to reduce the suction or down force created by the venturi. This seal-water interface pressure distribution and magnitude has been measured in tests and correlated with a rather complex model of the stern seal dynamics. FYI..the pressure at the gap can drop below atmospheric pressure.
The issue in these investigations was not seal drag, but instead we wanted to understand most or all of the mechanics affecting steal stability; some bounced and some didn't and unstable ones created serious ride quality problems. However, it stands to reason that the ones that were unstable contributed a drag component since each 'bounce' resulted in contact with the water surface.
In calm water, a correctly designed lobe seal probably has little if any drag contribution since it is not contacting the water surface and has no perceptible effect (visual anyway) on surface elevation forward of, or under, the sea itself.
SES Specialist Rick Loheed (private communication) provided some interesting comments which seem to fit nowhere else, so I include them here for the reader’s benefit:
“Our tuning trials were typically for optimizing motion controls, but during our testing we always did a “Stern Seal Delta-P Vs Speed” sensitivity test without any other variables changing. When venting the cushion for control in waves it was found running slightly tighter than design allowed a higher vent valve ‘effective bias’ dynamically, yielding a little more bi-directional control and keeping more cushion pressure longer as the seas got bigger, resulting in higher speeds with the same power.
“In observing the seals during initial trials, I used to adjust the Delta P until it was observed to contact the water, and then back off just enough so it didn’t too quickly arrive at an operating point near the optimum. I was seldom wrong by much.”
Mr. Bill McFann (private communication) added: “I always made it a point to make the yard install windows into the cushion for seal observation. I also thought they should be there so the crew could check the seals for damage, but the classification societies typically made them remove them. I think maybe a couple put covers over them and managed to keep them. Others argue for a video camera. It is never as good and they can fail- I still think the crew needs to be able to observe the seals directly.
“In calm water, watching the stern seal gap is fascinating because the water smoothly flows beneath it, then rapidly begins the rise to the surface because the pressure is off and the venturi is helping turn the flow. Everything is a bright green- it is seldom as dark in there as you would think. It looks like the smooth inside curl of a breaking wave just under and behind the aft lobe. This of course means the venturi ‘wraps’ around the lower lobe also- it does not exit flat as if shedding from a transom.
“Typically I could not see any spray from inside the cushion- the surface was usually very smooth. It may have gone more turbulent near the surface where atomization events like ligaments turning into droplets could occur more readily as the escaping air tears at it.”
The forces in a skirt system are in three classes:
- Internal forces
- Attachment Forces
- Dynamic Forces
We have already seen that the basic internal force in a fabric structure is due to the inflation pressure. If there are no stays, wires, restraints, etc., then the structure will take on a circular shape and the fabric will be loaded to a Hoop Stress which is equal to Stress = Radius X Pressure / Thickness. This stress drives the selection of the number of fingers or lobes in a seal, and is driven by the allowable stress of the skirt fabrics.
Attaching a fabric skirt to a rigid craft is not simple. The challenge is to try to create an attachment system that is continuous, e.g. a bolt rope, in order to avoid stress concentrations. The purpose of this attachment is to provide the restraint forces needed to hold the skirt in place against the “thrust” caused by the pressures. through illustrate a few methods of accomplishing this.
Sometimes the skirt must include point-load restraints, such as stays or webbing straps. In this case the manner of attaching these items requires doubler sections and grommets. Indeed, skirt manufacture is very like sailmaking, and most of the techniques for handling reinforcements in a sail are also used with skirts.
Skirts are not made in single elements. Especially in the case of fingers, it is desirable to make the skirt in segments. In the case of fingers it is normal to have each finger fitted with a removable “cuff” at the bottom edge. This is where most of the finger wear takes place, and with this method one can simply remove and replace the cuff, rather than the whole finger.
Similarly, bag segments in a multi-lobe stern seal may be made removable. This makes possible afloat maintenance, as seen in . Segmented construction also results in controlling the weight of any single component, easing maintenance and installation.
Attaching segments to each other is usually accomplished using point loads. I have seen both bolted attachments and lacings used equally successfully. Bolting is straightforward, and requires local reinforcement. In the case of lacings the system consists of simple grommets through which a cord is woven and tied, exactly like tying a shoelace or corset.
Figure 213 ‐ One type of bolt‐rope style method for attaching the edge of a fabric skirt to ship structure
Figure 214 ‐‐ Another bolt‐rope style attachment method
Figure 215 ‐ A piano‐hinge type of skirt attachment
Figure 216 ‐ Bolted attachment of fabric elements on an ACV
Figure 217 ‐ A detail of the Anti‐Chafe ring. This prevents the nuts and bolts from being damaged by contact with the ground on an amphibious ACV
Figure 218 ‐ The components of a bag‐and‐finger system, highlighting some of the attachments that take place.
There are static forces due to inflation. There are local issues due to attachments. There are also some very important dynamic forces present in even a simple skirt system. The most important dynamic force is flagellation.
Flagellation takes place especially at finger tips or any other similar unsupported edge. The trailing edge in such a situation will flutter and flap, exactly like a flag in a breeze. The trailing edge itself flaps back and forth several times a second, subjecting itself to high accelerations. Finger tip accelerations have been measured to exceed 8000 g’s - Wow! This gives rise to a form of fingertip wear that looks exactly like abrasion – see . It also, however, gives rise to internal heat build up that can destroy the skirt fabric from the inside. The rapid flexing of the fabric results in energy that shows up as heat, and can cause burning or melting of the fibers or of the rubber coatings of the fabric. This problem gets worse as fabric gets thicker, because the thicker fabric has a harder time shedding this internal heat.
Figure 219 - An SES bow skirt, where the wear at the tips of the fingers due to flagellation is clearly visible
Skirts do fail. Most failures are simply wear, rather than catastrophic-event type failures. Wrinkling, delamination, and abrasion of finger tips is common and should be provided for by designing removable cuffs. Bow skirt wear rates are on the order of one millimeter of fabric lost per hour of high speed (>40 knots) operation. This yields finger cuff replacement intervals of about 1000 underway hours.
Stern bag wear occurs at the edge of the tube. Many designers use a wear drape or feather in this location. Wear can also occur along the feather used to cover the drain holes. Stern seal wear rates are much lower, with stern seal repair / replacement intervals on the order of 5000 ship hours.
It is possible to tear a seal, say by striking a log or other obstacle. If tearing is expected to be a problem due to the nature of the operation then it is recommended to design skirts that include rip stops (similar to crack arrestors in early steel shipbuilding.)
In the extreme case, a skirt can blow out. Blow out is usually associated with snatching or snap-back loads in a wave encounter. This results in a force which is basically the same as the restraint forces and steady state pressures TIMES a dynamic load factor.
The materials used in modern full-size skirts are virtually the same as used in inflatable boats: Natural rubber reinforced with nylon, etc. The issues in selecting a skirt material are:
- Strength, to withstand the skirt forces (including local loads)
- Heat tolerance, to withstand the heat generated by flagellation
- Flexibility, to yield the comforming behavios sought in a skirt
- Adhesion between the fiber and the matrix, to ensure long fabric life
- Repairability, including the feasibility of using adhesive patches, stitching, etc.
Table 12 presents some data on two skirt materials produced in China, taken from Yun & Bliault (Reference 16.) In the western world the only skirt maker I know of is Avon Engineered Fabrications, a division of Avon Rubber (the makers of the successful Avon line of inflatable dinghies.)
At model scale, some model makers use sailcloth to fabric skirts in the towing tank. There is debate as to whether this is satisfactory, as sailcolth will not have the same weight / stress / strain properties as scaled full-scale fabric.
Table 12 - Data table from Yun & Bliault describing two skirt fabrics available in China
Table 13 - Data table from Yun & Bliault describing skirt materials and life from some built SES and ACV