Text Box:  
Fig 1 Header
The second air cushioned craft - A model ferry



Contents - scroll to read as one text or click on title




Design decisions for the new ACV


The lift system


Building the lift system

The intake

The divergent transition from round to rectangular

The divergent 180 degree bend

The "hull"

The "bow" and the side pods

The stern and moveable dam

The upper deck and the intake fairing

The drive


Assessment of this design


The first hovercraft was very much an exploratory design. It was not built as a fun boat but to find out whether physics could be applied successfully to a fan unit to use it first to power an air cushion and secondly to make it into a thruster. Whilst it took quite a while to build it was still only the minimum effort required to find out whether the design would work as planned. I think that it was much more successful than I envisaged when I started. I am much more confident about my understanding of fans. Nevertheless I must build on what I learnt from this proof-of-concept design.


The fans were cheap but I do not know whether this mattered. Certainly two thrust motors have failed which means that I have worked them to their limit. That limit came when the battery voltage was 8.4 volts. At this voltage I could hear that all three fans were producing as much thrust as is possible because the flow over the blades was just breaking down and causing the noise level to increase. In fact I reduced the speed slightly to get rid of the noise. The two thrusters are used for steering as well as driving the vehicle and, whilst I thought vaguely of increasing the speed of one and reducing the speed of the other to steer, in fact the drag on the craft was so low that the only way to steer was to stop one fan. This left the vehicle to be driven by one thruster. Surprisingly this does not reduce the speed of the vehicle by very much but the turning circle was quite large and I usually sail it when no one else is afloat because it is not really very manoeuvrable. Any attempt to make a thruster run backwards and aid the steering makes a lot of noise but little difference to the turning circle.


The lift system has proved to be very effective and I cannot think of any way to improve it in principle although a fan of better design may make it work better still. The practical details of the essential ducting appears to be satisfactory and the recovery of kinetic energy appears to take place.


After a time I became acutely conscious of the fact that the thrusters were very large for the thrust that they produced and that the vehicle could rightly be criticised for having a lift system that occupied three quarters of the volume of the vehicle and the rest was taken up with the electrics. I felt that I could not have answered had anyone asked whether it could be modified to be the basis for a practical vehicle like a ferry. A ferry riding on an air cushion might find acceptance on a continental lake where it would make little noise and almost no wash yet go at quite high speed. I felt that I needed to make a model in which there is a sensibly sized space for passengers. If the Seacats are anything to go by that space should be just a rectangular floor with blocks of seats and, no doubt, a refreshment bar and loos. On the full size the essential mechanics is found a place wherever it will fit in and the same could be applied to a model. When I considered the possible layouts for the model it became clear that, in order to create passenger space, the lift system would have to be made to occupy much less space and the model would have to be longer.


The length of the first craft is 32 inches and it was determined by the size of the piece of liteply that I bought at a regatta. Ply from a shop could have a length up to four feet. I set a practical limit of 54 inches because this will go across the back of my car and thought in terms of 48" plus. But, before I went ahead, I needed to be sure that the existing lift system on the proof-of-concept model could go faster without running into some limit that I could not foresee. I dug out a fast electric boat to use as a tug and, with someone driving the tug, towed the hovercraft with the lift system and the thrusters going. It went at about twice the normal speed so there is no limiting factor in this speed range.


This cleared the deck for a new air cushion vehicle that might be the basis for a real ferry and would certainly look a great deal more attractive than the proof-of-concept model.


Design decisions for the new ACV

I do not think that a model-sized, fast, air cushion vehicle propelled by fans is possible. The laws of physics get in the way. If the new craft is to go twice as fast as the first one the thrust will have to be at least doubled. Scale effects mean that fan-driven thrusters might be possible on a large vehicle but, at model size, the required thrusters would be very large and cumbersome when compared with the vehicle as a whole. It struck me when I saw the Hoverspeed craft that its four airscrews were very large and anyway I cannot use airscrews because they are not appropriate on a public lake. This leads me to consider using water screws.


This has attraction. First water is much more dense than air and much less of it has to be pushed about at much lower speeds to drive a waterborne vehicle. The vehicle must have side pods for buoyancy when off the cushion and these could be adapted to contain the motors, step-down belt-drive, and drive shafts and not encroach unduly on the floor space or the roof space. Furthermore the water screw can be reversed to much better effect than a fan that becomes very unsatisfactory when reversed and very noisy. This would mean that steering using the water screws is likely to be much better than that using fans and, at the sort of speed envisaged here, thrust can be increased at will by increasing the propeller speeds or by changing the propellers. The whole drive is much more adaptable than the use of fan-powered thrusters.


If this is to be a design for a ferry I have to think about the passenger accommodation. I have said that a Seacat has a main concourse with blocks of seats and a smaller upper deck with a similar arrangement. I could have two decks with the upper one being removable to give access to the electrics This would mean that my model would have two rows of windows and these could be styled to look reasonably attractive although it is bound to look like a box. The lift system would have to be adapted to fit in physically with these two decks. Where the original lift system occupies 23" the new one will have to be much shorter perhaps as short as 12". Then, if the hull is say 54" long there would be 36" notionally to accommodate passengers.


There is not much choice of commercial fan units and the GWS unit that I used on the first model has proved to be satisfactory so I am rather limited to using an identical unit. This meant that I have to consider how the required duct work could be rearranged to shorten it by 50%. When I was building the first model I looked at the duct and pondered on whether it would be possible to make the 180 bend divergent. The geometry looked to be troublesome and I decided to make this return bend of uniform cross-section. It was still troublesome. The consequence of this decision was that most of the divergence had to be in the duct between the two return bends. This made it long and it occupied about 2/3 of the available space. If, now, I want to reduce the space occupied by the ducting for the lift system the first return bend will have to diverge. This will not be enough on its own and the craft will have to increase in length from the existing 32".


There were some unknowns in all this. I was proposing to use the same fan for the new lift system as I used for the old but now the length of the channels under the hull were going to be 16" longer, a step up of underwater length of 50%. There was no certainty that the fan was man enough to maintain a flow under this extra length. I reasoned that if I made a lift system that was 48" long and tried it in the fish pond when fully loaded and found that it would not lift I could cut the length until it would lift just like the first one.


If water screws are to go under the side pods and outside the lift system there was no certainty that I could avoid cavitation. Most outboard motors use propellers that are quite near to the surface and have a relatively small anti-cavitation plate. I thought that I could make anti-cavitation plates that are rectangular and part of the side pods and fitted so that they just rest on the surface when the boat is up on its cushion and running. I checked the feasibility of this using my fast electric boat that was used for the towing test and it did not need to be very deep without a plate to stop cavitation. I reckoned that if I was generous with the area of the anti-cavitation plates I might avoid cavitation and if necessary I could extend my anti-cavitation plate.


With these thoughts I decided to start building.


The lift system

The first thing to sort out was how to make the existing system more compact and fit into a practical design for a model ferry. The circular intake of the first design did not lend itself to blending into a design for a ferry with two decks and lots of windows. It would be more easily blended in if it were to be rectangular. It seemed to me that the possibility that a rectangular intake would supply air at uniform speed to the fan was very remote and I did not entertain the idea. An intake that started square and changed to round was as far as I was prepared to go. I thought that it was worth a try but it was a serious risk because the performance of that intake in delivering air to the fan blades is crucial to the success of the whole design. I thought that I could build the whole lift system and then check to make sure that it would lift and, if absolutely necessary, change back to a round intake.


Text Box:  
Fig 2

As I had the same fan with a cross-sectional area of about 6 square inches and I saw no reason to change the outlet area from the divergence to the channels between the strakes from 15 square inches that I used before to give an exit profile of 15" by 1". What I needed was a steady increase in area over a length of about 22". There could be no question of doing this without making a diverging 180 bend. It was then about putting dimensions to the three parts that make up the divergence.


Figure 2 is what the finished lift system looked like. Air comes in through the intake which changes from square, to suit the design of the boat, to round to fit on the fan unit. The divergence starts with a change from round to rectangular and then changes to a larger rectangle in the 180 bend and finally to the full area in the third divergence before the final 180 turn to go under the "hull" between the strakes. It is much more compact. I think that it is the most difficult piece of modelling that I have ever done and I have made some difficult items in my time. The difficulty arises from the need to have the finished surfaces on the inside and to have splitters in the divergence that must marry up at each joint.


I finished this lift system to the point where it could be tested and it lifted 8 pounds with what sounded like half speed. There was going to be no trouble making a likely lift. (In the event the finished craft weighs 11 pounds 7 ounces or 5.2 kilos, including 10 ounces of cyanoacrylate, and as soon as the lift fan is switched on it pops up like a cork and the lift fan does not run at top speed.)


Building the lift system

The intake

The intake must be built on a jig. In my case it is a square base plate of medium density fibre board with a peg fitted centrally with the end turned to fit the body of the fan unit. Figure 3 shows the body on the jig with a collar that is made by lamination from 1/64" ply and stuck with epoxy glue. The rest of the construction is made using cyanoacrylate.



Figure 4 shows the frame for the intake with four ribs to join the frame and the collar. The ribs have a radius to blend into the collar and the intake has a cross-section at any level made up of a square with radiused corners. During the transition, the size of the radius increases. The shape of the ply linings can be drawn and the figure shows the four pieces in place ready for the sub-ribs to be cut from liteply and fitted as shown in figure 5. Then the curved "corners" can be lined as shown in figure 6. The lining is done in lots of strips without being too fussy and then carefully finished with epoxy-resin filler like LeakFix that can be dressed to give a smooth inner surface.


Figure 7 shows the finished intake as installed on the model and figure 8 shows the fairing that completes the intake to give smooth flow from behind the intake and round the fairing and into the fan.


I am aware that some readers will think that this is a great deal of fuss about nothing but this is a crucial part of a lift system that lifts well over 11 pounds using less than 35 watts of power supplied to a cheap fan unit. I do not think that anything can be left to chance. Air is remarkably non-cooperative.


The divergent transition from round to rectangular

I have discussed at length the problems of designing the divergence for the first air cushioned vehicle. It is on this web site. In effect one must design so that the whole divergence is equivalent to a conical pipe with a divergent angle of about 8 degrees. This first divergence goes from a diameter of 3" to a rectangle of 4" by 2" in 3.125".


The construction is the same as for the intake but now there is no radius to contend with and its shape is made up of flats and parts of cones.


Figure 9 shows the basic carcase. Note that two of the ribs diverge and two converge. The collar is again laminated from 1/64" ply and the flange is laminated from 1/32" ply. Figure 10 shows the sub-ribs that will be lined to give the shape of a "bent" cone.


Figure 11 shows the planking. It looks rough on the outside but, again, it is finished internally with LeakFix and given a good finish that is varnished.


The air cannot be allowed to find its own way through the transition. We want a rise in pressure from both the slowing of the flow and, if possible, the conversion of the kinetic energy of rotation in the flow from the fan to more pressure energy. Flow straighteners are required and a streamlined cover for the motor. These are shown in figure 12. There are four straighteners in this transition but only three in the fan unit. This mismatch cannot be resolved but the four flow straighteners do match the next divergent bend.


The fan unit fits between the intake and the first transition in the two collars and the three held together with red plastic tape as shown in figure 2. This facilitates the servicing of the motor if necessary.


The divergent 180 degree bend

The nearly-complete bend is shown in figure 13. I suppose that it would be relatively easy to make if it were not for the need to divide the flow into four channels. There are those who might say that this is gilding the lily. The fact is that when fluid flows round a tight bend like this it sets about changing the flow pattern into that for a free vortex and largely succeeds. It even starts upstream of the bend. Once the air is through the bend this alteration in the flow pattern is reversed and, in the process, energy is lost. This loss can be reduced if the flow is split as I have done here.

Text Box:  
Fig 13

The snag for the builder is that the splitters have to line up between the several parts of the duct otherwise any gain from fitting the splitters is lost at the poorly-fitting junctions.


Text Box:  
Fig 14
The first requirement is to draw the shapes of those three curved pieces of ply for a transition to 6.75" by 1.5". These so-called pieces of ply are really two walls of the bend and a splitter and they are laminated from 1/64" ply by forming round specially made ply formers that are of slightly smaller radius than that finally required because the laminations tend to spring open slightly after forming. The main thing is to get the splitters straight across the joints without any secondary support.


Figure 14 shows what is fixed at the outset. These are the shapes of the inner and outer walls, the central splitter which is flat and the flanges at inlet and outlet. The inlet flange matches the one on the first transition but the outlet flange will be stuck to the "floor" of the ferry.


Both inner and outer walls have slots cut into them to take matching lugs on the splitter and the splitter has slots cut in it to take lugs cut in the second splitter that will be fitted in two pieces. It starts with the inner curved wall being attached to the ply support in the middle so that the inner wall is truly cylindrical. The partially completed bend in figure 14 is effectively the carcase of the bend.


Going back to figure 13 we can see the bend with the centre splitter in place and with one side planked with balsa. Note the balsa strip attached to the centre wall to give a jointing surface. The centre splitter has yet to be aligned on the nearest half.


Figure 15 shows the bend nearing completion with a support for the splitter as glue sets.


The assembled intake, fan, divergent transition and the return bend is shown in figure 16. Once inverted it is ready to go into the "hull" where the final part of the divergence can be constructed.


The "hull"

The hull starts off as just a sheet of 1/16" liteply that needs sides and front and back plates to make it into a float. Figure 17 shows the ply base and sides and the inside of the front return bend that will carry air under this hull. This inner bend is made from three layers of 1/32" balsa laminated round a dowel to give a one inch outside diameter and fitted with internal semicircular stiffeners to keep it cylindrical.


Figure 18 shows the next step of fitting semi-circular splitters that will lead to the strakes under the hull. They will create a total area for flow of 15" by 1". The return bend is completed with another half cylinder of balsa again made by lamination. It is shown in figure 19 where the fan assembly is shown in its intended position.


It is possible to see in figure 19 that the forward return bend comes out under the hull to feed the spaces between the strakes shown in figure 20 where the side walls have yet to be fitted.


Text Box:  
Fig 21
This cylindrical surface is not suitable to be the bow of this vehicle. It can have shaping fitted to it but that shaping is a compromise between two requirements. Both requirements are important. My ultimate decision is evident in the photo at the head of this article. This small air-cushion vehicle will inevitably have to deal with wind-affected surfaces, that is popply water or ripples and possibly waves and the curved surface that will have to do this is shown in the header picture of this article. One might argue that it should be higher to handle larger disturbances but the second requirement is to have smooth flow into the intake. Ideally all the inlet edges of the intake should be the same but that is not possible. At least the shaping that is fitted has no sudden discontinuities and the sharp edge that I suppose is the bow is quite a way from the intake.


Figure 21 shows the shape of the lower surface of the "bow". The vehicle has to run with the lip of the outlet from the forward return bend resting on the surface to form an air seal.


Figure 21 also shows the stiffeners for the bottom of the hull and the position of the divergent return bend. The outlet from that return bend has to be joined to the inlet to the forward return bend and be split into six separate channels.


In effect this is the end of the important parts of the hull. Clearly a stern is needed and there will have to be side pods.


Examination of figure 21 shows that the divergence cannot be completed by making divergent, straight ducting because the air must flow smoothly out of the divergent bend and be guided all the way to flow smoothly into the forward return bend. The channels will have to be curved, that is, to have S bends.


Figures 22 and 23 show the bends and how the divergence was built The vertical splitters were preformed by lamination. All the inside surfaces were sealed with sanding sealer sanded smooth and varnished.


The "bow" and the side pods

Reference to the heading photograph shows that the bow blends into the roof of the lower deck. This then fairs over most of the duct leaving the rest to be faired over with the exception of the intake. Some additional structure is required and this is shown in figure 24. In figure 25 the bow is fully shaped with 1/16" liteply and the fan assembly is shown in position. The craft has its side walls fitted and is ready for the side pods. The joint between the first divergent transition from round to rectangular and the divergent 180 bend is made using 10 BA brass screws into the ply which was drilled and tapped.


The essential function of the side pods is to increase the buoyancy over that of the bare hull when the model is not up on its air cushion but it has the new function of housing the motors and the propeller shafts and to incorporate the anti-cavitation plates that I hoped the craft would ride on.


I bought two opposite handed Graupner Speed 480 motors and two 39 mm propellers of opposite hands in their Hydro-prop range. The propellers are designed for surface piercing but they are nothing special. These motors were said to be suited to direct drive but I was doubtful. I decided to fit them with direct drive and add a step down drive if it was needed.


The structure of the side pods is shown in figures 26, 27 and 28 and the anti-cavitation plate is seen to be just a rectangle of ply but the pods needed to slope upwards in front of the anti-cavitation plates just to reduce drag.

Text Box:  
Fig 29

There is nothing special about the forward shape of the podsText Box:  
Fig 28
but I fancied something that might catch the eye and I like making funny shapes and watching them form from bits of balsa. All that was needed was birch ply on the vertical surfaces to resist knocks and balsa sheet facing over the rest. The result is shown in figure 29.











The stern and moveable dam

Text Box:  
Fig 30
The craft needed a movable dam like the one on the first hovercraft. The actual stern is no more than a flat plate to make the hull watertight but it also carried the operating gear for the dam as in figure 30. It is normally covered as in figure 31 to make it look presentable. The dam is shown from underneath in figure 32.


The upper deck and the intake fairing

Text Box:  
Fig 33
The final form of this air-cushioned craft is shown in figure 33. It turned out that the two decks at 3 inches per deck blended into the duct work of the lift system quite well. A few curves here and there and some paint stops the boat from looking too angular.


Clearly there was quite a lot of ordinary modelling to building the top works. I had been working on the basis of a notional scale of about 1/27 that gave head room of about 3 inches. This odd scale fits reasonably well with the quite arbitrary size of the basic structure of the craft.


Then, if there are two decks, the duct work can be concealed. Figures 34 and 35 show how it all fitted together. The sides of the lower deck are in 0.8 mm birch ply and the Text Box:  
Fig 34
window holes were cut out on my Text Box:  
Fig 35
milling machine because they were too long for me to cut using a hand fret frame.


These two pictures show that the proportion of space that could be free for passenger accommodation is a very useful part of the whole. For a full sized vehicle the lift system would proportionately take up much less space because of the effects of scale.


The only technical part of the top works was the fairing over the intake and fan. It is tempting to think of the flow into the intake as just coming from ahead and flowing in smooth curves into the intake. I knew that the flow would also come from behind the intake. I checked with smoke with the fan running and smoke moved quite quickly from behind the intake and flowed round the lip of the intake as it is in figure 33. It followed that the fairing over the back of the intake should be curved. This fairing had to be removable to service the fan if needed. The outcome is shown in figures 36 and 37.


The rear view of the duct work in figure 34 shows how little space the bends take up. The supports for the sides is also evident. The glazing was not fitted at this stage but provision was made to slide the windows in after construction and painting.


The drive

The basis of the design is to use an air cushion to lift the body of the craft out of the water and to then drive it with two water-screws. On a full-size craft one might expect to see the propellers and the lift fan driven either electrically or hydraulically from diesel engines. For a model the drive must be electrical. I am not persuaded by all the hyperbole that goes with brushless motors, after all they are only the old induction motor that is used everywhere on alternating current mains electricity made to small size and supplied with a variable frequency controller. They are not easily reversed and it is hard to see any valid reason to change from the well-tried commutator motors that are normally used in marine applications.


I have said that Graupner make handed Speed 480 motors that are designed to be used for direct drive and have a specified direction of rotation. This permits the improvement of the performance of the commutator by minimising the sparking at the brushes. It is necessary to have a speed controller for each motor and, if a computerised transmitter is to be used for steering, the speed controller must not have any automatic system to match the maximum stick travel to the maximum speed. I used Electronize speed controllers and these permit the adjustment of the motor speed corresponding to maximum stick position.


The Electronize speed controllers work by repeatedly turning the motor on and off and varying the ratio of time on to time off. This used to be called a mark/space system and the variables were the mark/space ratio and frequency that the motor is turned on and off. Electronize seem to be of the opinion that it is best to use a low frequency for the low speed operation of the motor and high frequency at high speed of the motor and provide a facility to do this. I used it.


As soon as I could I tested the direct drive and it was immediately obvious from the current drain that direct drive was not an option for this combination of motor and propeller and I fitted a toothed belt 2.5/1 reduction gear using wheels and belt from www.motionco.co.uk The motor then is mounted inside the hull as shown in figure 38 and the main wheel on the propeller shaft in figure 39 the tensioning pulley is a ball race. The cooling jacket is a piece of 3 mm bore, thin-walled, silicone tubing from a fishing shop wrapped round the motor eight or ten times and held on with adhesive tape.


At maximum speed each motor draws less than 10 amperes and the lift fan draws less than 5 amperes from an 8.4 volt NiMH battery. A battery having 3.3 ampere hours of charge drives the craft for between 10 and 12 minutes.


Assessment of this design

This, the second of my air cushioned vehicles, added nothing new to the physics of the lift system over that for the first. It is just a rearrangement of all the parts to produce a different model. Success depends not on the application of more science but on my knowledge of how fluid flows and on leaving nothing to chance.


When this model was partly completed but capable of being tested I took it to the lake. Someone watched it go and then asked me whether it was a hovercraft driven by water propellers. I said yes and he asked whether designing a model without a flexible skirt did not defeat the object of a hovercraft which he perceived to be to operate on water or land. I did not argue but I do have a view.


The hovercraft is just one of many designs for amphibious vehicles. I watched the SRN 1 on the runway at Farnborough. It was supposed to be used by the military. The idea of creating an air cushion obviously worked but I had no confidence that any rigid vehicle that relied on an air cushion for its support and hovered at about 9 inches would ever cross a typical common covered with gorse bushes or indeed take to the road in times of military action. The evolution of the deep skirt was intended to make the hovercraft into an all-terrain vehicle but success has been very patchy. In my view it is not necessary to regard an air cushion and an inflated skirt as inseparable nor is it necessary to further identify a craft having an air-cushion with crossing water and land. I think that one should look to design air cushions to suit particular applications and, if a skirt is desirable and feasible then so be it.


The ferry for use on continental lakes is just an application of the air cushion. I made my case as follows "A ferry riding on an air cushion might find acceptance on a continental lake where it would make little noise and almost no wash yet go at quite high speed." I restricted it to a continental lake because wind-driven surface waves could not reach the size of waves in say the English Channel and the air cushioned ferry could be large enough to cope with these smaller waves. There are lots of other sites that have similar conditions.


As such a ferry will never be called upon to make its own way across the land the design of both the air cushion and the drive can be chosen to suit the craft on water. The side wall design seems to me to be best.


After I had written this I looked up conventional hovercraft with skirts and came across "the hump". Occasionally I had wondered how, for a hovercraft with a skirt, the flow of air forwards from under what passes as the bow, behaves as the speed increases. As the idea of a skirt makes no sense if the craft is not to be all-terrain and I was restricted to being afloat on a lake I did not pursue the problem thinking only that it is troublesome just like the transition to planing for a displacement hull about which I know nothing. By chance the sidewall hovercraft of the design above does not have that problem because there is no perceivable bow wave. I checked the speed of this model and it is about 8 feet per second or 5.5 miles per hour. The sailing fraternity use "hull speed" in handicapping equal to 1.4 times sqrt l knots where l is the waterline length feet. This is a guide to the critical speed at which the resistance to motion starts to increase very rapidly with speed. This craft is 4.5 feet long so its hull speed is 3 knots or 3.5 mph. This ferry is well above that so there is no "hump" problem.


So this miniature craft that was made solely to find out whether an efficient lift system using a fan of very low power could not only lift a craft of considerable weight and permit the craft to run at a speed that is quite fast but also do it with the lift system occupying much less than half of the available floor area has been successful on all counts.


This second lift system appears to be more efficient that the first. It lifts 11.5 pounds on 5 square feet of lifting area instead of the 7 6 pounds on 3.3 square feet of the first using the same fan unit. One might argue that the increased lift is due solely to the increase in area but the fan has to drive the air an increased distance and there must be some limit to that distance. I think that this model is close to that distance. As a matter of record the model starts to "nod" as the battery becomes exhausted and the voltage drops (see video at YouTube using my name as the address ivorbittle. It behaves as if the air starts to flow in waves under the hull.


In the first model the first return bend was of uniform area and simply redirected the flow. This caused a loss that could be avoided by the use of a divergent return bend and this appears to have been the case. The system lifts 11.5 pounds (5.2 kilos) without any difficulty with the fan running quietly at less than full power on 8.4 volts. The power from the battery for the lift system, including that lost in the battery, is less than 40 watts. I can see no leakage from the system except of course the discharge of air under the dam at the stern.


The propeller drive is compact and uses less than 150 watts. The run time, mostly on full speed, is up to 15 minutes from a battery of 5,100 mAH so the apparent current drain is of the order of 20 amperes which matches my figure for power above.


I think that I should give the important numbers for this model The ability to lift also depends on the area of the lifting surface. For this boat the area is 15" wide by 49" long, 380 mm by 1250 mm. The mean pressure on this surface is 0.017 psi or 0.00108 bar. This equates to 0.43" of water gauge or 11 mm. The change in area between the first and second craft was a 50% increase and this seems to have increased the lift without affecting the performance of the fan in producing a rise in pressure. When one considers the fact that there must be an increase in the volume of water displaced just to satisfy Archimedes' principle this is surprising. The craft is lifted by about 1.5" (35 mm) by the air cushion. The power consumption is 8 watts per square foot and the lift per square foot is 2.3 pounds. I can think of no way to reduce this still more except perhaps by making all the duct work by accurate injection moulding. That is almost certainly prohibitive.


I came across figures for the lift to power ratio for full sized hovercraft. It seems that the early craft needed 73 kilowatts per ton of dead weight and that, with development, this has been reduced to 15 kilowatts per ton. My model air cushioned ferry requires 7 kilowatts per ton.

Figures 40, 41 and 42 show the ferry under way. The speed is adequate indeed it is probably too fast for me to control if there are other boats on the lake to avoid. This suggests that, if I chose to work on the drive system, the air cushion is able to handle higher speeds. In truth, there is not much in the water, certainly the strakes are probably clear of the depressed surface and only the sidewalls are dipping into the water just far enough to provide an adequate air seal. It is likely that the speed could be doubled with a change of motor and propellers. Figure 41 shows that the steering is better than that of the first craft largely because the transmitter is set up to make the motor that is on the inside of the turn go into 25% reverse and so produce a drag. If the motor is not reversed the vehicle goes into a wide turn and it is then better to come off the cushion, turn, and then go back onto the cushion. If this were to be a full-sized craft it would probably be operated as a displacement vehicle when manoeuvring to come alongside. In displacement mode it can be turned in its own length.

Text Box:  
Fig 42

The craft rides on the lip of the forward return bend that feeds to the strakes and on the two anti-cavitation plates as planned as is evident in figure 42. This trim is obtained by movement fore and aft of the electrics pack to move the centre of gravity and then by adjustment of the dam to raise or lower the stern.


I said at the outset that a suitable application for this design would be a ferry for use on a continental lake and that it would have a small wake. Figure 43 shows the wake at full speed. The wake that is evident is not from the bow but from the side walls where it is partly generated by friction between the sidewalls and the water and partly by the anti-cavitation plates. It may be possible to adjust the dam to reduce this wake still more


I took the model to the lake when the wind speed was about 20 mph. The water was very popply and, in the short time that I had, it was evident that the bow could deal with the popple. When the craft was running off the cushion the sharp bow allowed waves to lap up to the intake but no water went into the fan when it was started. The boat ran into the wind but a sudden gust spun the boat through a U-turn because there is nothing in the water to resist a cross-wind. This must be a problem for all hovercraft but, here, it could be solved with retractable dagger boards.


The appearance of the craft is important. It seems to me that styling of any boat calls for quite different ability to that needed to do the engineering of the design. Some people have a flair for creating an attractive appearance but I have no skill in that direction, I just plod along designing as I go with an eye to appearance and if it works it is a bonus. I am quite pleased with the general impression of this one and the way that the intake duct blended into the passenger decks. The simple "bow" gives the impression of thrusting forwards. The side pods came out quite well. The thing that jars is the bridge. It seemed to me that those controlling the craft will need to be able to see in all directions when in transit between their start and finish and be able to see the edges of the boat when coming alongside and manoeuvring. My bridge permits that but it is very angular. There is a flag staff to go at the stern and this softens the lines to some extent but it is still angular.


Most of my boats have given me pleasure but this one is special. It is the first boat that looks like a real boat but has no discernible bow wave. At one time I would have thought that it is impossible. It was worth making.


Text Box:  
Fig 43
There is a short film on YouTube that was taken to examine the behaviour of the craft. It can be accessed using my name as one word, that is, ivorbittle.


Ivor Bittle Christmas 2011