Vanguard Thrust Duct

So there are a variety of ways to make a large duct for hovercraft. I chose this one as I thought it seemed like it would allow for experimenting with a few parameters - most notably having a diverging exit. The more classic UH (Universal Hovercraft) method (example here) was going to be my choice until I saw Bryan Kearns site.

First some math for the divergence part. The suggested best rate of expansion after the prop is said to be about 4 degrees (that is 4 degrees each side - 8 degrees from side-to-side). I calculated what that would translate to in increased radius for the formers and cut two disks and added spacers.

Trouble is that I used a small 1/4" hole for the center point when making these and during the drilling out to 1" for the shaft shown it managed to move the center a mm or two (sorry for mixing inches and mm - my natural tendency is to use metric but its hard to do so in this country!). Anyways it was easier to cut new disks and use the final 1" hole when routing the disks out. The picture below shows the new disks and the ply wrapper untrimmed. Here starts the complication of the diverging exit. If the exit is parallel the disks are the same size and anything you wrap around it can be straight and easy. Make them different and everything wraps into a cone so must be cut into very large radius curve in order to fit. Rather than deal with getting that correct I simply attached the four pieces of ply left extra wide on both sides with slightly beveled ends to make a poor-mans curve. It worked. You can see how the overlaps rises and falls.

You can see the amount of divergence here.

Once the ply was complete I trimmed it with a small trim saw by rotating the former and clamping the saw to the stand.

The large curve cutting theme has to be continued with everything - including the foam. The foam in this case is 1" 1lb density whitebead foam. I held the 22" wide sheet of foam around the ply and marked out the edge lines to give me the the curve to cut - remembering the foam extends 5" forward of the ply and flush at the back. I used my hot-wire to rough out the curve - again final trimming is done on the former.

To get the required 3" of foam I added three wraps. I first started filling the gaps with expanding foam (Great Stuff) but later removed it as I thought it would interfere with the hot-wire trimming. I could fill these gaps later. I made a hot-wire jig to cut a thin strip of foam for the transition to the ply (ply is 4mm okhoume marine ply BTW).

The bellmouth shape is important to get right. Luckily there are a lot of smarter people than me so I chose to use an ellipse shape - with a 5" major and 2.25" minor radius. First off I needed to rough trim the front so I could form the ellipse.

I wanted to try to cut the ellipse rather than sand or carve it my hand. I bought some nichrome ribbon (like the wire but 3/16" wide x 0.018 thick) and formed it to the shape of the ellipse and supported it close to its ends. Vendor plug:- M PAK Systems for nichrome wire and ribbon.

As you can see in the picture below it worked rather well and I was pleased. The only downside was the the flat bit cutting close to the ply was so thin that it seemed to pit the foam a bit. I think next time I would add thicker foam and therefore cut more off.

The waste from this process is a little better than a huge pile of foam dust from sanding.

I used the same idea for the outside simple radius. The tool works well and I think could be made in many different guises. All I needed to do was sand by hand the edges and the glue lines.

The final hot-wire was a trim of the outside flush to the trailing edge. I left about 1/2" of foam thickness at the back. This will give me some extra strength for some hard points to support the rudders. Then add a flat side for the bottom to make it easier to rest on the deck. This duct is not getting foamed/glassed on the hull itself.

And finally it is glassed with 9oz satin weave cloth (56x56) and epoxy. I had a little trouble using the old Epoxical so got some new marine epoxy from Tap Plastics and it sped up the process a lot. Everything wetted out much better. Still I have to say working with epoxy is harder than polyester or vinylester. There also was very little stick to the white foam when applying it. It seems to have stuck once cured but there is no cling to it when wet. I also wonder if I used too little epoxy as I can see parts where its not totally saturated to the foam. Perhaps I should've liberally coated the foam first then laid up the cloth? Oh well, too late now.

Excuse the Blur in the foreground. (bike, not the shoddy camera work)... at least it was easy to get off the formers being diverging and all...

Here's a shot of the duct on the hull.

And finally some pics of the whole assembly ready for mounting.

Household interjection

Oh and in the middle of this project I had a little side project. Its not hot here often but I like my comfort so I installed a 3-ton A/C unit. At least everyone in the house can enjoy this and try to overlook the smell of resin setting in the garage...

Weight

The weight of the unsanded, unpainted duct is 16lb 6oz. This means there is not a significant weight savings through building a duct. I can't recall exactly the weight of the prop cage on the CanSev but it was about 20lb and it was 72" in diameter (made from 3/4"x0.063 aluminum tube and 14ga 1x2 welded mesh). But as everyone else has concluded it looks much cooler!

Cutout

It became clear that in order to get the prop in the right part of the duct that I would need to make a cutout at the bottom for space for the idlers and the air-drive itself. I cut the fiberglass with a high-speed diamond cutting tool then just used a hacksaw blade to cut the rest. A bit of sanding and more epoxy and paint later and it was done.

Installation

The duct is finally mounted. The support tubes are 0.75x0.065 aluminum tube with a short length of round solid at each end to support the bolts. I weighed the duct and rudder system (and the morse cable) - they were 29.5lb total. Tip clearance is about 1/4" - I was shooting for less but instead opted to push the prop further back in the diverging duct. Hopefully this works OK. I read that tip clearance for a diverging duct is not as important as you might think. The mounts make it feel pretty rigid although I am not sure how much structural strength it has. It is foam, ply and fiberglass after all....

Alignment Issue

I noticed that when the prop is exerting thrust that it moves forward quite a ways as the air-drive frame is not super rigid fore-aft. With the prop extension and the pulley overhang this means the prop moves upwards a bit too. Applying 100lb of "thrust" (with a large spring gauge) bought the top of the prop into contact with the duct. First thing I did was tilt the duct forward at the top a bit so that it is properly aligned when under thrust conditions and I discovered it is out-of-round a tad so I added a tensioning cable (see pic below) across the top to pull it back into round. This lifted the duct top about 1/4" and it was round at that point. These two efforts meant that the tip clearance was back to 1/4 - 3/8" all around when there is thrust. The top clears a bit more with no thrust.