Nitro Powered RC Hovercraft

My interest in hovercrafts was first incited by Junkyard Wars. I absolutely loved that show. From the constantly changing cast of characters to the equally unique challenge posed every week, I couldn't get enough of the weekly scrap heap showdown. I really wanted to make my own version of whatever the teams where creating any given week, but the hovercrafts really caught my attention. Something about the odd way hovercrafts slide across the ground like everything is covered in ice caused dreams of such a machine for me to pilot. I'm a little big to ride the radio controlled hovercraft that resulted from my obsession, but it's still a fun vehicle to drive around...

For a dimensioned drawing, check out my website: thewidgetforge.com

The design is based around an 0.049 cubic inch un-throttled Cox engine that came off a control line airplane. It is very powerful for its size and can really move the craft along. Unfortunately, the downside is often difficult to start, especially if you feel the urge to run the craft in the cool air of fall. The glow plugs which cost about $10 and are difficult to find also burn out very easily. There is a fine line between enough juice to start the engine and a fried plug. It would help to have a larger, throttled engine for the greater control and less finicky starts. If you're interested in a Cox engine, eBay has many on sale, often in lots with plenty of extra parts included.

To help with the control issues of an un-throttled engine, I came up with the most unique part of the design: The radio controlled deflector which splits the air between thrust and lifting. It can deflect a maximum of 50 percent of the available thrust to the lift the craft or close off the lift duct completely. While it lets you control the ride height of the craft to some extent, it is most useful for panic stopping by completely cutting off air to the skirt, allowing the craft slid to a stop on the ground.

Steering consists of two balsa uprights which support the two rudders. The rudders are two inches wide and provide plenty of maneuverability, although more rudder surface would probably make it even more maneuverable. A cowling would likely make the rudders more effective too, but the hovercraft is very maneuverable as it is, so it isn't really worth the added frustration in starting the engine to add one.

Materials and Parts
The body is pink insulating foam from a local home center. This stuff works great for this purpose: it is easy to work with and quite stiff on its own. It comes in a variety of thicknesses and usually two foot by eight foot sections. The one inch thick foam is plenty strong for a craft this size. The rest of the structure is made of 1/8 inch thick balsa and, for the engine mount, 1/8 inch plywood from a local craft store.

The remote control system is a two channel set-up meant for cars, boats or other crafts that ride on the surface. It has a range of about 100 feet which is plenty for this hovercraft. The 2 oz fuel tank, fuel lines, fuel, control horns and pusher prop were all bought online at an RC specialty site.

After I laid out the main features on the base with a red marker, I routed the necessary channels with a Dremel using the small base attachment that comes with the tool and a 1/8 inch bit. I clamped a ruler down to the foam as a straight edge. This made for some very accurate and clean cuts. The engine mount structure and the uprights for the rudders are all glued into routed grooves.

The servos sit in routed cavities and are firmly screwed into balsa blocks which are glued down to the foam. I free handed the holes for the rest of the RC gear. The routed cavities for the RC gear worked out real well. They can be easily removed when necessary but don't slide around or come loose.

For extra stiffness, you can use balsa strips glued into routed cavities It wasn't necessary on this craft, but I did this on a future design to stiffen the compromised structure of the foam around the lift duct hole.

The engine mount is a simple 1/8 inch plywood stand which bolts directly to the engine. The fuel line simply connects underneath the engine to a connector on the carburetor. It may be a good idea cut out a hole in the back of the mount to better feed the air intake even though there is clearance as it is. The main engine mount is supported by the four balsa braces, which are also mounted in routed slots in the foam.

The rudders are connected to 1/8 inch by one inch wide balsa uprights. The top section is just glued with a butt joint. As a future improvement, this structure could be stronger and the top section could be better connected to the uprights but it is still strong enough as it is. The lift deflector is attached to two pieces of 1/16 thick balsa glued directly to the foam on the top surface, which is visible, and the inner surface of the lift duct.

All of the movable surfaces where attached with the same technique: fabric hinges. I cut a basic fabric into one by two inch strips. Four of these strips per surface were plenty on my craft. I sanded the leading edges of the surfaces to a sharp edge so the fabric could cross over the edge cleanly. I used glue (Weldbond or any heavy craft glue would work) to attach the fabric to the control surface, saturating the material similar to the way fiber glass is used and alternating sides of the control surface as shown in the pictures. Once that dried, I could attach the other half of the fabric to the uprights and lift duct respectively in the same way.

The control surfaces attached to the servos with simple metal connecting rods made form a metal wire that comes covered in plastic. I'm not sure what it's for, but I think it is used for training plants. Z bends at either end of the rods connect to the plastic control horns. To actuate the second rudder, I drilled small holes in the top of the rudders and used more wire bent in a c shape to connect them. This works, but they had to be secured with rubber-bands when the connecting rod popped out of place. This made the rod dig into the top of the rudder. It would be much more secure to go ahead and use two more of the control horns and a connecting rod with z bends to link the two rudders.

The current set-up includes a piece of foam which holds the tank up high enough for the engine to be able to easily suck in fuel (around or above the height of the carburetor.) The foam kept coming loose in the beginning, but was fixed with a tighter fitting piece. It can't be glued in since the tank and foam need to be removable for filling the tank and priming the fuel lines, which can already both be difficult and frustrating tasks. Unfortunately, the tank soon loosened up and was moving around and falling out of the foam support, so it is currently held in place with rubber bands lashed around the tank and foam.

Instead of this design, I would suggest gluing balsa braces on either side of the foam so the stand can securely slot into the body without using glue. That would keep the tank upright and secure without making it any more difficult to work with. A rubber-band would still be needed to hold the tank from falling out, but the stand would be much more stable.

The skirt made of four pieces of fabric sewn together at the edges so the edge that touches the ground is smaller than the edge which fastens to the craft in order to catch air efficiently. The top of the skirt is pined onto the sides of the foam body with regular straight sewing pins. This holds the skirt on well with little air loss, but could be more secure, as pins often come loose and become ineffective. The picture at left was taken after running the craft on a gravel covered parking lot.

This open-bottom skirt design is very simple and effective at holding air while creating little friction, but it does have some issues. When running on surfaces with sand and other loose debris, it tends to scoop up said debris and weigh down the rear end of the craft. Running on rough surfaces also tends to wear and fray the skirt. A tube like design with holes on the inner wall would help to solve this problem and would still be easy to attach to the foam, but I haven't tried this design yet.

Hovercrafts seem to be relatively resistant to a mispositioned center of gravity. It is still important to check that the craft is level while the engine is running. Air running through the lift duct at the rear of the craft and the engines torque can both affect the balance enough to change the way the craft handles. It's a good idea to test the running balance before permanently securing the radio equipment. Since the engine and servos were positioned in relation to the rudder and lift duct fans, the positions of the receiver, batteries and fuel tank were the main variables. I started by positioning the fuel tank as close to the center of the craft as possible so it wouldn't change the center of gravity as the fuel is used. The rest of the RC equipment was positioned to place the center of gravity in the correct place: about an inch towards the tail end of the center of the body to counter the thrust of the lift duct and about 1/4 inch to the left side to counter the torque of the motor. I then tested out the running center of balance by taping everything on for a test run. After a few tries I was ready to route out the cavities for the RC equipment.

After some tweaking and testing, I got the hovercraft working consistently, although it could still use some further improvements such as a better skirt design and a better fuel tank stand set-up. Its biggest drawback is the difficulty of cold starting the engine. Once running the craft performs quite admirably. Acceleration is decent while the speed tops out at about 15 mph. It can also do standing 360s and takes high-speed, sliding turns with ease.