Intro: Airboat ReDeux With Found Parts
Every generation makes an airboat --- it's a required right-of-passage for any Maker. Here's my oddball version.
This year I wanted to offer our local Maker Group something new, a boat that's powerful enough to cross the entire pond. I also wanted it to be RC controlled with no Arduino involved. Yes, I can do a project without Arduino.
I also wanted it to "look" different, something funky and a little organic. I sketched for a while and kept my eyes open for unusual items.
When someone donated a crashed quadcopter I decided to use that. When last year's Christmas décor hit deep-deep discount at the craft store I knew I had to make a semi-transparent "waterbug" style boat out of the clear bulbs. A trip to the $1 store provided the T-frame squeegee to connect the pontoons/decorations.
Boom, the design was done.
The whole build came together in a day or so once the design was set. This project is really easy to assemble and pretty cheap - four $1 items plus some glue and wire.
Step 1: Materials and Methods
- Old Quadcopter and Transmitter - I'm starting to get these in donations now, and this one had a working transmitter and flight/motor-control board. The frame was trashed, and it was a no-name clone so I didn't want to risk buying a crash-pack for an unknown product. I think this was meant to mimic an early Hubsan quad.
- Batteries - The donated quad had two tiny LiPo batteries with a custom connector. (Splicing in a standard JST connector system would allow use of a much larger battery) Luckily, the charging cable was included.
- Two "Flattened Pill" holiday decorations - There are several sizes of these clear plastic bulbs available at most craft stores. I used the 3"D x 1.5" T version
- One Candelabra holiday decorations - These are also available in a number of sizes and shapes. I used the 2" D x 4" L version
- Squeegee or similar T-frame - found this one ay a $1 store
- Stiff Wire - To make the motor mount. I used some aluminum "armature" wire and some scavenged copper
- E6000 Glue - the go-to, fixes nearly everything adhesive --- 'nuff said
- Silicone Caulk - to waterproof the openings of the holiday decorations
- Twist Ties - helps to hold pontoon/decorations while the E6000 dries
- Paint - maybe some plastic-compatible spray-paint if you want to paint the squeegee body.
- Pliers - small jewelry-maker style to bend the wire precisely
- Snips/Cutter - to cut wire and snip out the back of the quadcopter body
- Prep Board - cardboard to catch glue and caulk overflow
- Test Tub - I needed to test the design in my bathtub before I took anyone to the lake, you might not need to
BASIC DESIGN and ASSEMBLY
The entire assembly process took me less than half a day, but the design and research process took months. When I say design and research I mean dreaming, doodling and scanning store aisles and my scrap pile for something that would not only work but look cool. I came up with a lot of ideas and designs. And I spent 10x more time partially assembling failed designs made with balsa, test-tubes, Styrofoam hemispheres, cosmetic bottles etc. But when the right materials were selected, the entire build became super simple.
I'll give more details in other steps, but this is all that's involved:
- Seal the clear holiday decorations with caulk
- Glue the clear plastic bulbs to the ends of the squeegee, flat disks on the "back" and pointy end of the candelabra to the "front"
- Fix the quadcopter body if needed
- Cut a slot in the quad's body
- Put the flight/motor controller board in the slot
- Wrap some wire around the quadcopter and attach it to the back of the boat
- Carefully bend the wire motor-mount so that the flight/motor board is as level as possible when in the water
- Turn it on, sync and go....
Step 2: Assemble the Body
I had been working on designs for making a T-framed triangular trimaran already. I tested models out of various materials for months. I was failing every time. Every design was too complex for a single 3-hour build with young Makers. So when I saw this squeegee I decided to gamble the $1 on it.
I tested the basic fit when I got home and was very happy to see how well the holiday bulbs fit onto the squeegee and how the gentle curve of the handle helped give the pontoon bulbs an attractive raking angle. It looked like it was going to work. But first I had to seal the bulbs.
Caulk and Seal
Put caulk into the caps of the Holiday decorations. Make sure all holes are filled. Now press the caps back onto the clear plastic bulbs. The pointy candelabra had a large hole, so I cut a small piece of thin Styrofoam and caulked it in place over the hole. The caulk should squish out of the joint to show you have a good seal all around.
The caulk should not only seal the bulb/pontoons, but it should effectively glue the caps onto the bulb body.
Glue and Affix
Next, put a glob of E6000 anywhere the bulbs naturally touch the squeegee. For this squeegee it was at the rim of the decorative caps and one place on the clear part of the bulb. And I'm talking a large glob of adhesive. We don't want the adhesive to fail in the middle of the lake.
I used twist ties to make sure the bulbs stayed in place while the glue dried. In fact, I left them in place on the first open-water test run just to be safe. I might even leave them on the final edition (or something similar) if I can find a way to make them look a little more decorative and intentional.
Let Them Dry
Okay, so I started testing the body immediately, even putting it in the water and submerging it. You, however, are smarter than I and will let the caulk and glue cure overnight.
Step 3: Mod the Quad
If you have a quadcopter that you can sacrifice, (or $25 is okay to spend), just use a new quadcopter. But you will need to modify the body so much it will probably never fly correctly again (without totally replacing the body.)
The donated quadcopter that I used was completely trashed, but free - my kind of budget. Three arms were nearly broken off and the fourth was hurting. So I had to glue the arms back to the body as a first step.
Re-attach the Arms
I wanted to be able to get back inside the body so I unscrewed the back plate and separated the two halves of the body. Then I used E6000 to re-attach the arms on both body halves. The glue reacted with the plastic of the body and made it temporarily soft. I waited until the E6000 was no longer tacky, then press fitted the two body halves back together. When both body-parts were put back together the motor housings were all pointed in roughly the same direction.
While they dried I checked occasionally and made sure the two halves were not getting glued together.
The result was good enough for an airboat but would never work for a quadcopter. Once the glue dried the plastic became rigid and solid again.
Cut and Gut
I knew I would have to change the location and orientation of the flight controller (see below). I was afraid I would have to cut the wires to the motors and solder in extensions. I was okay with that because I intended to use JST connectors to make the motors transferable to a number of future projects. But when I tested, I found that, if I cut a small slot, I could leave the board connected to the motors while I rotated the board perpendicular to its original position.
I cut a slot in the back plate of the body and slid the controller board into the slot. The edges of the slot helped keep the board in place for early test.
When I knew the board was going to work correctly again, I used hot-glue to affix the controller board into the slot. I tried to place the board at right angles to the main copter body to make it easier to align when mounting to the boat frame.
But Why Change the Controller Location
Normally the quadcopter is oriented flat to the ground, but for this project we change the orientation to vertical. This means we have to change the location and orientation of the controller board. The flight controller senses when the quadcopter is tilted. The board then increases or decreases the power to the motors in order to bring the copter back into level flight.
If the board was tilted to the vertical, then two of the motors would cut out and not provide any thrust. If it is tilted front to back we would loose top or bottom motors. If tilted side to side we would loose the left or right pair. But if we move the controller board so that it's once again flat (parallel) to the ground, then all four motors will run at full power.
Buy a couple of "crash-packs" of spare parts for a cheap quadcopter. This should give you four motors, a two extra bodies and lots of extra propellers - but no motor/flight controller. We don't need a flight controller for this project. In fact, it just complicates things for no added value.
- Use the transmitter and receiver from an old RC car to control the motors. You will probably even get the ability to reverse the motors, but you will still (probably) only have two communication channels.
- Or use a tiny Arduino and driver board. You will probably have extra motor channels left over to control a rudder, lights etc. plus the ability to reverse power the motors. Choosing which communication protocol is up to you - Bluetooth for phone control, WiFi for remote, LoRa, XBee or regular RC.
Step 4: Mounting and Alignment
I used some thick aluminum wire for mounting the quadcopter onto the boat. I am familiar this type of wire as "armature wire" the kind clay sculptors use to make body-frames for their sculptures. Jewelry makers also use it make arm-cuffs, ear-cuffs, tiaras and bracelets. But you can use any stiff wire - and if it's insulated that might even be better.
You will also need to bend the motor mount so the flight/motor controller board is flat to the ground plane.
MAKING the MOTOR MOUNT
- Measure out enough wire to create a "U" that wraps around the body of the quadcopter, with a vertical length that keeps the copter blades out of the water and is also long enough to wrap around the rear crossbar.
- Create the motor-mount by wrapping the wire around the body of the quadcopter with the bend of the U at the top (nose) of the copter.
- I bent the loop back over the nose (now at the top) of the quad to keep that end secure.
- Then run the wire snuggly against the two sides of the copter body.
- Then bend the wire around both of the lower copter arms to keep the bottom secure. This might take a few tries, but keep at it.
- Make sure you leave enough wire to create a vertical stand that keeps the propellers from hitting the water or the body of the boat.
- Then bend the wire around the rear crossbar of the body, one loop on either side of the center bar to keep the motor-mount from sliding from side-to-side.
Your copter might be completely different, but the technique should work the same.
Create another loop of wire to hold the battery. I ran the loop through an opening in the body. You may notice (see photos) that this part failed a few minutes into the test.
LEVELING the FLIGHT CONTROLLER
You probably know that the flight/motor control board helps keep the quadcopter level. It senses when the copter tilts and increases or decreases power to the motors to bring the quadcopter back to level flight.
This presents a problem when we reconfigure the quadcopter for a vertical orientation. If we left the flight controller attached like normal, the controller would cut power to two of the propellers in an attempt to bring the copter back to level. So we would never get more than half of the available thrust. We have already moved and reoriented the board in a previous step. Now we have to subtly bend the motor mount to make the board as level as possible.
This is a precise, yet brute-force process:
- Bend the wire of the motor-mount to level the controller board, then test to see how flat it looks when the boat is floating.
- When you have it nearly level, test the motors by running them up to full speed with the left/right control centered.
- All four motors should start at the same time and run at the same speed.
- Continue fine tuning the levelness of the flight board until all the motors act in unison.
- When you accomplish this, test to make sure the motors respond correctly to the left right stick.
Step 5: Lessons Learned
Here are some observations about the process and results:
I spent more time going down blind alleys and false starts than I did on the final project. Everything was too complicated, expensive or required tools we don't have. All this time was not wasted though. When things came together the design, both visually and structurally, were far superior to the early tries.
Now I just need a way to convince young Makers that constant, repetitive failure is fun - we call it iterative design.
Ease of Assembly
A success. Very easy to assemble, but only after many failed designs
I wanted something unique. And this is, at least, an odd looking solution. But I had a similar concept for a 3D-printed version. Unfortunately, we don't have a 3D-printer available so I shelved the concept. But I think the early concept pointed me towards certain pre-made materials in order to achieve some basic design goals.
Now that it was successfully tested, I wish I had painted the main body before assembling. I wanted a sort of shimmery, bug-like appearance to match the dragonflies that inhabit the pond. Right now it looks more like a Klingon Harbor Patrol craft before battle-paint. And that's okay.
Everyone loves RC toys, but not everyone wants to create an Arduino communications platform. So I wanted to use an existing RC system with minimal modifications.
The quadcopter's flight controller caused issues. By re-orienting the board we solved many of the problems. But we must be loosing at least 25% efficiency.
However, that same board has motor controls and RC communication, so it made the rest of electronics tasks very simple. Overall, a winning trade-off.
Personally, I still wish I had an Arduino and a multi-channel motor controller for rudder controls and thrust reversing. But hey, that's what second versions are for..
It is faster than I anticipated. The motors are so much smaller than the toy DC motors we typically use on our homemade airboats, and the prop-wash did not feel significantly stronger, so I thought it would barely putter slowly across the pond.
But this little sucker gets up and moves. Except for the nose digging into the water (see the "stability" section below) there isn't much drag so it accelerates quickly and scoots along smooth water nicely. I anticipated it barely being able to overcome a gentle breeze, but it did. A strong wind could probably stop it though.
If I ever do an Arduino, brushless ESC style set-up then this configuration would scream.
Again, I was surprised at how well it turned. It does take a good 5-10 yards (m) to complete a 360, but that's better than I anticipated. It does skid a little, but not as much as I thought it would.
I do wish I had a rudder, either an air or water turning version. And I wish I could selectively reverse thrust. Both of these options would make the boat turn much more tightly. But I would need an Arduino and motor driver to do these things and that was not in the project specs.
Another option would be to splice in a longer power cable, then place the motors at the far ends of the back bar to improve turning capabilities. But I wanted to make this project something a kid in elementary school could accomplish with classroom tools.
I don't know enough about nautical design to predict whether having a keel fin would improve the steering.
I tested the design in the bathtub before going to the pond. It seemed stable, too stable for something that I designed with no references and threw together in a morning. So I was worried - it seemed too good to be true.
The back floats work well, but the nose dives and digs into the water when under power. The nose sits in the water correctly when at rest and I thought the sloping shape of the front pontoon-bulb would cause it to rise up "on plane" when under power. But I think the thrust from the propeller causes the boat to rotate-oppositely around the pivot formed by the back floats. I'm sure this submerged nose costs speed and battery life, but it sure does look cool - so I'm okay with it.
The boat also seems to have a low profile (cross section?) where wind is concerned. It gets pushed around a lot less than "water bottle" style projects, and maybe even less than Styrofoam and balsa versions. Even with no keel (the big fin under sail boats) it doesn't react much to breezes.
I want to 3D print a version where the T-frame plugs directly into the bulb and seals them. The T-frame emerges more smoothly from the shape of the clear bulbs. But those holiday bulbs, they are much cooler than the original design so they will be used in any future version.
The motor housing could be detachable and could be switched out to eventually hold a brushless motor and ESC.
An Arduino and good motor control board would provide rudder controls, reversible thrust and other movements.
Sensors, lights and additional actuators might be incorporated.
For a first try, I would give this project a strong A- with bonus points for fun. Hope you enjoyed it.
Third Prize in the