UNDERWATER QUADCOPTER

[Raptortech]

I'm in a robotics class, and my team has a $1000 budget to build something awesome. Half of us want an ROV (submarine) and half of us want something that flies. The obvious compromise is a machine that can do both.

After some discussion and research, it appears that the best way to go about this would be a multirotor. This because it is already heavier than water, and has all the mechanics it would need to maneuver both underwater and in the air.

The basic idea for this multirotor (probably a quadcopter) is that it would land gently on the water and sink below the surface. Once submerged, the pilot can turn on the rotors again (at a significantly reduced power), and control the craft. To takeoff, its as simple as ascending to the surface, and once all the rotors are clear of the water, powering up and lifting off.

We found a video of a guy doing just this in his sink with a waterproofed toy quadcopter:

Fortunately, brushless motors are waterproof (to an extent), and FT just released a video on waterproofing electronics! This craft is designed to remain underwater for more than just a moment, so we would likely make a water proof container for the flight controller and receiver. This will also help with buoyancy.

The main challenges for our group:
--Ordering parts because the school has specific websites that are available to order from due to some funding policies. We are working on getting dx.com approved.
--Waterproofing. It's never perfect.
--Radio range underwater is significantly reduced, so we may need a more powerful transmitter and improved antennas over stock 2.4GHz radios.
--We do not know how a quadcopter will handle underwater. It may be the case that the motors simply stall. This could require modification of the flight controller firmware to optimize handling, as well as special ESCs or under-propping the motors.

The reason I am posting this is because I would love to know if anyone has an ideas about how we should approach this project, or any experience with underwater RC. It's also kinda cool, so I thought you might be interested

 

[T-Richard]

If it were me I'd try to use a variable pitch quad with two different idle speeds one for underwater and one for air.... or perhaps limit servo travel in water mode to avoid stall and have it increased for air.

 

[Raptortech]

If it were me I'd try to use a variable pitch quad with two different idle speeds one for underwater and one for air.... or perhaps limit servo travel in water mode to avoid stall and have it increased for air.

That was one of the first ideas we started throwing around. Although the cost probably wouldn't be an issue, I would like to keep the quad as simple as is possible. Fewer parts means fewer things to break. Also, we would have to waterproof the four servos and somehow protect the pitch mechanism.

 

[Craftydan]

Most hobbyist subs I've seen are built neutrally bouyant with the water, then they use a servo actuated syringe connected to an external port to change the bouyancy from positive to negitive -- if you've got a failsafe, best to set that servo to "positively bouyant" . . . don't want to loose it at the bottom of a lake.

Either project done right can be a good sized project. Both together will benefit from a bit of engineering simplification. If I were to build one (yes, I've considered it) I'd lean toward a tube-hull H-quad with the electronics mounted inside, battery/service hatch at one end, a small, low power waterscrew on the tail on a 2-axis servo gimbal, and the air-motor booms and motors above the frame(so they're just above the water when it's floating). I can see Y-connecting the pitch/roll to the gimbal and the water screw to the throttle. Probably best to not connect bouyancy to rudder -- don't want to re-arm while trying to resurface.

Touch down, sink almost to the props, disarm the flight board and decrease your bouyancy to sink . . . Throttle back up (disarmed, so the waterscrew is going) and tootle around underwater . . . then float back up until the props break the surface, arm the flight board and throttle up to pop clear of the water. Your waterscrew will still work, but it'll move so little air, it shouldn't affect your flight.

For Comms, if you can get 72Mhz Gear . . . or even 35Mhz (harder to find, and you'll need a Ham license) it'll penetrate the water better.

Simpler layout than attempting to get the flight board to go easy on the throttle and hope your board can make sense of the new environment (if nothing else, your PIDs should be MUCH different).

 

[Raptortech]

Awesome! Tons of good ideas there. I like the H quad idea with the tube. Maybe something along the lines of a boxcopter would work. The buoyancy failsafe is probably a good idea. I was sorta liking the idea of using the same drive system for flight and for underwater navigation, so idk about the pusher water-screw. I guess I'll just have to do some underwater testing with a flight motor and prop and see how it goes. As for the radio, I am a HAM and already have a few 72MHz crystal radio systems. We will be testing in the school pool, so we'll try out a a couple of setups to see what gives us the best range and depth. I suppose the eventual plan is to use this in a lake or the ocean (though salt water is tough).

 

[BEEAH]

Water proof quad under water, takes back off!

 

[Cyberdactyl]

If I were to build one, I would just use a small tupperware container. There's no need to use giant containers as I have seen used. Then use about 3/4" I.D. x 1" long threaded tube with silicone washers on both sides, with all 12 motor wires siliconed in the 1" tube. Nothing else would penetrate the container. Keep it as isolated as possible. It wouldn't be pretty, but it would be quick, easy and very unlikely to leak.

 

[RAM]

Some good info in this video series.

 

[Raptortech]

Turns out we can order from amazon, so I have put together a shopping list on there. What I have in mind right now is:

Frame: 450 size
Motor: 1000KV
ESC: 30A SimonK
Prop: 10x4.5
Battery: 3000-4500mAh 3s 30C
FC: KK2.1

Each of the above items in the list is linked. I've never built a multirotor before (though I've got a couple hours flying my friend's). Does this parts list look acceptable? Any input is greatly appreciated!

At my instructor's insistence we will be buying twice as many of everything just in case we get duds or manage to fry them in the water. For this reason, I decided to buy fairly cheap stuff to keep the total price relatively low ($324 including wiring/connectors, batts, and chargers).