First ROV Project - Construction Starting Soon

I am new to the ROV scene, but have always been fascinated with them. So I have designed one in CAD and have actively been sourcing parts and equipment. I have some programming experience (C, C++, VC++, VB, LISP, and AutoLISP). I purchased (2) Arduino Mega 2560’s, 8- Arduino UNO’s, Manuals, and various tubing, cameras, etc. I plan to form the hull out of fiberglass with a 1/2" thick acrylic camera window in the bow. I have two 12vdc solenoid valves for the ballast tank control. The air tank will be 7" dia. steel pipe, 10 inches long. 2- 18ah agm 12vdc motorcycle batteries for power. The tether is 1/16" stainless steel braided wire ~150’ long and 2- cat6 data cables to controls on the surface. I will be using a Logitech Wingman Pro 3D extreme joystick for ROV controls. The overall size of the ROV is 40"L x 14"W x 10"H excluding thrusters and vertical control surfaces(rudders). Max dive depth will most likely be under 120’ as it is intended for freshwater inland lake use.


1 Like

First thought in my mind when I looked at the picture was of an old Typhoon class Russian submarine :slight_smile:

Question about the ballast system … are you using it to just sink and rise to the surface or what? Was curious as to your design configuration for that system.

 

Yes and no. The ballast system is basically to trim the boat out. It is a pretty simple system consisting of the air tank (80-120 psi) with dual 12vdc solenoid valves, a check valve and four ballast tanks utilizing modified 3 inch dia. bicycle tires. The ballast tanks are about 9.5 inches long, 3 inches in dia and made of pvc pipe sealed on the ends. The fist solenoid charges the tanks for ascension, and the second discharges for decent. The check valve keeps outside water from back filling. I am not sure how many cycles the air tank will provide at this point. The ability to dive to depth using the ballast instead of relying solely on the thrusters was the motivation. I would prefer to save as much power for navigating at depth if a longer search is required. There is also a bow mount transducer planned for this project as well(more energy consumption). I have attached an image of the system as planned. And yes, it does have the shape of the russian typhoon class, but somewhat modified. My previous design was based upon an 8 inch diameter schedule 80 pvc. Not enough room without making the ROV long and skinny. Would handle more like a carrier than an ROV :).


If you are using bicycle tires as “air bladders” why are you worried about check valves in your four ballast tanks? Frankly the lift is going to be based upon those being inflated or deflated. Also, are you planning on some sort of list and trim system since you have more than one ballast tank? I would assume that you could allow the unit to do its’ initial dive to depth and then you could zero the bubble by putting a little air here and there as needed.

One final question. When the system is totally devoid of air (the bike tires) or at least as much as they will deflate, are you going to be negative by a large amount or how much were you planning on? I am asking that in case you considered using them as trim tanks too and maybe had a fifth one that would contribute more to depth control by itself than the four if you are trimming the system.

 

There is only one check valve, and it’s sole purpose is to keep the water out of the solenoids. I am assuming that the four ballast will inflate/deflate at near equal amounts as water pressure(from outside) and air pressure(from the air tank) will be the same for all four. As far as the ballast being devoid of air, I am not yet positive on buoyancy. Once all systems are installed and watertight, I plan to adjust buoyancy from that point to have just slightly negative buoyancy with ballast empty of air. At this point, I haven’t designed separate controls for each ballast.


Update: I have decided to scrap using steel pipe or PVC for an air tank. Instead, I will fashion an air tank out of fiberglass to the specs I require using a combination of fiberglass woven cloth, fiberglass mat and #90 Isopthalic Polyester Resin. 3/8" thick side and end walls with additional thickness around the valve and feed line. It will be more than adequate for the 100 psi I need for air supply. As a bonus, it will reduce construction cost for the air tank as well as a small reduction in weight. The 1/4" stainless steel tube arrived today, as well as an additional Arduino Mega 2560 board, Ethernet Shield W5100 Expansion Network Board, Mini Ethernet 10/100Mbps Network Switch/Hub-5 port, 2- 3000 Lumen Cree T-6 lights, USB 2.0 Arduino Shield, and 2 bread boards. I am still awaiting the Arduino Uno’s (3), and the 8 - L293D Motor Drive Shield Expansion Boards. I plan to purchase 4 12vdc brushless motors and ESC’s to use for programming the Mega and Logitech Wingman Joystick. Once I am satisfied the software works correctly, I will then purchase 4- T100 Thrusters for the final install. I plan to use a laptop topside with the joystick to control the ROV. The use of an ethernet switch/hub will hopefully eliminate the need for additional Cat6e lan cables to the ROV. Additionally, I am researching the possibility for integrating the sonar(fish finder) into the software to eliminate the added equipment(fish finder screen). I would love to have a picture-in-picture screen with sonar display in a smaller frame with video feed as my main image. I will admit, it is a bit ambitious with my rusty programming skills.

Here is a picture of the ballast air controls preliminary assembly. I still have to teflon tape all the fittings and tighten.


Update: Fabricated a hot wire foam cutter and cut the pieces for the main hull. I have 5 pieces of 2 inch thick styrofoam, 14 inches wide and 42" long.They are stacked, glued and drying. Received the (6) 1-1/2" aluminum channels which will support the thruster and rudder shaft’s and bearings. I have the shaft holes positioned. I will probably use flashing tin for the end patterns as guides to shape the hull for the hot wire foam cutter.

The tin flashing worked, the main hull body mold is nearly complete. I have to purchase a 1/4" wooden dowel rod to use for the bow and stern. I will shape 14ga nichrome wire to the correct shape and attach it to the dowel which passes through the body. Once in place, I will adjust voltage until 14 ga wire is hot enough to spin in a 180 degree arc to produce the shape of the bow, then repeat the process for the stern. Then I will place the fittings accordingly, and begin fiberglassing to a thickness of ~3/8". I will alternate layers between glass weave and glass mat.

Charles,

I’m enjoying following along on this! I’d love to see some pictures!

-Rusty

Thank you Rusty. I finished the rough cutting on the main hull today. I used 14ga nichrome wire on a 1/4" wooden dowel rod to shape the forward and aft ends of the styrofoam mold. It is a bit rough, but nothing that a bit of plaster won’t smooth out. I measured in seven inches from the ends and drilled 1/4" holes from top to bottom for the dowel rod to fit through. then attached the 14ga nichrome wire in the shape of the hull. Turned on the transformer and adjusted the heat, and spun a 180 degree arc to form the bow and stern. Bear in mind, this mold will be destroyed upon removing it once the fiberglass hull is formed over it. It is critical however to get the shaft supports aligned correctly before glass the hull. I am using duct tape to as a barrier to keep the resin from reacting with the styrofoam(hopefully). As I mentioned previously, I have opted to use a stronger resin than standard fiberglass resin. I needed the stronger resin for the air tank, so opted to use the same resin throughout the build. It is nearly double the price, but well worth the safety factor. The shaft supports are 9/32" OD, and 1/4" ID. There is a .003" variance between the stainless shaft and the shaft support. I am hoping thew combination of marine grade grease and nylon bushings on either end of the shaft support will create enough of a water tight seal to prevent leaking.


Casting a concrete copy of the bow to use as a form for shaping the 1/2" acrylic bow camera window using Henry self leveling mortar. The coffee container is there to increase volume without mixing additional mortar. This will give the acrylic the general shape of the bow. It will be cut and placed on the bow form and fiber glassed into place. At least the theory sounds good on paper, we shall see. I intend to cut a groove on the edge of the acrylic to further aid in binding the window to the fiber glass hull.



Charles have you considered putting a mounting ring on the front (embedded in the fiber glass) and then bolting the window to that with an O-ring surface between? This would allow you to remove the window etc. for repair / internal access.

 

@Harold, that is an idea worth looking into. I have designed in a large enough access hatch to give me room to access most components inside. The access hatch on the top will be built somewhat like that. I will use a rockwell sonic crafter to cut the hatch from the hull. Then using a dremel grinder bit with a metal spacer set to 3/8" I will grind the inside of the hull to a uniform thickness to mount the aluminum bar stock. I plan to drill out the holes for the bolts through the hatch and bar stock. I might have to use some short pieces of 9/32" aluminum tubing flared as an insert so repeated tightening of the hatch plate screws doesn’t deform the fiber glass hatch holes. Then use 1/4" thick x 2" wide aluminum bar stock to build a hatch plate seal from the inside. I am looking at using B-half fuel sealant to form a gasket for sealing. My biggest design goal was to have as few openings as possible, to eliminate any chance of leaking. I plan to use the same sealant around the inside of the camera window to strengthen the seal as well. I will have a lot of electronics on board, items which do not function well in contact with water. As a side note, I am a disabled vet, 9 years active service, having served in a special weapons unit in the Navy. A good friend of mine who is also involved somewhat in this project is a retired Air Force Mechanic (E-8) with 20 years of service (and experience) with F-16 aircraft. I am lucky to be able to bounce ideas off him if I am unsure as how to proceed on some issues. He of course, prefers objects that fly, where as I am more interested in things that involve water. Having said that, I appreciate and value any advice or ideas that are offered.

 

Update: Rough cut the foam for the air tank mold and started coating it with plaster. As soon as the plaster dries I can sand it smooth and seal it. I will get the fittings ready to set and the next step will be to apply the first coat of fiber glass and resin. At this point I am contemplating the use of carbon fiber instead of fiber glass. I will do a bit more research prior to deciding on which to use. I have attached a few photos of the roughed inner mold for the air tank. (I am very partial to the look of carbon fiber weave, thinking the hull would look great!)



Update: I built a cradle/stand to set the ROV in as I work on it and also to use for transporting in the future. I have been applying thin layers of plaster to the foam body to smooth out the rough areas from the hot wire foam cutter. I have also ordered the 2" x 1/4" aluminum bar stock to be used for the hatch plate. In addition I ordered 1/4" bore double sealed flange bearings and collars for the thruster shafts. I will be ordering some 9/32" stainless tube to use instead of the aluminum tube. This will provide more strength and a closer tolerance. The aluminum tube had a tolerance of 3/1000th inch whereas the stainless tube has a 1/1000th inch tolerance. It will make sealing the thruster shafts much easier and less likely to leak, while giving more strength as well. They will only be turning a maximum of 180 degrees, so heat and friction will be negligent. With the addition of marine grade grease I should be able to get a very good seal. The 3- L293D Motor Drive Shield Expansion Boards have arrived as well. Each is capable of controlling 4 motors, or 2 servos or a stepper motor. I believe I can get away with using 2 as the fwd and aft thruster shafts will rotate in unison. Then I can keep one on hand as a spare.


Update: I have installed 3 steel shafts(temporary) and the shaft support brackets after having shaped the hull mold. I then applied the first sealer coat of enamel and will wait a day or two to apply a second coat. Tolerance on the shafts is within +/- 1/32" of center so I am happy. I am looking at triaxial carbon fiber for the hull construction. I have decided to mount the light housings on the outside and glass them in place. The transducer will be located between them on the underside of the bow.




Update: I have decided to use 1/2" dia. stainless steel tubes for the thruster shaft housings vice 9/32:. The inside shafts(thruster shafts) will now be 7/16" dia. stainless steel tubes vice 1/4". This will give me added strength for thruster support. The tolerance between the inside of the 1/2" stainless steel shafts and the outside diameter of the 7/16" stainless steel shaft is 0.0065". This is a tolerance I can work with to water proof the shafts. The 1/2" precision ground shafts have arrived and the 7/16" shafts are ordered and on their way. I also have an Adafruit 16-channel 12-bit PWM/Servo Driver-I2C Interface that I will be using instead of the L293D Motor Drive Shield Expansion Board for Servo control. The 1/2" diameter set screw collars I ordered aren’t actually 1/2" ID. I had to ream them about 1/1000th of an inch to make them fit the precision ground 1/2" diameter outer shafts. Not a big issue, just the difference between U.S. tolerance and foreign. When the 7/16" shafts arrive I can cut and set this all in place and get ready for the carbon fiber hull construction. I have also cut and shaped the clear acrylic camera view port for the bow. I used the concrete mold to shape the acrylic plexiglass after heating the plexiglass in the oven at 340 degrees for a period of ~24 minutes. I also placed the concrete mold on top of my wood stove for 45 minutes prior to using it. Fingers crossed the minor distortion from forming the plexiglass will not be to noticeable when submerged. I do not have the facilities to regrind the plexiglass to correct for any optical distortion at this time. I also used a dremel tool to cut two parallel groves on the edge of the plexiglass to aid in sealing to the carbon fiber hull. I ordered 4- 3650 4370KV 4P Sensorless Brushless Motor with 45A Brushless ESC to aid in programming the logitech joystick and Arduino Mega w/L293D motor controllers. The motors won’t be used on this project and are simply to be used for programming. I also ordered 1-Robotis Dynamixel AX-12A<b> </b>servo. If this servo proves to be strong enough for thruster shaft control, I will order an additional 2-AV-12A servos. I am still looking for beveled gears to use on the shafts, but if I can’t find something that will work, I will opt for gear and chain drive instead. I need to be able to pivot the shafts(fwd & aft) a full 180 degrees. I want 90 degrees up and 90 degrees down for maneuverability. Both shaft servos will be in sync. In addition, I want to be able to spin in place(YAW control) by using one forward thruster and an opposite aft thruster. The rudder will be used for directional control while the ROV is in motion(forward and reverse).