Keith Mitchell's

Scratch Built Radio Controlled

Disney Nautilus


Click any image to enlarge


 This is a front on view of my Nautilus. It is 52 inches long, about 1/41st scale. I think it is about half done at present. The weird pink and blue colors are due to a thin layer of automotive "bondo" applied over the fiberglass to smooth out the curves. Construction materials are mostly fiberglass, bondo, brass, and scrap printed circuit board (the green stuff) with the copper stripped off.
2. This is an oblique view of the model. I expect this model to be around for a few hundred years so I built it with maintenance in mind. It comes completely apart easily so the forward arch, fore and aft chines, superstructure, pressure hull panel, pilot house, parlors and internal technical stuff are all easily removable. All attachments are made with hex socket head stainless hardware into brass threaded inserts recessed into the hull. The "made in china" sticker is my grand daughters' idea, I like it, and it may stay there.
3. The pressure hull with some of top end parts removed. The hole is to accommodate the pilot house base which had to be deepened to install the lower deck and stair well.
4. This view shows the pressure hull lid, the forward raker arch, and the pilot house removed. When complete the pilot house will be sealed and will have internal lighting so the controls, gauges, the machine on the wall that goes tick….tick….tick, railings and stair well will be visible. The forward plastic blisters distort and limit the internal view so I am not going to spend a lot of time on details no one will be able to see.
5. I am way too old to wrestle around a 60 pound model so this one has a flooding hull. That should require about 25 pounds to sink it to the water line. I have installed an internal "dry package" which will house the technical bits and will be sealed against water. This is a view of the internal cavity and you can see a big gray knob looking thing on the aft (right) side of the package. It houses a thrust bearing which transfers the longitudinal thrust from the propeller to the hull frame and keeps it out of the propulsion unit. There is a square shaft which extends from the propeller shaft into the package which will couple to the propulsion unit.
6. Modular construction has always impressed me so I carried it to my Nautilus. This picture shows the three main modules in the package. The one on the left houses the electronics, servos and a water cooled heat exchanger for the power components. The middle module houses three 6 volt 4 amp hour batteries. I am not sure how they will be configured yet because I have not finished the speed control. The final module is my propulsion unit explained in the photo number 8.
7. This is a view of the package with the modules installed. A water tight cover will keep them dry.
8. This Nautilus is a surface ship that I want to be able to dynamically dive and skim with the forward blisters just at the waters surface. To maintain scale, my nautilus required a huge (by model ship standards) almost 3 inch propeller. Spinning a propeller like with enough energy to make it scoot that will require a whole bunch of torque. I ended up with two huge 5 pole motors that are coupled to a single drive shaft with a chain driven 4:1 gear reduction transmission. All shafts rotate in ball bearing mounts. It is very smooth and extremely powerful. A full power bathtub test caused a dreadful mess as the water left the tub, but it sure does push water. It should be able to give me the effect I want.
9. This view of the propulsion module shows the square socket at its center. The propulsion module is dropped into the front part of the cavity and slides aft over the square shaft to couple to the propeller. The battery module is dropped into the forward section after that. To the right of the propulsion module is the water cooled heat sink. I spent an embarrassing amount of time worrying about the heat build up from a speed control unit in a sealed electronics package. I was so focused on removing the heat from the package I could not see the solution. I remember changing my thoughts slightly to "how do I get the heat out of a submarine" and it hit me. The mind is a strange thing. Using a windshield washer pump the heat sink can suck almost 600 watts of energy without raising its temperature more than a couple of degrees.
10. This shows the propulsion module slid halfway down the coupling shaft. Several people have asked me why I chose chain gearing. While I was experimenting with gears I realized that gear trains must be very precisely built and aligned to work at all; and not completely wear out the gears within a short time. Metal to metal gears; metal to plastic; or plastic to plastic; all combinations had the same problems. Gears must also transmit all of the power over a few percent (and a few teeth) of their driving surface and more power means bigger gears. Under stressful conditions teeth can be easily stripped from the gear and cause total failure. Unless the modeler uses expensive (and even more difficult to align) helical cut gears there is a “whining” sound that I do not like. It seems the greater the power, the louder the obnoxious whine.

Chain and sprocket drives are surprisingly easy to build and do not require a great amount of precision. Sprocket tooth ratios work just like in gear tooth ratios and are a lot easier to count. Like with atom bombs and hand grenades, “close” works just fine and the chains can be quite “Sloppy.” The slop in my chains is easily visible and as long as the chain on the take up side does not get in the way of the chain on the drive side, it simply does not matter. Adding “idler’ sprockets allow chain drives to transmit energy over a considerable distance, to multiple driven elements, and even around corners. The power is distributed over half of the sprocket drive surface so it eliminates most wear and the potential for failure. Besides all of that; when it is running it makes a quiet hissing sound and no whine at all.
11. This is a picture of the rack gears pushed out of the trailing edge of the tail fin to demonstrate their location. When the rudder is set the gears are entirely internal and are not visible except for two small square holes in the back of the tailfin. The gears and cables are also easily removable by pulling them aft out of the races in the tailfin. You can see the lower bearing recessed into the bottom fin. For access the top bearing is removed, the rudder post is pulled vertically from the fin, and the rudder falls from between the tail fins.
12. This is an attempt to get a shot of the teeth down inside of the rudder post cavity. The cables are connected to two pieces of brass rack gears set in square brass races on either side of the rudder post and are also imbedded into the fin. The tiny pinion gear on top of the rudder post meshes with the racks and as the cables are worked in a push pull manner, the rudder turns easily and smoothly. The rudder post square shaft is sized so it passes between the teeth and slides through the square channel imbedded in the rudder. After the round lower portion of the rudder post engages the lower bearing the upper gear couples with these teeth.
13. Here is what the rudder post looks like before being fully inserted into the shaft to mesh with the gears. The lower portion of the shaft is long enough to already be seated in the lower bearing and the pinion gear is about an inch from engaging the rack gears. Sequencing the parts makes it a whole bunch easier to remove and return the rudder to its position and it only takes a few seconds either way. When assembled the rudder has no visible working parts. Dropping in the top bearing lines everything up and the rudder works surprisingly well.
14. About Rivets: After much dread and worry about how to create some 16 thousand individual rivets, I decided not to because I do not have the skills or patience. I performed some experiments and ended up making a set of tools that emboss crisp and realistic .050 inch hemispheres into .010 inch sheet styrene plastic on .125 inch intervals. That scales up to 2 inch rivets on 5 inch centers which I think is accurate for the model. My idea is to make the individual plates and emboss the rivets, fill the back side with epoxy and cement them into the proper position on the hull surface.

This picture shows the tools. The thing on the right is a die with a .050 inch hemispherical depression in the center. The thing in the middle is a pin with a .030 inch hemispherical ball on the end. The thing on the left holds them in alignment. I also have another unshown gizmo that holds them all together when I am pressing the plastic. I plan to produce another set of tools to make the smaller (one scale inch (.025 inches) I think) rivets too.
15. It took less than two minutes to produce this first test plate. I expect it should cut my rivet making time by over 90%. A year or so ago I contacted Lee Seiler and the resulting discussions and a little research convinced me that mid 18th century iron ship building technology would produce slightly irregular rivet patterns with some buckled and occasionally overlapped plates. This method should allow me to reproduce just that image on my Nautilus. I am attempting to produce a Nautilus that looks like a frequently used and war scarred veteran submarine that has been roaming the seas tearing the keels out of many sailing ships from that evil nation with no name. We shall see how it goes.
16. My tail fin is slightly less than half an inch thick and figuring out how to make the rudder work without an external pushrod just about drove me nuts. I ended up with a "push pull" set of control cables that loop around under the dry package; aft up through the waist and imbedded into the tail fin. You can see them hanging near the forward end of the cavity photos. The cables are connected to two pieces of brass rack gears on either side of the rudder post and are also imbedded into the fin. A tiny pinion gear on top of the rudder post meshes with the racks and as the cables are worked in a push pull manner, the rudder turns smoothly. The photos show the component rudder parts. There is a rudder post with a tiny pinion gear attached to the top, the top bearing and of course, the rudder.
17: The information posted on the DNC site from Theo and Lee was extremely helpful with my wheelhouse detailing. I am going to illuminate the interior with LEDs hidden inside the wheelhouse. I noticed that the large forward blisters distort the view and limit what can be seen so I attempted to show the larger controls and instrumentation that is closer to the front of the wheelhouse.

It is still incomplete but this view shows the starboard side of the wheelhouse and the double wall construction necessary to develop the thickness and details on the inside of the forward blister port holes. Most of my construction is raw brass because it is easy to work with. It will be painted where appropriate. If you look close you can see the top two steps in the stairwell. I do not think they are visible from the outside, but what the heck….
18. This is a view of the port side. It is also incomplete but reveals the two level nature of the wheelhouse deck. The penny gives you some idea of the scale. The ships wheel is not accurate and came from a now deceased and much missed family member, so it has sentimental value and will remain at the helm to honor him.

The curved interior roof that has been discussed on the DNC site was installed and became an advantage because it is hides the wiring to the crocodile eyes and interior LEDs.
19. I included this perspective to show the scale of the deck when compared with the rest of the ship. For laughs I had a clear teardrop shaped jet aircraft "bubble" canopy tacked down over the wheelhouse. It was sort of a "Nautilus as built by Boeing" version. It looked kind of neat, was totally impractical considering it could not survive ramming ships. On the other hand, can you imagine the view.
20. I tried to get a view of the inside wheelhouse from the outside and was not very successful. The blisters still need to be polished and the reflections from the exterior lighting interfere. When the wheelhouse is internally illuminated I think most of the internal stuff will be easily visible.
21. A flash helped penetrate the blister to some extent.


About this Nautilus

Sometime around 1996 I decided to build a model of my favorite submarine, and I doubt if I really need to explain "why" to any members of the DNC group. I have been a model aircraft builder since I was a kid, and had not built a ship before I had this idea. Experienced modelers will probably notice that this Nautilus is built like an airplane and not a ship because that is all I knew how to do. I have since discovered that model ship builders do things very differently. I suppose that a model is a model; and if the end result is recognizable as the Disney Nautilus, it does not really matter how one gets there. It took me a couple of part time years to research the Nautilus and a couple more to produce my plans before construction began in the summer of 2000.

When I began my research there was not much detailed technical information about the Disney Nautilus on the internet at that time. I could not find a kit for sale so I decided to just scratch build one. Luckily I also discovered Bjorn Lundbergs' website which had downloadable AutoCAD drawings of the Nautilus. One of the features of the AutoCAD system is the ability to scale a file full of drawings to whatever size is desirable, and all of the dimensions of the other files component parts in the program are automatically scaled proportionally. I have since documented my entire Nautilus on AutoCAD.

AutoCAD is also a neat tool for making parts. After I design a part I print it out full size on a piece of paper and it is accurate to four decimal places. I use rubber cement to temporarily glue the drawing to the material I am making the part from, and cut/grind/drill and or shape it. Everything always fits accurately the first time.

I have been in communication with your DNC moderator, Theo Wilhelm. I included a few pictures of my wheelhouse detail and thanked him for the information he and Lee Seiler were providing. He recommended that I submit more photos of the model to the members "as is" even though it was not complete. I was planning to wait until it was further along but he talked me into showing it as it evolves. He thought the DNC members would be interested in seeing its history and watching its progress through additions to his web page as I stumble through the rest of the building process.

Some thoughts on "Scratch Building" a model and why not to: This Nautilus will probably be my last scratch built model. While it has been an interesting and rewarding learning experience it has also been a constant problem solving exercise and has required an unbelievable amount of time, thought and effort. I have always been a bit of a gearhead or gadgeteer, but I am not a professional modeler, machinist or materials scientist. It sure would have helped if I were. The parts and components you see on this DNC web page are ones that finally worked. What you do not see are the many earlier attempts that did not. If you like the propeller or the water cooled heatsink you would probably also like the three previous ones in each case that became nothing but impressive paper weights. Figuring out how to make a small brass tube bend smoothly around a radius without kinking or flattening took many frustrating days.

After many unsuccessful attempts to create a mold for casting multiple parts like the raker teeth, hatches and others I discovered silicon rubber. Silicon rubber is a very thick goop that requires the addition of a catalyst to make it cure. Mixing in the catalyst creates air bubbles in the rubber. If left in the rubber, the air bubbles show up as lumps on the castings and ruin them. Soooo… before I could cast the teeth, hatches and other parts in polyurethane casting resin from silicon rubber molds, I needed to build a vacuum pump out of an old automobile air conditioner compressor and a washing machine motor. It was the only way to suck the air bubbles from the silicon rubber molds and achieve accurate looking parts.

The point I am trying to make is that figuring out how to make a tool, and then actually producing the working tool, to make a particular part or assembly; takes as much time and effort as making the model itself. (I hope that makes sense) Documenting each finished part or assembly on AutoCAD adds even more time to the project. I have been working on this model for over 5 years and it is still only about half done. My wife of 39 years says I am a little crazy and extremely stubborn, I prefer to think of it as virtuously persistent.

The only reason I built this model like I did was because no one produced a kit or even a set of plans at the time I caught the Nautilus bug, and I did not know how to make one. I needed to figure out everything by myself. Whether you are interested in a static Nautilus or RC version, the kits offered by some of the DNC website members are truly superb and if they were available when I was looking for a kit, I would have bought one of them in a heartbeat. DNC group artists have performed an extraordinary amount of research; committed at least an equal amount of very meticulous work and have solved all of the problems. They have created the tools and ironed out all of the kinks, and it is absolutely the best way to go for anyone that wants an accurate model of the Disney Nautilus. It is amazing to me those DNC members that sell their kits or finished models are able to produce their fantastic models for the low amount they charge, and please trust me on this one; they are worth every nickel of their cost - and then some.


The Second Half

One of the more enjoyable aspects of "scratch building" models is that there are no rules, standards or dictates that control the model. All modelers claim some amount of artistic license, and can do anything to their model they wish. My Nautilus would have a set of tank tracks on the keel and the ability to crawl right out of the water if I felt like it. The purists might have problems with it but in the end the modeler is the only person that needs to be satisfied.

Several years into my Nautilus construction I discovered and became a member of the DNC group. I was truly impressed with the diversity and quality of Disney Nautilus membership knowledge, and discovered many errors in my model. This awesome pool of talent, technical information and intelligent speculation has helped me correct some technical errors and I can not thank this group enough. There are times when all I can do is keep up with the information flowing in from the DNC group and document it for further investigation to decide if I wish to incorporate it into my model. Construction on the model takes a second seat when that happens, and I am sure that I am not the only one out there grinding fiberglass and spinning brass that has the same wonderful problem.

I have many folders full of notes, sketches and incomplete ideas I may or may not include in the finished model. This Nautilus will continually evolve until the project is completed and I honestly do not yet know what the final result will be, but at least I know what it will look like. The following paragraphs contain some of my ideas and thoughts at present. They will most certainly change as I encounter unsolvable problems; think of better ideas; or due to the DNC Group - come up with entirely new ones. Here are some of the features on my Nautilus that need to be finished, developed or added as the model progresses from present. This Nautilus may be one of those rare models that are never completely "finished."

Shaping / Sanding / Rivets / Painting / Finishing

The dorsal fin, hatches, and other tacked down elements must be properly aligned and permanently or hardware attached to the hull. The many various sized blisters need to be polished out and bonded to the Parlor frame, crocodile eyes and forward ports. Two irises need to be made from brass shim material. My Nautilus still needs a lot of final shaping and sanding before the application of the riveted plates, scoops, anchors, external ladders deck texture, all of those little thingies on the keel and other final surface detail.

I have discovered that the Disney Nautilus is an unusual ship to model because it has a tremendous amount of surface detail visible only to those who have made a study of it. The more you study the Disney Nautilus, the more you will discover. In the same manner that General George Custer found himself in a "target rich" environment; the Disney Nautilus becomes a "detail rich" surface and it can become a trap for a detail oriented modeler.

Power and Speed Control

I am using three 6 volt 4 amp hour lead acid batteries. I might end up using them in parallel as a 6 volt 12 amp hour system, or a series 18 volt 4 amp hour system. I am not sure at this point. I only need enough energy to run the Nautilus for about 30 minutes at a time. I will carry extra batteries or make some sort of charger operating from my truck battery to recharge them if I wish to operate the Nautilus longer than that.

When applying 6 volts of power to a single motor and gradually increasing my grip on the shaft, the motor turns with enough friction energy to make me release the shaft when the current reaches about 5 amperes. I do not think the propeller load when cruising will be anywhere near that much so I expect about a 2 amp draw for each motor at flank speed. My 4 MOSFETS are good to 28 amps each so there will be no problem there. The "H" drive circuitry for the MOSFETS was fairly simple and easy to build from junk components. I have not decided whether to spend the time to finish developing my own reversible proportional control, or just purchase a commercial one and bond it to my water cooled heatsink. I have developed several versions, but so far I am unable to create a successful decoder device between the power components and the receiver.

My problem is that I want a servo-less speed control, which plugs directly into a receiver socket. That requires programmable or proprietary integrated circuits which demodulate the receiver servo signal to produce a signal that will control drive circuitry speed and direction variables. I have been toying with the idea of taking an old servo apart and experimenting with the electronic portion. Maybe it produces a signal I can integrate with my own power drive circuitry. I have not tried it yet so I do not know if it is possible or not.

Auxiliary Dive Planes

The Nautilus dive plane system is about the only system on the Disney Nautilus that I do not think would be effective on a "for real" submarine. However, considering awesome ability of the Disney wizards to transfer 1950s nuclear submarine technology to an 1860 time period and make it completely believable, who could be the least bit critical.

I plan to use two external detachable dive planes made of clear plastic that will plug into servo driven sockets. They will be recessed into the bottom side of both forward chines between the forward dive planes and trim planes. When resting at the water line, the Nautilus floats like an ice cube with about 95% of the ships mass already under water. I think a powerful propulsion system and about 10 square inches of plane on each side of the Nautilus bow will be capable of driving it down into the water with enough force to keep its blisters at or just below the waters surface, maybe more. I have absolutely no experience with this sort of thing, it just seems to make logical sense that it should work that way. My nautilus will always maintain positive buoyancy so in the event of some system failure it will always end up floating on the surface for recovery.

Illumination

Any sci-fi inspired boy in 1950s USA (except DNC member Nestria) was certain of at least one absolute truth. Anything that glowed green always had three distinct properties: 1. It had mysterious, awesome and unlimited power. 2. It was created by a mad scientist; or better yet a brain sucking pack of Martians. 3. It was ATOMIC, and no one cared that no one knew what it meant. Even in the grade "B" black and white movies, anything that was glowing - you just knew it had to be glowing green.

As a 10 year old boy at the theater watching 20,000 Leagues with my Dad I still can remember suffering through Kirk Douglases' "Whale of a Tale" performance and sadly thinking that I had been suckered into sitting through yet another senseless '50s era musical. Suddenly: "On deck…floating object off the larboard quarter." It blew up a ship, it knocked Kirks' boring-singing-dancing ass into the sea, and it was glowing green. I experienced one of those unforgettable galvanic childhood moments and was instantly and forever hooked! Even today when I see that part of 20,000 Leagues, my minds eye has me still in that theater next to my Dad. He is still giving me his icy Marine Corps stare because of my spontaneous solitary applause when Kirk hit the water and I dropped my sack of popcorn. Back then people still applauded in movies and popcorn really came in a sack, not a platoon sized bucket.

I will completely illuminate my Nautilus with LEDs (Light Emitting Diodes). Under recommended operational conditions LEDs never burn out and put out enormous amounts of light with extremely little electrical consumption. LEDs convert electricity directly to light without wasting energy to make a metal filament hot to emit mostly heat and a little light. LEDs can be permanently "built in" to structures and there is no need for sockets, insulation, waterproofing, or access for replacing burnt out members. The evolution of LED technology over the last few years is truly impressive and the newest generation of LEDs operate somewhere around 20,000 (for some reason I seem to like that particular number) MCD (MilliCanDella). A millicandella is a term which expresses LED brightness. On many occasions I have used the internet to attempt to understand LED brightness values (candelas) compared with something I am familiar with; like incandescent light bulb values (watts/lumens). I usually end up with yelling at my monitor because of my ensuing cerebral hernia. It is obviously way too complicated for a mere mortal like me to understand so I gave up.

Out of desperation I bought a variety of LEDs to tinker with and to try and figure them out for myself. As a comparison, the LED power indicators on your monitors or DVD players are rated about 80 to 100 MCD in brightness. I played with some LEDs in the 5,000 to 8,000 MCD ranges; when I look at them held at arms length they hurt my eyes, project a big bright spot on a dark wall 6 feet away and no matter how long you run them they do not even begin to get warm. The instructions on the 20,000 MCD varieties tell you to not stare directly at them because they will flat out damage your eyes, and "bright" does not begin to describe them.

One curious property of the white LEDs is that they all seem to produce all of their light in a narrow 20 to 30 degree cones and operate as miniature spot lights. When viewed from anywhere out side of that cone there is almost no light. The pilot house interior needs to be flooded with light from a source similar to an incandescent bulb, and even multiple narrow beam spot lights still do not look correct. Diffusing the light absorbed too much of it and I still ended up with most of the light in a spot. It took an extra few weeks to figure out an optical trick to make a clear LED radiate in all directions like an incandescent bulb without any diffusion.

Model Builders Need A Little Fun Too

Other than the obvious lighting in the crocodile eyes, parlor blisters and wheelhouse, I plan on hiding two of the hyper-bright 20,000 MCD (unless someone develops even brighter ones by then) green LEDs in each of the 28 scuppers (notches) around the base of the superstructure and angled slightly down. Cruising anywhere up to half speed I plan to drive the scupper and crocodile eye diodes with a ¼ level of maximum intensity. After that; as the propeller speed increases to maximum turns, the intensity of those LEDs will also proportionally increase another 300% to maximum. As a semi-intelligent guess; all 56 green scupper LEDS, each operating at 20,000 MCD should present the equivalent level of several incandescent kilowatts worth of green light with a current draw of less than 1.5 amperes. The narrow 20-30 degree projection angle of the scupper LEDs should concentrate the light into a large shallow downward conical "fan" shape around the base of the superstructure and illuminate the water.

When seen from the surface the LEDs will not be directly visible, but the light reflected and diffused by the water should be. I am again guessing, but I think a 1.12 mega-MCD (1,120 candella?) moving patch of bright green light in the water; a dark deck and wheelhouse silhouette at the center and a big set of green glowing crocodile eyes on top should be an astonishing sight in otherwise dark water, especially to someone not expecting to see it.

During a cruise about dusk time, I believe that I could have some fun with people fishing at dusk at our local lake. But what's this…. that guy fishing from that rubber boat pretending to mind his own business….he dares to fly no flag……

Just a passing thought…, maybe I can even start some "Lake Monster" rumors.


About Me

I am a self employed semi retired telecom engineer living in Oregon.

My hobbies include optics, astronomy, electronics, modeling, aviation, radio and general technology. I am a private pilot (w/sailplane endorsement) and scuba diver.

US Navy 1965 to 1971

I grew up in Fallbrook California through the 1950s, San Diego in the early 1960s and was a serious Disneyland / 20,000 Leagues exhibit-a-holic. I returned from my military service and discovered that the Disney management had destroyed the 20,000 Leagues exhibit. Disneyland was never the same after that but at least I have my memories.


For more pictures click here.