Thread Tools
Mar 12, 2011, 09:42 PM
Damp and Dizzy member
Brooks's Avatar
Thread OP

A rubber band powered Nautilus

My first rubber band submarine was built from a "Tetra" sub kit by Seaworthy Small Ships. After seeing how much fun one could have with a simple, dynamic-diving sub, I made a Disney-esk profile Nautilus, also rb powered - it worked fine. Neither sub could control it's depth - they would dive to the bottom, travel 30 feet or so, then surface as the rb wound down.

The sub reported here is a dynamic diver, also rb powered, but adds automatic depth control by manipulation of the dive vanes. The dive vanes have trim tabs set to force the vanes into the "dive" orientation. A line runs from the vanes to a floating surface object. As the sub dives and reaches design depth, the line tightens, pulling the dive vanes into a "surfacing" orientation (or, at least, reducing the dive angle of attack of the vanes, allowing the hull buoyancy to move the sub upwards). The constant battle between the trim tabs and the tug of the line on the vanes results in the sub proceeding at a stable depth. Design depth is set by adjusting the length of the line. Testing today was set at a 6" depth.

For testing, a simple foam float was used as the surface object. Now that the concept has been shown to work, a more Nautilus-themed surface object will be substituted - what could be more appropriate than a Giant Squid? Two squids are under construction. Spending the afternoon watching "toothpicks" (periscopes) traverse the pond, or watching a squid chase a Nautilus .... hmmm :-)

The dive vanes are located mid-hull, rather than at the bow and stern as with modern subs. This mid-hull placement is characteristic of subs from the mid 1800's thru early 1900's. Pierre-Yves Garcin has photos of 1904 rb subs, with mid-hull vanes, and reports that they work well. Jules Verne's 1870 Nautilus would have had mid-hull vanes, most likely. Various Verne Nautilus designs have been gathered by Michael Crisafulli.

The advantage of mid-hull vanes is that the hull remains level during dive and surfacing; this is good for a free-running sub as it reduces potential oscillations in depth. The disadvantage of mid-hull vanes is that they don't use the thrust of the prop to help the sub dive or surface. With modern subs, giving the boat an up or down bubble (ie making the hull point towards the surface or towards the bottom) speeds depth changes. With mid-hull vanes, however, we can't do that, making the forces available for our use in depth control much less.

This Nautilus has a Disney tail and a "League of Extraordinary Gentleman" (movie version) shaped hull - long, thin, and deep. The dive vanes are modeled after Humpback whale fins. As a "Nautilus sub", it has a chambered nautilus painted on the side. The yellow eye/salon window looks particularly fierce, to my thinking.
The river ice went out this week, so we could get to the river for testing. Because of low water, we were able to use the pool and spillway just above an irrigation diversion dam. Since the water was less than knee deep at the spillway, the only individuals at risk were the occupants of the sub and target; no full-scale humans were harmed during the testing.

Intially, the sub would not dive. However, by adding ballast to bring the deck down to just awash, the vanes were then able to fly the hull under - Success! Depth control was good - the foam float was not pulled under, nor did the hull alternate breaching and diving. Under rb power, the Nautilus moved at about 1ft/sec. I was winding the rb 200 turns (easy if you use a 5:1 winder). Power provided by 4 strands of contest 1/4" rubber. This size rb should be good to 900 turns, so the power available is certainly much more than we used in today's tests.

The sub would run about 50' underwater at today's power level, then finish with about 10' of surface running before the 5" prop stopped. The propeller gear is from a Guillow's Sky Streak balsa airplane. I also tried a 4 bladed prop (by putting 2 red props on the shaft, held in an X configuration, more or less, by small rubber bands). The 4 blade seemed to make the sub run slower - one would expect a longer run time, in that case, but we did not test this. Both the 2 and 4 bladed configurations would dive the boat.

I could not directly observe whether the vanes were actually controlling the depth. It's possible that the buoyancy of the foam float was simply holding the sub up, like a sky-hook dirigible. The river still had ice in protected areas, so the water was too cold to stick my face in to watch :-)

I think the sub was actually under vane control, though: we could tether the sub in the fast water above the spillway, and watch it dive and maintain depth. River speed at the entrance to the spillway was 2-3x as fast as the speed the sub could make under rb thrust alone. Consequently, if the vanes were stuck on "dive", they'd have generated 4-9x the diving force in the spillway area, compared to the force they'd generate when the sub was moving due to rb power (lift/dive force is proportional to water speed squared). If the vanes were not moving, then the foam float should have sunk 4-9x as deep as the rb runs. Since the sub did not pull the foam float under any farther during the spillway experiments, versus when we ran the sub under it's own power in still waters above the dam, this suggests that the vanes were being adjusted by the float, per design :-).
The other tests performed today were ship vs sub. A small brig was anchored in the pool area, and the sub was sent to ram the brig's keel. The 2-part keel is designed to drop ballast when the "flapper" is struck, leading to foundering. The sub was able to sink the brig 2/2 when it struck the "flapper", and 0/2 when it struck the solid portion of the keel. Test run was only about 6-9 feet to target; the goal was to see if the flapper design worked, not to see if I could hit the target :-)

Future tests:
a) other vane shapes; today's vanes were modeled on Humpback whale fins with a good aspect ratio; an alternate design, awaiting testing, is triangular, with a reptilian look.

b) other vane locations, such as at the stern or bow.

c) construct durable, ram-the-target-capable vanes; today's test vanes were made of foam core board, fast to make, fast to fail :-)

d) RC control of rudder and vanes; a water-resistant cylinder to house the electronics is under construction.

e) sink a moving target brig, possibly with the assistance of the BATAM - Brooks' Analog Torpedo Aiming Machine (details to be posted later).

f) compare actual dive vane forces to computer prediction (NASA FoilSim program for flat plate lift in water).
Neon Tetra rubber band sub:

Pierre-Yves Garcin's 1904 rubber band subs:

Michael Crisafulli's collection of Verne's Nautilus designs

Target Brig details here:
Last edited by Brooks; Mar 12, 2011 at 10:18 PM.
Sign up now
to remove ads between posts
Mar 13, 2011, 06:17 PM
Damp and Dizzy member
Brooks's Avatar
Thread OP
Tested the delta-shaped vanes today. They were a disappointment, both in the stern and in the mid-hull positions. Although they were of the same wing area as the humpback vanes, they were completely outclassed by yesterday's vanes. Apparently, wing aspect ratio is important even at these low Reynold's numbers.

In the stern position (the modern sub configuration), the deltas were unable to dive the sub, even in the fast water above the spillway. I added stationary vanes at the bow, set in a dive orientation (about 10deg down at the leading edge). This allowed the sub to dive at the spillway, but without the strength of dive of the humpback vanes. Whereas the humpback vanes would send the sub to the bottom when the vane control line was disconnected from the float, the uncontrolled, stern-positioned delta vanes, even with the help of the forward vanes, were unable to get the sub to the bottom.

Water was clear at the spillway and sunlight was positioned so I could observe the vanes. I could manually pull on the control line and see if the vanes responded. The control line did change the angle of attack of the vanes from "dive" to "surface". Trim tab tests showed that small tab deflections did affect the angle of attack of the deltas.

When run under rubber power, only about a third of attempts got the sub underwater. In contrast, the humpback vanes were uniformly successful once I'd added the extra ballast (same ballast used today as yesterday). It was hard to get the sub to dive with deltas in the stern position.

The deltas performed a little better in the mid-hull position: achieved a couple dives with underwater cruising. The delta vanes needed a steeper dive angle than the humpback vanes. But if the angle was set too steep, then the sub was slowed too much by drag, and the deltas wouldn't dive the sub. Of course, limiting dive angle of vanes is perfectly legitimate. But the humpback vanes sub seemed less sensitive to angle of attack.

The lack of dive power also showed up in the need for a more perfectly trimmed hull - ballast position fore&aft was critical. Again, this is not a killer requirement, but does lead me to prefer the less critical setup of high aspect ratio vanes.

Because of the prolonged experimentation time needed to get any dives out of the deltas, we were unable to practice target attacks. The brig's underwriters were happy :-)

So, to conclude:
A) delta vanes are inferior in performance to high aspect ratio vanes eg. humpback.

B) mid-hull position is superior to stern position for slow speed, rubber band powered subs. The rb sub designers of 1904 knew what they were doing.
Last edited by Brooks; Mar 19, 2011 at 08:47 AM.
Mar 14, 2011, 08:13 AM
Registered User
Rmay's Avatar
You know Brooks, I'm going to follow all your post. You always have interesting stuff. Do you post in the Airplane section? I don't know if I can keep up
Mar 14, 2011, 09:39 AM
Damp and Dizzy member
Brooks's Avatar
Thread OP
Thanks Rmay :-). No airplane stuff yet this year; my last posting there was to show a better fly wire arrangement for the Eflight Neiuport-17 (prototypical works better to absorb flight loads (and is easier to install) than the plans' setup).

btw, I'm reading a 1993 English translation of "20k leagues under the sea", trans. Miller & Walter. As I guessed, Verne's sub had mid-hull vanes:

Nemo describing Nautilus pg. 86 -"But I can also make the Nautilus ascend or descend, on a vertical plane, by slanting its two fins. These are attached to the sides at the center of flotation."
Mar 15, 2011, 10:34 AM
Registered User
Rmay's Avatar
"20,000 leagues under the sea" is my all time favorite novel. I only have a cheap hardbound book, but pick it up and read it every year or so. The Disney Nautilus will be, if I ever build a boat that “supposed” go beneath the waves, the first submarine I build.
Mar 20, 2011, 06:43 PM
Damp and Dizzy member
Brooks's Avatar
Thread OP

Today's tests and a video

New wooden vanes replaced the foamcore humpbacks. They worked fine once I added pb to the leading edge of the vane (hammered flat wheelweights, 14gram per vane). The foamcore vanes also required these leading edge weights (7gram per vane). Apparently, the trim tabs are (so far) unable to keep the vanes in a dive angle w/o the weights. More experiments with trim tabs are planned.

To protect the vanes during collisions with the brig, I added 3/32" bronze brazing rod guards. The guards can be clearly seen in the video.

I've changed the vane control from a brass frame and fishing line setup to a pure line setup (polyester carpet thread). This simplifies the mechanism, reduces drag, and makes the sub more crash-damage-resistant. While testing the sub above the spillway today, I could see that the float was moving the vanes from full dive (about 30deg dive angle on the vanes) to maybe 20deg dive. Sub depth held steady.

Under rubberband power, sub had no trouble diving from the deck awash surface condition. There were no failed dives. Depth held steady in rb tests (no porposing of hull, nor cyclic submerging of the float). Shifting the ballast to trim the hull to slightly bow-down led to deeper dives, while trimming to slightly bow-up led to shallower dives (sub remained entirely underwater in both conditions, however). I am still not sure how much of the depth control while rb-powered is due to vane movement, and how much is simply due to the skyhook zepplin effect of the float. I'll have to fix the vanes in the full dive position and see whether the float is pulled under farther. I know the vanes help in the high speed water above the spillway, but do they make a difference in lower speed rb cruises?

I am also testing a different prop, an APC 5.1x4.5E thin electric, LP05145E. The red Guillows prop is a 5x8.5 - rubberband props are the opposite to gas props in that the pitch is greater than the diameter for rb props. The APC is sort of a compromise between pure rb and pure gas. I went to a lower pitch with the APC to see if the sub would travel faster. Today's tests were not timed, but our impression was that the sub is much more zippy. A couple times Nautilus got away from us.... Wet pants resulted :-) I tried to count prop rpm, but it was way too fast at the start of the run. Total cruise length seemed shorter, but I don't have numbers for comparison with the red prop cruises.

The sub ballast is now enclosed in a 3/4" pvc tube, reducing drag. I'll have to go back to the red prop to see how much of the sub's speed pickup is due to the slicker ballast tube. The new tube can also be seen in the video.

Rather than a leisurely cruise towards the brig, the sub now moves with purpose and (evil) intent. No sinkings resulted, however:

a) my aim was abyssmal. In mitigation, I'll claim swirling waters in the pool - river is higher... in fact, you could see the sub sashay in invisible currents during the run-in to target, and silt stirred up by my boots quickly disappeared.

b) I set the design depth too shallow with the new vane control lines; when the sub hit the flapper, the hit was too high to get good leverage to pop the brig's ballast pin.

I am glad I left the vanes white. White is appropriate for the chambered nautilus "shell" design, of course. But the white vanes show up clearly even in turbid water. Of course, you have the float to show the sub position. But the float does not always track well, sometimes shooting off to the side like a water skier. So, w/o the clearly visible vanes, you might not know just where the sub is relative to the float (or brig). This knowledge will be critical to the skipper for guiding Nautilus to the target, once rudder control is installed.

Video here, sub run-by:
Last edited by Brooks; Mar 20, 2011 at 06:50 PM.
Mar 27, 2011, 04:43 PM
Damp and Dizzy member
Brooks's Avatar
Thread OP

Squid and Rudder Escapement

Squid - Finished the squid/float. He's carved from pink construction foam. To smooth the foam, I applied a coat of water-diluted vinyl spackle. After light sanding, I waterproofed the beast with 3 coats of Varathane waterbased polyurethane (Lowes). That is very nice stuff to work with, no odor at all. Red color is from a coat of acrylic paint. The tentacles were spun from 3 strands of acrylic yarn (Walmart) on my miniature rope walk. They are 40% longer than a scale giant squid's so that they will reach the sub hull when the sub is submerged to 12" depth. The tail fin (red) and keel (clear) were cut from a blisterpack. The eye is a pearl-plastic button with a black painted pupil. I've seen eye buttons on the web, but the simpler/cheaper plastic buttons look good (Joanne's Fabric).

While pursuing prey, squid move in the direction of their 8 arms. The arms are kept close together to streamline the body. Once the long tentacle pair have captured the prey, the arms are flared in the classic squid pose. Japanese researchers have captured some of the hunting actions of live giant squid on film. They are scary even in photos :-).
Rudder Escapement - My RC parts for the water resistant cylinder (WRC) are taking a long time to arrive (Hobby King). For fun, I made an escapement steering mechanism for the sub's rudder. Tugging on a fishing line leading to the escapement cycles the rudder between port/starboard/center. This mechanism will ring a bell with any pre-proportional RC'ers (or clockmakers) :-). I made this one from Lego Technics parts; it's powered by it's own rubberband.

How it works - Under torque of the rubberband, the escape wheel (Large Lego gear) attempts to spin counterclockwise. An pushrod from one of the teeth moves the rudder concurrently with the wheel rotation. The rotation is arrested when a tooth is stopped by the first pallet. Rocking the yoke back and forth by tugging and releasing the fishline allows the wheel to continue its rotation (and thus moves the rudder):

a)When the fishline is given a tug, the 1st pallet is lifted out of the way of the tooth. The wheel then rotates until the tooth hits the 2nd pallet.

b)When the fishline tension is released, the 2nd pallet releases the tooth, and the wheel completes it's first segment of rotation.

Successive teeth give the rudder it's port/starboard/center movement.

The escapement mechanism is mounted on the port side of the black tail fin, while the rubberband torque mechanism is mounted on the starboard side. Due to an error on my part, the escapement is not mounted correctly, thus there is only one center rudder position per revolution of the wheel; normally, the sequence of rudder movements would be port/center/starboard/center. I had to remove one of the teeth to get a true center, and my right and left positions are not symmetrical. In the water, the sub will be able to turn faster towards one side due to more rudder movement. Correcting the escapement mounting (by rotating the whole escapement mechanism about 20deg CCW) would yield the standard and symmetrical movements one would expect from an escapement.

The first version escapement needed more rubberband torque; it worked fine on the bench, but,in the icy waters of the river, failed to trip reliably. This first, simpler, version just stretched a rubberband, transferring the force to the escape wheel via a string wrapped around a drum. The 2nd, more powerful, version, uses a wound rubberband, and a 2:1 gearing setup. It works fine on the bench, snapping the rudder to successive positions with authority. We'll have to see what it does in the real world of icy Montana streams :-).

When the RC parts arrive, I'm not sure what I will do. It's handy to have a fishline to the sub since it simplifies retrieval after a run :-). So I may continue with the escapement, saving the WRC for another sub. Or I may opt to return to the simplicity of the un-controlled rudder, and scrap the rudder control entirely.
Mar 30, 2011, 12:49 PM
Damp and Dizzy member
Brooks's Avatar
Thread OP
The analog torpedo aiming machine (BATAM) thread may be found here:
Apr 02, 2011, 04:53 PM
Damp and Dizzy member
Brooks's Avatar
Thread OP

Squid and Sub videos; Hydrophone test

2 videos with the new squid float.

Too windy to try for a brig attack, we went subbing just as a squall moved through. Of course, the wind gusts did not bother the sub one bit, but surface vessels would have had a hard time of it :-).
PVC hydrophone - 1 1/2" pvc pipe, 3"-to-1 1/2" reducer, diaphragm of cut-up latex glove (held on with 2, #64 rubberbands).

This worked well in detecting 2 rocks clicked together underwater (UW). Detection range was about 30-50 feet in a high noise environment (test just upstream of the dam spillway). I could not hear the rock clicks w/o the hydrophone.

I could also gather some directionality information by noting the loudness of the click. The next test will be with a pair of hydrophones separated by a 7 foot spar. WW2 US subs used a 7 foot separation for their hydrophone mounted on the foredeck, but from what I've read, they used electronic microphones, rather than a giant stethoscope. The goal is to make a sub-finder for when I enter the world of lost RC subs :-).

One note, the diaphragm must be tight across the reducer. I tripped over a log, and apparently stabbed the end of the pvc into a small rock. This left a dent in the latex (and water in my waders, brrr). Whereas before the accident I could hear UW rock clicks fine, afterwards...silence. I stretched out the latex to remove, mostly, the dent, and the hydrophone was operational again :-).
Last edited by Brooks; Apr 02, 2011 at 05:15 PM.
Apr 05, 2011, 05:24 PM
Registered User
WOW! I never even knew that RB powered submarines ever existed, I will be following your posts!
Apr 10, 2011, 03:29 PM
Damp and Dizzy member
Brooks's Avatar
Thread OP

Video of Nautilus hitting brig

A very short video of the Nautilus capsizing the target brig is posted.

A new target ballast release system was tested and it worked. Details will be posted on the brig thread. Basically, a line runs along the leading edge and the trailing edge of the target's keel. The sub bow, or a barb on the sub's vane guards, snags the line and pulls the target's ballast retention pin. The new method increases the target area underwater to the full width of the keel (and increases the sub's effective attack area to the full width of the vane guards). The disadvantage of the new system is that the sub can't hit and run, that is, sub can't continue it's underwater cruise after striking the brig. The sub becomes entangled in the release line, so ends up circling the target. This does force the target off course, so maybe the sub should get half credit for a capture, even if the target does not sink :-).

Hitting the moving brig is hard, as expected. Some of the difficulty was due to lack of manpower. You really need a brig skipper, a sub skipper, and a camera operator. We had only 2 people, only one of whom had waders, so we were limited to short runs for the attack. Jenny ran the shore ops, including camerawork, while I ran the sea ops. I'd launch the brig, then take the sub and wade to an attack position. The brig was too fast for me to have time to wade more than a few feet away. Also, if I waded fast to the attack position, I stirred up the water so much that launching the sub on a predetermined course was futile. Both of these problems will be remedied when my 9yr old assistant gets out of school in a couple months :-)

There were plenty of misses...Mostly behind the target (not leading enough...I did not have time to use the BATAM, so had to aim by eye, which is obviously out of calibration). On two attacks, the barb stuck in the gap between the flapper and stationary portion of the keel...since the gap is only 1/8-1/4" wide, you could say I was defeated by superior aiming, shades of early WW2 US sub difficulties *smiles*.

While we had several collisions between sub and target, only one resulted in a sinking. I'd locked out the flapper sinking mechanism to concentrate experiments on the new system. Some of the brig escapes would have been thwarted if both systems had been operational, I think.
Last edited by Brooks; Apr 10, 2011 at 08:09 PM.
Apr 10, 2011, 04:13 PM
Damp and Dizzy member
Brooks's Avatar
Thread OP
I forgot to mention that the escapement rudder works now. I stiffened the Lego structure around the intersecting gears, which kept them in alignment and engaged.

The cold water (32degF) apparently shrinks the axle bearings (holes in the Lego beams). The escape wheel turns slower when in the pond than when the mechanism is at room temperature. This affects the timing of the skipper's tugs on the escapement control line (running to my hand on shore). I expect the unit will work more freely once the water warms up.

I also doubled the rudder area. The sub now turns both port and starboard.

The complication of escapement-steerable sub operation is not as attractive when it's so cold out. Simplicity rules when your fingers are wet&frozen *grin*.
Last edited by Brooks; Apr 10, 2011 at 11:03 PM.
Apr 10, 2011, 07:54 PM
Registered User
Why was it that you needed the hydrophone?
Apr 10, 2011, 08:16 PM
Damp and Dizzy member
Brooks's Avatar
Thread OP
Mj, no need for the hydrophone with this sub. That was simply an experiment for when I build an RC sub.
Guys who have built fancier hydrophones report that simply listening to what's happening underwater is fun in itself. They were speaking of saltwater, so maybe the fish and crustaceans are noisier in the sea than in a freshwater pond?

I have read that some British model submariners install pingers in their subs, and enjoy tracking them. RAM has some sub sound modules that might be fun to install in a WTC. My boating buddy John says I should stick a playing card in the way of Nautilus' prop :-) It would be one way to develop a prop rpm vs sub speed chart....there is, apparently, no limit to how crazy one can get even with a simple rb sub, hoho.
Apr 10, 2011, 10:42 PM
Damp and Dizzy member
Brooks's Avatar
Thread OP

Sub dimensions

A viewer asked for plans. Here are the dimensions of the sub. The only critical dimension is the size and shape of the dive vanes, as discussed earlier.

Changes during testing not reflected in first photo:
1. I extended the ballast hanger frame downwards to enable the pvc ballast container to hang below the rubberband. Originally the ballast was placed above the rubberband. The ballast hanger is offset to the side so that it does not interfere with the rubberband. This offset does not materially affect the trim of the boat.
2. Prop change to single APC prop, as discussed in previous post.

The dive vane control line simply runs up from the vane, thru a screweye, then to the top of the sub. At the top, the 2 lines (port and starboard) are tied together, then a single line runs aft along the deck to another screweye, then up to the squid. The gap in the eye of screweye is closed with a drop of medium CA to keep thread from jamming or jumping out .

Squid dimensions don't include the clear keel.
Last edited by Brooks; Apr 10, 2011 at 10:58 PM.

Quick Reply

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Rubber-Band powered helicopters! Accu157 Life, The Universe, and Politics 21 Mar 17, 2019 05:45 PM
Discussion Rubber band powered micro's? metroidrc Scratchbuilt Indoor and Micro Models 21 Nov 20, 2007 02:35 AM
Discussion places that sell rubber band powered, balsa wood, indoor plane parts alienshards Alabama's Rocket City Renegades 11 Mar 20, 2006 01:45 AM
Rubber Band Powered to Electric BackYardFlying Scratchbuilt Indoor and Micro Models 6 Mar 23, 2005 03:55 PM
How to determine pitch on rubber band powered propeller? aeropete Electric Power Systems 4 Mar 26, 2004 04:09 PM