Akhenthot

Dolphin #23

2011-10-07T09:52:00.000-07:00

With excellence in form, even brute force will be graceful. The Dolphin does what it was designed to do, and does it well. Many things could have been improved, but such is the fate of all prototypes. What is important is to answer without doubt, the questions pondered that led to this project. A craft may be built that will run under water, at great speed jump out of the water reaching significant height before plunging down again.

I had to take these pictures myself while simultaneously controlling the Dolphin. Not too easy. It's kind of unfortunate that I didn't get any pictures with huge amounts of air, but there will surely be more opportunities before winter. At any rate, everything seems to be working well. I'm amazed that the ESC and motor doesn't fry, uncooled as they are. I've tried two times now to capture some nice underwater shots, but there's simply too much gunk in the water, and not enough light. Think I'll try and make myself a subsea light pod with some hefty LEDs. Also, next time gotta remember to bring a peon to man my photonic capture device.

Dolphin #22 - No propellers for you!

2011-08-31T17:19:00.000-07:00

A video! Very very early, limited, silly and poor, but nonetheless. Using a tiny prop with 2-3 broken blades I do a little bit of testing. Center of gravity is improved with a new titanium front nose, and light plastic-geared rear servos. Still running in the tiny, filthy nearby pond which obscures sight and breaks propellers. I now have zero propellers.

Dolphin #21

2011-08-30T14:29:00.001-07:00

Trimmin' my sub in the tub. Here I have taped a co2 cartridge (~60 grams) to the front of the sub. Barely visible is a piece of cork taped underneath the rear, giving approximately 50 grams of buoyancy. This is sufficient to give a reasonable center of gravity, while maintaining about 150 grams of overall buoyancy. I can remove some rear mass by switching the two rear servos from metal-geared to plastic ones. That'll save about 40 grams and can be done right now. As for the front, I could put a small counterweight inside the nose, but even better would be to make the nose itself out of a different material. Titanium would be ideal, taking the mass from 23 to 78 grams. I'll see what I can do.

As for the test runs, they went.. fairly well. I managed to jump about half a meter into the air once, but shortly thereafter the propeller had an encounter with a rock. What used to be a three-blade propeller instantly became a one-blade propeller. Unfortunately, there's no footage yet. I only have one propeller left right now, and would like to at least partially fix the center of gravity before having another go.

Dolphin#20

2011-08-28T05:14:00.001-07:00

Finished the new hull. Looks pretty neat, doesn't it? Okay, it does look a little 'dirty', but that's just because the Kevlar is only one layer thick, and the spotty foam underneath shines through. Final mass: 325 grams. I'm pretty satisfied, considering it is made from no less than six different materials: aluminium, carbon fiber, epoxy, expanded polystyrene, micro-glass balloons and Kevlar.

While charging the batteries for today's test run I waded through a nearby swamp for an hour, looking for a nice shallow pool with little current. And I did finally find one, twenty meters from where I started looking. When the batteries are charged, I'll head out. Unfortunately there's no one around who can do any filming, but I'll se what I can do with my GoPro on a stick. It'll be lo-fi for sure.

Jumping out of the water won't be a problem

2011-08-27T12:29:00.000-07:00

Holy Mowgli. Holy macaroni of mammolini. Holy baloney. I finished the test assembly and decided today was going to be the day. A baby harness was constructed out of nylon thread and then off I went to do first submersion. When I reached the riverside the face of the water was pitch black. Not much light anywhere else. Couldn't even see how deep the water was. I fired up the dolphin and hoisted it into the water, hoping it wouldn't escape its harness to go die in some sandbank. (Skunkworks engineers can tell you there is nothing more suicidal than prototype vehicles) Rudders check. Engine? Give it 5%. Dolphin very nearly leaves the water, despite the fact that I fitted it with a puny 2-bladed 40mm propeller before leaving. Looking good. This baby isn't going to jump, it's gonna fly.

I didn't dare do much else, given that it would have been hopeless to recover it if it got away. I buzzed around a bit, packed up and went home. The only disadvantage I could notice was that the center of gravity is very far back, leading it to float nearly nose-up. The new hull will probably help a bit, and I can also add some floaty stuff in the back and some heavy stuff in the front to even it out. Honestly though, it didn't seem to be a problem in practice. As soon as I gave a little thrust it leveled out nicely. Tomorrow will be interesting.

I'll be sure to record some videos of it. A compadre may for instance be submerged along with a GoPro camera, to get some subsea footage. Should go nicely with the Das Boot-soundtrack. Anyway, time for a celebratory can of Coke.

Dolphin #19

2011-08-27T10:30:00.000-07:00

I haven't made an update about last weekend, but suffice to say that I spent nearly twenty hours machining the remaining components. The front fin system is ready. The rear fins are ready. The engine is ready. The internal structure is ready. The new hull.. uh, still need to glue on the last piece of the kevlar shell. In the picture you are seeing final assembly for testing. I have spent a few hours now assembling, disassembling and reassembling it to fix minor errors, all the while going slowly insane. Today or tomorrow I will do a minor submersion test with the old hull, to verify all seals and such. In the picture you can see a new propeller, hilariously oversize. I bought a bunch of them on ebay in various diameters and pitches, and the largest one is the only one I have received so far. It's guaranteed to fry the motor if I actually try to use it. On top of that it is terribly unbalanced. Looks cool though.

Dolphin #18

2011-08-12T07:21:00.001-07:00

The carbon fiber beams have been cut to length, and had threaded pins glued into them. The assembly structure is now finished. I tested it with the old hull, and everything seems to work perfectly. It was quite easy to screw the three parts together by turning the nose. One thing to consider is that there is nothing maintaining the alignment of the fins, save for the friction created by tension and the hull o-rings. If the body parts slide out of alignment things could get complicated. I'm betting however that the friction will be sufficient.

The front fins are locked together, by means of the aluminium shaft seen here. A linkage will be used to connect this to the horn of a servo lying horizontally below. The servo won't be very well aligned with the middle of the aluminium shaft, but I'm hoping this too won't be a problem.

The front fins! Unfortunately, this is as far as I can get at the moment. I need some coordinates and angles for the curved cut that gives the correct spacing between the fins and the body. In addition to that, my summer vacation will soon be over. I can complete the body during the weekdays, but milling/turning will have to be put off till next weekend. It's just a question of completing the fins and making the servo mount. Should be doable.

Dolphin #17

2011-08-06T14:37:00.000-07:00

More progress! I have finished the 'bulkheads', which are the two plastic rings supporting the batteries. I also milled the X-wing* with honey drizzler attachment shown in front, which is the structure keeping both ends of the dolphin in place. The drizzly bit goes into the very schnoz of the dolphin, which pulls the whole assembly together. The schnoz has two holes in it, which will be used for rotating it, using a tool I have yet to make.

The front end, with X-wing honey drizzler in place. The X-wing was very amusing to mill, not least because I didn't have any coordinates for it, and had to calculate all the intersection points by hand. It's sweet and fitting to blow out the sinuses of ol' Pythagoras once in a while. Also shown are parts of the front control planes, which I have started making.

I have also made some progress with the new body. Final mass is estimated to 310 grams, a near 140 gram improvement of the old body. It'll be ready in a few days, although I am beginning to have some doubts regarding how to proceed with the Kevlar surface. Should I glue a pre-hardened shell onto the surface, perhaps resulting in an improper fit, excessive use of epoxy etc etc, or should I apply a fresh layer of Kevlar, which will be impossible to cut to proper size beforehand, therefore resulting in a messy lay-up and frayed edges everywhere? Ah, engineering, where every option has downsides.

Edit: For the record, I am completely against the use of honey drizzlers. They're quite silly devices, about as practical as Mimosian anti-matter chopsticks.

*Edit: I am an idiot

Dolphin #16

2011-08-03T03:46:00.000-07:00

In spite of the tenets of physical culture I find myself in the basement during summertime, toiling away in the hopes of offering unto Primus, lord of Modrons, a working dolphin before autumn sets in. Quite a bit of progress has been made, none of which has made it to this blog, due to simple neglect on my part. I have now finished the servo system for the tail fins, meaning that the rear 3rd of the dolphin is essentially complete. I have also completed several other components, which will be documented in due time.

For controlling the dolphin, I bought a Futaba T6EXHP, the best FM transmitter I could easily get my hands on. By using V-tail mixing I can use one axis to move the tailfins in unison, and another to move them oppositely. This makes it easier to control turning and rolling. Unfortunately the V-tail mix is designed for planes, meaning that the stick inputs end up being unintuitive. I suppose I could open the transmitter and swap the axes around manually.

I have also made some observations regarding the dolphin design. I designed heaps of tiny metal components that have turned out to be a pain to make. My crappy homemade CNC mill is also showing its limitations in this regard. Further, the decision to mount servos and motors onto the front and rear parts of the hull means those parts ended up heavy and with heaps of critical features. It would probably have been better to make an internal skeleton, onto which all servos, motors etc were fixed. Third, I chose to make the tailplanes independent, rather than the front planes. This means one more servo in the rear, where the heavy motor is already displacing the center of gravity. I hope this can be mitigated by the new, lightweight hull, which should allow for a counterweight in the nose.

Dolphin #15

2011-05-22T14:55:00.001-07:00

Now with 100% more tailplane! Also note the background, which is twill weaved Kevlar. I'll use this for the outer shell of the new body. I don't know exactly why, but Kevlar just looks infinitely cooler to me than carbon fiber. Probably has to do with all the useless imitation carbon fiber car parts there are out there. Kevlar is actually pretty cheap, and has some downsides and advantages compared with carbon fiber. Kevlar has a lower stiffness (E-modulus or Youngs modulus) which is a downside in parts like bike frames, although it can be useful in controlling where a component will flex, and also in absorbing impact energy without fracturing. The greatest advantage of Kevlar lies in the toughness of the fibers, giving high resistance to cutting and abrasion. That's why it's used in cut-resistant and biking gloves. In fact, Kevlar is so tough it's very difficult to cut with regular scissors. Professionals use serrated scissors or pizza-wheel type cutters to abate this problem.

I actually did the first casting of the new body today as well. The two aluminium rings mentioned in the last post are now sitting on a mandrel with a carbon fiber tube between them, hardening right now. Rolling the fiber cloth onto the mandrel was awful, like trying to bandage an eel.

Dolphin #14

2011-05-21T13:13:00.000-07:00

Milled the first of two tailplane horns. (It's the metal handle sitting on the tailplane shaft) This was an utter bitch to make on my horrible homemade mill. A new xy-table (ironically one of the parts of the mill i DIDN'T make myself) is becoming ever more of a necessity. As for the body, my plan is to prep the core of the mold tomorrow, hopefully letting me do the first casting in a few days.

Dolphin #13

2011-05-20T11:39:00.001-07:00

Well, progress has been very slow as of late, due to various reasons among which laziness is easily a top contender. Not all is lost, however, as I can document three small updates. The two aluminium rings shown above are intended to be at each end of a new body for the dolphin, which will be made primarily of carbon fiber and expanded polystyrene. With this upgrade I'm hoping to reduce the body mass by perhaps 100 grams compared with the old acrylic+syntactic foam body.

Another update is that the front nose of the dolphin is now finished, save for two holes I will drill into it to make rotation possible. (This rotation will assemble/disassemble the entire dolphin)
Lastly, The tail end of the dolphin is also finished. It's the bolted-on part right behind the propeller. This component holds the driveshaft in place when reverse thrusting. The four bolts make it easy to replace the driveshaft in case it gets bent, or when testing a propeller with a different mounting type.

Dolphin #12

2011-03-26T02:17:00.000-07:00

I have tested the Dolphin motor in my aquarium. Sort of silly, but it certainly looks promising, and who doesn't like bubbles? I've also received the new ESC (Mamba Max Pro) and milled a cradle for the Dolphin to sit in. Pictures to follow.

Composite Mill #1

2011-03-04T16:01:00.001-08:00

Progress is painfully slow on most of my projects. The root cause is, I have crappy tools. The CNC mill I built has a shoddy XY table with a tiny Y-axis, trapeze screws with horrible backlash and so on and so on. Conclusion: I must build a better XY table with linear rails and ballscrews. Unfortunately this will make the whole shebang a lot heavier, and that's not something I appreciate. Therefore, in addition to a new table, I will have to replace the knee that forms the basic structure of the mill. The obvious choice of material is carbon fiber. By molding over a styrofoam core I can make a rigid, lightweight backbone. If resonance becomes a problem I can carve out the styrofoam and fill this core with sand. I haven't heard of anyone trying this before, which means it is sufficiently crazy/silly for me to question whether it is possible at all. In other words, the kind of project to which I can apply infinite effort.

Dolphin #11

2011-01-30T10:36:00.000-08:00

Back to the Dolphin. I've milled and drilled the motor mount. Now for some bolts.. Strange as it may seem for a guy like me, I have to hit ebay. Don't know how I've managed to come this far without having every conceivable bolt within arms reach. So who's up for some used pink bathrobes or rare mint snowglobes? Oh, and M3 bolts of course..

Coilgun #5

2011-01-16T13:17:00.000-08:00

After some more coding the coilgun is now entirely autotuned, save for the first stage. I'm timing how long it takes for the projectile (which is a fixed length) to pass in front of the optodetectors, and for each stage this value is used to calculate how long to wait and how long to fire each coil.

Peak performance right now is about 11m/s, but it can probably be improved. By letting the coils stay on even as the projectile enters the next optodetector, I can squeeze out some extra performance. All that's required is to rewrite the code to use an interrupt to turn off the coils, which will free up the microprocessor in the meantime.

The coilgun accelerates its projectile because there is a repulsive force between it and the coil. The coil obviously has a symmetric magnetic field, which means that only 50% of the coil can be used for repulsion. Perhaps one might construct a conical coil, or fit a regular coil with a ring magnet on one side which would cause the magnetic field to become asymmetric, such that a greater distance of the coil would be usable for acceleration. It would look more like a long trumpet and less like two opposed funnels.

Edit: changed some code, got up to 12,2 m/s. Strange thing is - efficiency seems very high. Gotta do some accurate measurements.

Coilgun #4

2011-01-15T17:01:00.001-08:00

Achievement unlocked: multi-stage coilgun. Everything is very prototypeish, but it works. I've dialed things back a little for now, and have four stages running at 2 cells lipo. That's just 8.4 volts! Muzzle velocity is approximately 12m/s, with some very rough tuning. The big circuit board is the controller, based on an Atmega16. The board has a bunch of indicator LEDs, switches and eight output optocouplers. It's powered by the blue 3-cell lipo at the bottom right. At the bottom left are the MOSFETs which do the actual power switching. These are four stacks of four IRF3415. On-resistance is thus about 10 milliohm. The initiator circuit I spoke about earlier is not implemented, so the actual coil current will be quite impaired and lopsided. I'm guessing about 50 amps, tops. Note that I have absolutely no protection on these mosfets. I'm amazed that they can handle the punishment. Once the field is switched off the spikes must be several hundred volts. When I build the initiator they have to go however, as they are only rated for 150 volts, and I'm planning for the initiator to be closer to 300.

I'm monitoring the raw output of the optotriggers on my scope. Here is stage 3 (yellow) and stage 4 (blue). The troughs are when the projectile passes in front of the IR led, which is what is used as a triggering event. This definitely makes me wish I had a 4-channel scope. Incidentally, these Rigol scopes are great for hobbyists - cheap and powerful.

Coilgun #3

2011-01-09T04:29:00.001-08:00

Here's an LTSpice simulation of the circuit that will power my coilgun. L1 is a coil. M1 is a bank of MOSFETs which turn power to the coil on and off. V2 is the control pulse to the MOSFETs, and will come from a microprocessor. V1 is a bank of lithium-polymer batteries. This bank is capable of delivering currents of around 200 amps 'continously'. Unfortunately, this battery bank has a very low voltage, and the rise time of current in the coil would therefore be very long. To mitigate this I have a tiny capacitor C1 which is charged to the maximum voltage allowable by the MOSFETs. This capacitor will 'punch' up the current in the coil, whereafter the batteries will supply current for however long I want. With this setup I can get an approximate square wave current lasting anywhere from 0.1ms to several seconds. Of course, I haven't actually tested this yet and have no idea if it will work or not.

Coilgun #2

2011-01-09T04:00:00.001-08:00

A prototype! While waiting for the UHMW blocks which will form the trays for the 8-stage coilgun I whipped up this prototype tray from some plastic I had lying around. This is 4 coils and 3 triggers (don't need a trigger for the first stage). The tray didn't come out as well as I had hoped, you need a lot of cuts to make such a surface with even a pretty large ball mill. Also shown is the magnet which will be my projectile. It's a 19mm diameter neodymium magnet with approx. N40 strength.

Coilgun #1

2011-01-08T14:36:00.000-08:00

With winter being what it tends to be in these latitudes and longitudes I reckon it will be some time before the rivers are once again ice-free. Therefore I'm not in any great hurry to prep the dolphin, and have started another project in the meantime. It's a coilgun. These were mighty popular on the net a few years ago but have since become somewhat obscure. They were largely abandoned because although relatively easy to build they required bulky components and had very low efficiency, typically 1-5%, which resulted in peashooter-level performance.

I have a couple of new components I want to try out, for a bit of an alternative approach. A combination of ithium-polymer batteries and field-effect transistors will by no means give me great advances in power, but it promises great compactness and with microprocessor control easy tuning. Further, by substituting the traditional iron projectile for a neodymium magnet, I believe a greater coupling between coil and projectile can be accomplished, as well as letting me power the coils in a push or pull configuration, or even both at the same time.

Dolphin #10

2010-10-03T11:06:00.001-07:00

Finally some progress. The main body has been turned to its final shape, and given a light coating of epoxy. The tail has been turned on the outside, and the nose has been turned on the inside. Next up is some milling, and then I can turn the outside radius of the nose as well. Finally, this project is starting to look like.. some sort of oblong cigar-shaped device. It has already found some use as a Mega Man prop.

Dolphin #9

2010-09-02T13:45:00.001-07:00

The interior of the tail end is now milled and drilled. Little remains before I try to turn its outer shape.

Dolphin #8

2010-08-29T13:20:00.001-07:00

Finally some progress. The black piece is the tail end of the dolphin, now turned on the inside. A couple of milling operations need to be done before it can be given its intended curved shape. The beehive-ish object on the left is the main body, now clad in syntactic foam mixed with expanded polystyrene. As shown, I have started to turn it to its final shape, after which it will be painted. In light of this progress, celebratory Das Boot soundtracks were played, and clips of Kon-Tiki watched.

Dolphin #7

2010-07-31T14:48:00.001-07:00

Finished the driveshaft, shown here with double ball bearings, u-seal and a candidate propeller. The driveshaft was machined from Super Duplex, a very nifty grade of stainless steel. Also shown are the bulkhead mounts for the tailplanes. These are AISI316L. Also also shown: a possible sponsor? I wish.

Makin' chips. I use an ER32 collet to hold the workpiece. Beats clocking those big 4-jaws every time. Machining stainless was surprisingly easy. Light cuts were very predictable, and of course it looks great when polished.

Dolphin #6

2010-07-25T04:55:00.000-07:00

Success! By mixing about 35% (by volume) expanded polystyrene the density of the syntactic foam was pushed down to 0,5 grams per cubic centimeter. This material feels seriously odd. It is very hard, yet incredibly light, like petrified shaving foam. I'm pretty sure the density could be decreased even further, by using less microballoons and more EPS. Now all I need is somewhere to get EPS pellets. It took me half an hour to pry apart enough pellets to make the above test block.

Dolphin #5

2010-07-24T13:50:00.001-07:00

Syntactic foam, made with epoxy and microballoons. This will form the curved shape of the mid-body of the dolphin, shown below. It has to be as light as possible. At 55% microballoons by weight the epoxy turns into a thick paste that is rather difficult to apply. Furthermore, 0.7g/cm^3 is the bare minimum I need. Probably going to try adding some expanded polystyrene to the mix, and make the part in two layers, with the outer one being more of a pure epoxy glaze, to give a better impact strength. As an aside, syntactic foam is used in the flotation elements on ROVs. With proper materials they can achieve densities of 0.5g/cm^3 in elements rated to 3000m depth.
I'm considering a kind of conning tower, shown here in red. It would be a good place for an antenna, as well as charging and on/off bulkhead connectors.

Dolphin #4

2010-07-10T15:09:00.001-07:00

Made an Oldham coupling for the main driveshaft. The middle element is PEEK, one of the better engineering thermoplastics out there. An oldham is probably not ideal, but they're easy to make. Also pictured is the brushless motor with its purple anodised water cooling shroud.

Dolphin #3

2010-07-07T14:18:00.000-07:00

A tailplane. Not bad for a homemade 3-axis mill. It has a pseudo-airfoil shape made up of four circular segments. Material: Some kind of mystery meat plastic.

Dolphin #2

2010-07-06T13:53:00.001-07:00

Last weekend I made a small maquette of the dolphin and put it in a 20 liter pyrex bottle I had lying around for just this sort of thing. Food coloring was added to taste. (It's held in place by a thin fishing line, probably doesn't show up in the image)

Dolphin #1

2010-07-06T13:39:00.000-07:00

This is the Dolphin. It's an RC sub I am working on, whose one and only purpose is to attain sufficient speed to 'jump' out of the water like a real dolphin would. At the moment I am marooned somewhere in the hideous nowhere between design and execution. Weight in particular, is becoming a problem.

Some specs:

~500mm long

2.7 kg

1200 watt brushless motor

4 cell 5000mAh LiPO battery

2010-07-04T10:25:00.000-07:00

test