FreekS

Super detailed work and great photography!

I think I had this idea after an episode of “The repair shop”, a UK TV program. Someone brought in a model of a ship, cut into puzzle pieces. Having done a trip on one of the nicer sailing cruise ships, I thought it would be nice to have a puzzle of the ship. “Star Clipper” is a 115m 4-master sailing ship (with engine, note smoke coming from the mizzen mast) from 1992. It can carry up to 170 passengers and has the first mast square-rigged and the others “Latin” (is that barquantine?). It’s a wonderful holiday opportunity, sailing in the Mediterranean or Caribbean depending on the season. The basic idea was to construct a “bread without butter” model of the ship, with each board representing a deck of the ship. Each board/deck would have its actual layout painted on it, and then the boards were cut up into puzzle pieces. I happened to have in my vacation photos a picture of the safety poster showing deck layouts (and sail layout). I did not have any plans and thus I’m warning the community that this is fun-quality rather than museum-quality! Here the first three boards (and the deck layouts prints of my photos below the). These represent the ship above the waterline. They will sit on two additional boards that contain the sea and the below waterline hull. The 5 layers will be held together by the 4 masts only, for which holes have been drilled under the appropriate angle. The hull was shaped by hand with the boards in a vice using a plane and sandpaper. next, the deckhouses, tenders and masts were added. I made a simplified deck planes, with color coded spaces (white-common spaces, red-passenger cabins, green-crew spaces) now turned all boards over and drew the puzzle pieces. A couple of days on my dremel jigsaw and numerous saw blades later I had around 300 puzzle pieces, still neatly arranged by deck. off the starboard bow is even an orca in the sea. Assembled: and then came the main test: it took about three hours to revert from the puzzle pieces to the model again. The normal strategy of finding the corners and edges first does not work, as there are so many of them….the motorised jigsaw was critical, as each puzzle piece must be cut perfectly vertical or the puzzle won’t fit. finally (and a few years) I added 16 more pieces in the form of 3D printed sails. She looks great and the puzzle has been completed several times! thanks for reading!

yesterday at my model boat club O-13 made her maiden test. Water was perfectly clear and just still liquid. The evening hours made it a great test for the LEDs (but somewhat scary for everything else). Made two small rounds, one on the surface and one semi submerged - the ballast is “on the safe side” and she does not want to dive at low speed yet. Filming was a challenge too but a friend at the club recorded the event. https://youtu.be/nnN6TKy9Hg0?si=mhxLVOyGmUrWRlAB

Recent focus has been on testing the boat. Based on the last bath test, I glued about 40 grams lead in the bow, and fixed with a small bolt 3 plates of 5 gram lead each inside the water tight compartment all the way in the bow. This for future trimming (as still a lot of small details need to be built and mounted). The stern was already heavy enough. Now, most photos of O-13 show her trimmed on the surface with bow high and the deck sloping down towards the stern. To achieve that and still keep the boat horizontal when dived, I glued a 30ml piece of building insulation foam in the stern, but above the waterline. I then placed 30 grams of lead under the foam and beneath the waterline. This cancels each other out when submerged, but pulls the stern down when on the surface. In the bath it looks “good enough” for now. In the bath the pump took 25 seconds to fill the tank and submerge the boat. That’s about the same as a WW2 sub could achieve. The pump actually works faster than its spec! The water sensor on the arduino now switches the pump off just when the mast, through which the air from the tank is pumped overboard, becomes filled with water. So it seems the tank fills completely (hopefully leaving no air bubble that can move fore and aft). The arduino also correctly measures the battery voltage and alarms when too low. The fore planes deploy but there is too much play in them. The crew goes below just before the dive but “forget” to close the hatches behind them. Oops. When the tank is full, the boat is trimmed to “decks awash” the props and diveplanes can then submerge the boat under very low speed. But that is for the next test. so the work on the technology is nearing completion, but much work remains. These pre-WW2 subs had a wooden deck, I want to build the “net cutters” on the bow, handrails, periscopes, radio antennas and bridge equipment, and of course get to painting. O-13 was deployed in 1937 during the Spanish Civil War to escort ships into the Mediterranean, and had clearly visible Dutch flags painted on the conning tower. That’s what I’m aiming for. its almost a year to the day since I started this build , I’m sure 2025 will see a lot more work on O-13 but also some nice sailings.

Happy holidays everyone! Captain below working as intended! https://youtube.com/shorts/58Pdq6fnKrM?feature=shared

As usual, the first wetting of the boat has revealed some work. First, I glued 500 grams of lead in the bottom of the ship, below the WTC. Also the techrack has too much friction, so I straightened out the M3 wires and adjusted the nuts. Sanded all contact points with the PVC tubes. It now runs better. Also, I wanted a magnetic switch so the whole boat can be switched on and off without opening the bajonet, using a magnet. The normal units use relais, and I did not have the space for them. Instead I made a small circuit with one MOSFET, one resistor and a reed–switch. I’ve tested it switching up to 2.2A, and when the reed is open there is no measurable current use. Seems to work! The reedswitch is located inside the water tight compartment, underneath a hatch. I put a magnet inside a wooden crew member, and when I place this in the hatch, the boat is switched off. On removal before sailing, the boat is switched on. It just fits between the existing main switch inside the WTC and the Arduino controller. I think the techrack is now complete, and next step is to test the boat in the bath, further balancing, filling and emptying the temp to dive and raise the boat and unfolding of the planes.

I tried the dome cap - but it did not solve the issue. So I looked for modifications to the double hinged balance - while maximising the vertical hight the (tall) Dutch captain can rise out of the hatch. I came up with a double hinged balance, with side by side pivot points. That seems to do it. I also tried to replace the blue foam floats for 3D printed hollow shapes which can be fitted more precise in the sail. On the top of the picture is the printed float. The three arms keep the float and the captains platform level. The two half-cylinder shapes left and right are connected and are the float to open and close the hatch still not sure if this will work in water with minimal differences in gravity force and archimedes’ buoyancy. Only a practical test will tell… but first I did a basic balance test, to see how much lead the boat needs. the WTC endcaps were glued, and replied until airtight (I blow in a tube as a waterproof test). Then the boat was “launched” in the bath. it needs about 800g plus 620g that simulates a full dive tank to “hover”. And when then removing the 620g the boat was reasonably on waterline. Needs trimming though very happy with this!! I have some more work to do to test the props and pump for the dive tank. And I need to review the front dive plane mechanism, the plane looks like it was hit by a depth charge.

That looks like a great idea!

Working on the “captain below” feature. O-13 has 4 hatches. I’ve made those from wood before, but to have low friction hinges I decided to print these. All hatches should open when the boat surfaces, and close before submerging. The hatch to the forward torpedo room has just enough space below it for half a crewman to stock his torso out of the hatch. by the way; the 1:50 winch holds the mechanism below and is the first part made with my first ever, very old Unimat-3 Lathe! I’ve mounted him on a printed double-arm balance, so he comes vertically up. On the other side of the balance are a 4ml blue foam-float weighted with 2 gram lead. So a “driving force” of 2 gram to either raise the float when the deck floods or to lower it when the deck drains on surfacing. The hatch is driven by the 2ml float weighted with 1 gram of brass on the right of the pic. It all seems to work dry…. although there is more space underneath the bridge, the officer emerging from the hatch there has to rise full length, to be visible over the bridge rim. The double-arm balance did not fit there, so the officer is pushed up by a simple counterweight and is guided by a brass wire. It’s a little more iffy - likely needing optimisation. The nut will be replaced by a compass stand made on my lathe.

That’s exactly what I’m thinking of, plus I will try to also close the hatch mechanically once the captain is down. The drawing of the actual boat gave me some idea how to do that; the hatch can also be operated by a float below the deck

Now that the “insides” of the sub are more or less done ( still need to make the connection from the power inside the water tight compartment to lighting on the bridge and hull), I’ve worked more on the wooden boat itself. Next major step is a wet test - so the wood needs to be impregnated. I wanted most of the remaining work on the hull finished so I don’t compromise the epoxy impregnating with future sanding and sawing. Thus I’ve made the holes to flood the hull, and the holes to let the air escape as that happens. The front dive planes are hidden behind a structured that I read was actually a set of hydrophones along the hull (in the 1930s). The bow needed a face lift. O-14 had a single stern tube, firing over a teakwood “deck” located over the stern rudders. Presumably to protect the torpedo from bumping into the steel hull. the weather was great for a photo shoot of O-13, together with her now 12 years old younger sister K-18, built to defend the Dutch East Indies from the Japanese. Both models are 1:50, O-13 was smaller as it “only” needed to cover the North Sea and the Carabean - while the K-boats defended an area as large as Europe. the bajonet especially needs more tweaking to fit the parts better together and I’ve started on the mechanism to make the captain go below on diving.

Last month or so has been the development of the arduino to control: 1. the pump pumping water in- and out of the ballast tank 2. The servo that folds/unfolds and operates the front diving planes 3. A low battery alarm and a water in the boat alarm, both resulting in surfacing the boat by emptying the ballast tank. 4. the navigation lights and alarm lights on the bridge it could get very geeky if I explain the details, but basically I’ve built a circuitboard first on a breadboard, with loose wiring to and from all components. realising all the “mess” needs to go into a small compartment next, I’ve soldered the components on a 4x6 cm circuit board (the max space available)and spent days finding soldering errors at the same time I wrote the sketch or software program and tested it dry. finally today I actually filled the tank with 570ml water before the pump shut off when the water sensors said it should, and emptied the tank again! Success! also did a first attempt to tidy things up so the whole assembly can slide into the waterproof PVC tube.

Mirabel, Great question shedding light on the post-build trimming! control at periscope depth typically depends on two things; - near perfect trim. Once the boat is watertight I will trim it in a tank with lead ballast (and sometimes with pieces of foam too). One must also make sure that if the boat has a ballast tank filled with water when dived, that there is no air bubble in it that may move around. Then trim during sailing, as the shape of the hull may impart some preference for the boat to rise or sink at some speed. the position of the ballast tank relative to the center of mass, the distribution of weight must be optimized to sometimes grams. That optimization can take months after the maiden…(at least with me). -a leveler like Norbert Bruggen’ Lageregler LR3. This is a piece of electronics between receiver and diveplanes. In the absence of up/down commands, it will operate the diveplanes to maintain the boat on an even keel. Commands from the transmitter are mixed with these. Levelers (or gyros) tend to work well for faster subs (like the 1:12 Delphin in the pic above). There are also subs with trim tanks fore and aft that automatically pump water around (just like real subs do) but I’ve not gone there yet… The most impressive boats tend to be the ones that can sail at periscope depth with very low speeds, when the leveler and diveplanes done do much any more. finally, as boats inevitably will go out of control or get lost from view, failsafes for loss of signal, low battery and water detected in the boat are used to force boats to rise (the 40mHz frequency used does not allow comms back to the transmitter). PS, I still always have eyes on the boat and two fingers nervously on the up/down control!

@Mirabell61 yes, that is exactly the problem. Skippers of surface models don’t spot a submerged submarine! Sub skippers always keep their eye on the tip of the periscope, the periscope wave, or slight decolorations where the boat is under water. That’s why sub skippers never remember who they were talking to at the waters edge! Generally, you sail as deep as you can just keep track of it. In a lake that is usually at periscope depth, or with clear water and little water reflection somewhat deeper. once or twice per year some outdoor pools offer the chance to sail subs on the day before that start or end of the season. Then you can sail just above the bottom and sail all sorts of maneuvers. I can really recommend it as a hobby, but you need to educate the other skippers in your club…..

Been a while since the last update. Much time has gone into repair of other subs post sailing - either by being rammed by unobserving skippers of (surface) models, or by time causing small leaks. found time to continue work on O-13 too. The joint problem seems to result from the PVC pipe not being exactly centred on the bajonet. A few wedges have improved the situation a lot! The sail has been roughly built of wood - with some thoughts as to make moving periscopes, a captain on the bridge that disappears below when the boat dives, and room for LEDs. I might include one of the 40mm guns that sit in waterproof buns. I think the boat carries only one (and no deck gun) as it had chronic stability issues! Not sure about the movable persicopes, the WTC directly under the sail means they would only rise 5 or so cm, seems not worth the trouble to construct pushrods to a servo inside. the front of the boat now has the covers for the front dive planes and spaces for the anchors constructed. Needs quite a bit of fine tuning too.

Hi Keith, that workmate is 37 years old! But I only started modelling in 2009 so maybe it’s had an easier life than yours! yes that model is a 1:32 Holland class; its O-1, the first Dutch submarine built in 1905. It’s about 10 years old and sails well! Like O-13, it’s made of wood with a dive tank inside.

The diveplanes, rudder and propshafts have been finished - meaning they are functional and firmly attached, but removable for further sanding and coating. one very big risk is that I lack a flexible connection between motor and shafts - the prop - shaft - gearbox - motor turns easily but I’ll have to see if vibrations don’t cause problems. I’ve assembled the main control components, they slide onto 4 M3 wires to form the “techrack”. This assembly slides nicely into the PVC pipes that keep the water out. The home made bajonet needs some more thought - having only two bolts to connect the two halves of the boat means it rotates a bit too easily. ive now started on the sail - which will be built of 1mm boxwood on a frame - has some acute curves through.

Thanks Ron! I will watch that! I’ve been mounting and remounting the propshafts dozens of times to correct the alignement ! I’ll be mounting the gears on the shafts today I hope!

Further with the rear of the boat. the rudders and diveplanes are made by soldering a 0,5mm brass plate on the shaft, then covering it with wood before sanding. Here the diveplanes were added - these are controlled through a brass rod coming vertically down from within the boat. Should be near invisible The prop shaft is a little too low here - it is driven from the above describes brass gearbox and fixing this meant reprinting the rear bulkhead of the WTC with a small 5 degrees offset from the horizontal. Both propshafts are reasonably aligned and holes made in the hull. The brass fins in below picture will be soldered to a structure in which the shaft runs and to the hull. This is a quite tricky part - I have little or no room for a proper motor-shaft coupling - so alignment must be very good to avoid vibrations.

Ok, further with self-taught engineering. Many if not all concepts below are well known in the model sub world, I did not invent them! I want to have a final design of the water-tight compartment (WTC) and its connections to propellers and rudders because only then can I plan and make the required holes in the wooden hull and finish the hull. The WTC will consist of three connected PVC pipes connected with 3D printed parts - aft: 44mm ID with the motor, motor controller, and two servos for the rudders and diveplanes, - middle: a 69mm ID pipe with the dive tank (a 60mm PVC tube), the receiver, the pump, a self-made valve as well as the ballast tank controller. - bow: another 44 mm ID pipe with servo for front dive planes and the batteries. These components will be housed in a techrack which is pulled out of the WTC. I’ve been designing these in Fusion360 and printing them on my 3D printer. The techrack will be mounted on four M3 rods to give it strength. here on the left the aft WTC , connected to a printed bajonet (located on the “cut” in the boat in background), and then to the right the beginnings of the techrack and dive tank The dive tank (here left) has a water-in and air-out brass connection. The resistance between these is measured by the dive tank controller to determine if the tank is full. The pump is a 500ml/min gear pump, and since gear pumps are not closed, I’ve re-purposed a mini-servo to pinch the air-out tube closed so no water leaks in when pump is idle. A 40 MHz 8 channel receiver will be mounted here as well. the rear WTC with motor axle sticking out (through a seal) and two o-ring holders for the pushrods to the planes and rudder. The motor drives a gearbox “in the wet” which drives two shafts and 30mm Raboesch props. The white printed part will be glued to the PVC pipe, but the motor can be unscrewed from the outside so that the rear techrack can be removed for maintenance. The screws must be sealed. here the tank again with its controller forward. Many electrical and signal wires will pass over the tank (which will be inside the 69mm ID WTC). finally, I did manage to do a little woodwork, the main rudder consists of two 0,3 mm brass sheets soldered to the 3mm shaft, and then covered with two layers of 1 mm boxwood on both sides. then sanded in shape. Also the future mount of the diveplanes is visible. sorry for the technical stuff - subs are a technical game, but I will get to the nice part of making the conning tower later!

No! A chemist!

I’ve gone down a bit of a rabbit hole! Having developed the idea of the functioning front diveplanes described above - the ideas for improvement kept coming! I’ve now got a prototype ready to be built into the hull, where the two movements of the front diveplanes, the folding and unfolding of the planes flush against the hull, and the up/down movement of the diveplanes when sailing under water, are achieved with just one servo. the up/down movement is achieved with a white gear and a white linear gear. IMG_0430.mov prior to folding, the planes are rotated vertical by these white 3D printed cogs. On further movement of the servo the linear cog runs out of teeth, and the white gear contain a magnet leaving the planes in the vertical parking position. Then a brass rod pushed in parallel with the white linear gear folds the planes via the brass gears previously described. A spring made from 0,5mm brass unfolds the gears as the servo retracts both rods until the linear gear re-engages. crazy complex, but it runs with little friction, uses less than 0,1 amps on a mini servo and uses no current when not moving. Very happy - will need to program an arduino to tie the servo movement to the “dive” and “surface” commands and the diveplanes signals from the transmitter. now hopefully back to some woodwork, apologies for the distraction!

Started work on the water-tight compartment inside the wooden hull. The hull will flood when dived, and the WTC will contain the motors, batteries, receiver, two-way pump and a 500ml divetank to dive the boat. The insides of the WTC can be accessed by a bayonet between the aft 50mm diameter PVC tube and the mid 75mm PVC tube. this is the WTC seen from the stern. The white stern endcap will have two shafts driven by a motor inside the 50mm PVC tube. The holes are where the prop shaft and the rudder and diveplanes seals will come. Between the two white parts of the endcap will come a gearbox to counter rotate the two shafts. Further forward is the bajonet, seen below when opened. two M4 bolts on the start section fit into the keyholes (strengthened with brass) in the front section and thus the two boat halves can be twisted shut. The bajonet contains O-rings to seal the 50 and 75mm PVC tubes. this picture shows the bajonet shut and the bolts holding the two halves together. And finally with the two deck sections attached. now that I know exactly where the WTC and its connections will come, I can start to work on the prop shafts and their outholders, the rudder and diveplanes and the holes in the wooden hull to flood the boat and enable air to escape.

You mean “surfacing after a dive” ! it will be a while….. for O-13, but here is one from my also wooden K-XVIII

The four parts of the hull have been sanded and I applied first layer of epoxy filler. The basic shape is now done. I want to get the various holes cut in the hull (for flooding, allowing air to escape, and the rudder, planes and prop holes before I epoxy again - so that all wood becomes sealed. thus I have to plan all “holes” and one of the difficult ones is for the front dive planes at the bow. These planes fold flush against the hull, and are covered by some plating outside the hull. so I started designing the functionality with the allowed space (tiny). This is the result: https://youtu.be/hi7XH0Wy4v0?feature=shared the two yellow gears rotate the common axle of the diveplanes. The axle has a hingepoint to fold each plane. Underneath each hingepoint is a brass gear that turns a pair of messing wires to apply sideways force on the axis outside of the hinges. This turns the diveplanes flush with the hull and back. The two brass gears are coupled through 4 more to work in tandem (could have worked with 2 but they would have had to be bigger and cause conflicts). The large yellow cog and one of the 6 brass cogs will be driven by a servo 4-8 inches further back in the water tight compartment of the boat (this gear assembly will be “in the wet”). This took me a good few hours of brasswork (of which I learnt a lot) and 3D design and printing (at which I’m also a beginner). I’d like to replace the printed parts with messing but I’m not at all sure I would achieve the precision on the boreholes to make it turn properly. I anyhow need to make another version of this as there is some play in the yellow cogs.