WW2 mini sub build

I have now cleaned up the 2 halves, using one of the handles that takes a hacksaw blade, using a full size hacksaw I found I couldn't get close to the line, but a blade in a handle is OK, though the blade does bend if it snags as youre pushing.


I held the 2 halves together using masking tape, though I know that some suggest putting them back into the mould to hold the halves together. I then used a lot of resin internally followed by pre-wetted 50mm woven tape which was very fiddly right up in the bow and stern areas. I did do the taping in several lengths especially at the ends, to make things a bit easier, but in general not too bad.


When I took of the masking tape I wasn't entierly supprised to find voids in the joints in places as I knew the 2 halves didn't have that good a fit, not sure if that was because the hull haves had warped or just poor cutting to the line.

I wanted to fit the prop & shaft before making the stern fins to ensure that the rudder & 'plane hinges cleared the prop, but I had decide to make a stern bearing rather than use a commercial prop shaft assembly, as I wanted to give plenty of bearing surface and I didn't need the shaft sealing and wasn't sure about the brass-resin bonding. The bearing was based on an M12 x 30mm bolt bored for a 5mm shaft (I used the shaft from a standard prop shaft, just not the outer). I machined 4 countersink dimples on the threaded part which when embedded in resin should stop the bearing rotating, even if the fixing becomes loose.  I had to ease the bore a bit as well, as either the shaft was a tad oversize or the drill a tad under.


The shaft is a standard 6" x M5 with a 70mm dia 4 bladed prop, about the biggest I could get. Eventually it will need a collar on the inside to take the reversing thrust.
I fixed the bearing to the hull using resin loaded with glass micro-balloons (basically look like a white powder) added after the resin & catalyst were mixed, with enough added to make a thicker paste that would stay in place a bit better.
I used 2 batches, the first to fill in the gaps between the hull halves (on the outside) and the 2nd to 'pour' into the hull to retain the bearing. I did use masking tape again to hold the bearing in place, with an inner block to keep the shaft straight, but it leaked so I had to hold the bearing by hand with the hull on end untill it cured sufficiently to stay in place - tedious and a bit messy.
Once rubbery I could cut off most the excess with a sharp craft knife, including freeing the shaft from the bearing that had stuck due to the leakage.
Mixing small batches of resin (50ml & 25ml) even with a syringe to add catalyst seems a bit hit and miss, but I had no problems with the resin not curing using 1% (by volume) catalyst in an unheated garage.

I'm quite pleased with the way the resin has filled the end of the hull, though where I poured it it is a bit obvious.


It's now starting to look like something nautical!

I might add an inboard shaft bearing for additional support, and if I was doing it again I would think about recssing the bearing into the hull as it looks a bit ugly due to the gap between the end of the hull and the actual propeller. I could have used the supplied outer tube, but it is so smooth I was woried about it bonding and there wasn't enough material to go machining lumps out of it. Maybe solder a tappered perforated metal plate next time as it would obviate the need for an additional inboard bearing.

The next job is to tidyup the outer before starting on the fins.

I now have what looks like an X-class hull, at least from a distance!
The middle 'lid' section was much like the two hulll halves, again required quite a lot of trimming, a hack-saw-blade-in-a-handle thingy from B&Q helped here.



The trouble is that after trimming the joint is far from perfect. I used masking tape along the edges to mark the horizontal mating line, but the 'lid' is now slightly narrower than the hull, and also taller, I can't trim down the height as it would make the width mismatch worse, so I'm stuck with what I have I think, bit of a shame. I'm not sure why, the obvious reason is that I cut the hull plug into hull and lid before adding the surface skim, so it is possible that the hull ended up with more surface skim (wider) that the lid. I don't think the moulds warped as they seem fairly rigid (2 layers CSM) nor the hull warped (3 layesr CSM) but the end result will have to do.
Not that obvious is that there was an air void under the gelcoat in the mould so I need some filler just on the break of the upper hull, and I hadn't noticed that some stray fibres form the hull haves had wafted into the lid mould so are visible on the surface at the moment. It would have been better to have finished the hull halves before waxing the lid mould as I could have cleaned the mould then. As it was I was waxing all 3 moulds at the same time (many times!)

You can also see the rough blank for the upper rudder fin, these are all ply and I intend to reinforce the joint with brass pins and resin/filler paste inside the hull. They are way over scale in terms of thickness but as a first build I am worried about them breaking off if I go too thin.

Obviously once finished I will need to prime/paint the entire hull, I was hoping that only the fins would need it but there is too much surface 'texture' & filler paste for that.

The rudder & 'plane hinges will be short brass tubes for a 2mm steel rod acting as the hinge. These tubes will have 'U' shaped brass strips soldered to then that can be fixed to the ply fins.

I'm thinking of coating the ply fins initially in a coat of thinned resin (acetone?) to allow resin to soak into the ply, anybody tried that?

I also need to add some small stops fixed to the hull inner to support the lid and provide some degree of locking in place, then on to the WTC supports.

Trim the hatch for height both fore and aft. there are ways of fixing the mismatch in width.

first, apply heat (hair dryer or heat gun) until the material is pliable, then use formers to change its shape... a few wooden bulkheads inside the hatch should work. you will probably have to bend it past the proper shape, it will spring back a little bit after it cools off.

second, put an inner lip on the hull, forcing the hatch to match the shape of the hull when it is on the hull.

Thanks to Tim for advice on getting the hatch to fit better, and it does!
I ended up using a grinding wheel to trim of the GRP hach, there wasn't enough to use a saw, and using a file would be a long job, but going easy with a grinding wheel sorted the height, followed by use of a hairdryer (our spare, of course!) with 2 plywoord templates for expanding the width.
It took a while to get the GRP to any warmth and I didn't want to overheat, but it seemed that with a small hairdryer the problem was too little heat, as the hatch never got much more than hand hot.
The templates were basically rectangular sheets that fitted horizonatlly inside the hatch with the idea to heat on the curved section not the rectangular upper casing. These could be forced into place as there was enough spring in the GRP, then held in place whilst applying heat.
I had to make 2 sets of bulkheads as the first didn't fit well enough even for me.

I also made the WTC bulkheads, this time from two laminations of 2mm plastic sheet cut to shape and then bonded using plasweld solvent based glue. The advantage of laminating is that the 2mm flexes so can be trial fitted even when oversize so can be gradually fitted to the inside of the hull.
The resulstant 4mm bulkheads seem rigid and strong enough.

The issue now is that when I trial fitted with the actual WTC the prop shafts don't align well enough. Some mismatch is OK as I need a middle section so UJs will take up any small misalignment, but I need to lower the WTC by about 15mm, due to that miscalculation.



I did try as an experiment using plasweld to glue a section of the plastic sheet to a dummy piece of GRP, but possibly due to the gaps (inside of hull is very uneven) it didn't work, so its some kind of epoxy glue to bond the bulkheads in place.

Dropping the WTC reduces that all important meta-centric height. The offset of drive shaft and prop shaft can be corrected several ways. A simple dog-bone connector is one way.

It starting to look much better, and now has fins and the 'pintles' for the rudder and 'plane.


The thin rods will be the hinge pins, with the rudder & plane fixed to similar brass tubes rotating on the steel pins. I fitted the pins to make sure the
pintles aligned as the epoxy cured (JB Weld).
Looking from the stern the hinge shafts are square so all good so far.

The battery support foam is also fitted in the bows:


I've strarted marking out for the various deck fittings and hoping I can disguise the fixings for the main 'lid' underneath them. The intentions is to have some brass rods going into holes in the ply bulkheads that can slide fwd/aft to fix & release, but still working on that. I did think about magnets, but I want something a bit more robust, if only just in case the inner bouyancy bag gets loose.

This time its the rudder & 'plane, both made from 1mm brass sheet with 2mm brass tube as the hinge.
I did think about simply soldering the tube to one side to a simple cut-out, but decided to try something a bit more symetrical, with the seperate sections for the parts fore & aft of the hinge.
Cutting out was no problem, the individual parts shown below.




However, when I tried to dry fit to the hull it became obvious that the fins were not as symetrical as I thought, the croosing point is not in the middle of the hinge tubes. For the rudder this is not too bad, but the 'plnes is more awkward. I had to tweak the balancing sections in front of the hinge os that they would not foul each other in use. If you look you can see there are not symetrical.
I'm hoping that there will not be too much rolling force from their opperation as the 'cetre of area' will no longer be in the middle.

I had trouble with the blowlamp flaming out as well so the first attempt was a disaster, I think it got too hot and the flux burned so the solder didn't flow. I had to break them apart, clean and re-do.
The end result is OK though the plates are not exactly centred on the hinge tube, but worth the extra work.

Once cleaned and fitted they seem to work freely.


Whilst I was soldering I also made the bow towing tube from brass tube and the rim made from copper wire that will be fitted into a hole in the bow.

I did wonder if soft solder is going to be strong enough, but I decide to use it anyway as its so much easier, and the joint is quite long. If it fails I can always replace with a silver soldered version or even try some kind of resin moulding.

They also need a bit more finishing to get a better shape, and also fitting the actuation 'horns' before they can be primed and painted. I think I will use brass sheet for the horns and epoxy glue to fix them, but I might look at what is available for model aircraft. I also need to get an M2 die to thread the blank ends of the steel hinge pins, but that will also help with the actuation linkange as I intend to use the same 2mm steel rod.

I also need to sort out the end of the fins, to try and smooth them out they are too lumpy at the moment.

Now reached a major point I think of fitting the WTC into the hull and linking up to the fins :)
The fin control rods are roughly horizontal so use a swinging arm/intermediate shaft arrangement to link to the WTC controls.
I depart from true scale in a number of ways, the intermediate shaft is above the propshaft, ideally it would have been level, but its easier that way. Also the rudder control is to low, it penetrates through the pressure hull not the rear deck casing, again to make life easier. The rudder shaft arm uses the intermediate 'plane shaft as the bearing for simplicity. Hope the phots are clearer!






The photo actualy have the 'plane crontrol rod missing as I had removed it for adjustment.
The rod ends have M2 clevis' to fit to link to the arms/control horns, but the M2 threads are very varaible and seem a bit tapered so I may have to re-do some of the rods (all 2mm steel).
The intermediate shaft runs in short lengths of brass tube epoxied to the hull to provide some bearings. I thought of making the port side a blind bearing so its not vissible, but for simplicity and to ease assembly its a through hole.

On firing up the radio control I found both servos are reversed, but not a problem as I can swap the servo arms by 180deg in the WTC or invert in s/w, probably the physical change.
I also need a few collars on the intermediate shaft to keep everything aligned.

Also progress on some of the deck details, though I have not cut the limber holes etc. a few photos to follow.

All in all feeling pleased with the results so far.

This shows the rudder controls, and how it exits the hull rather than the deck casing.


the jig I used to create the various brass rod fittings


general view. Masking tape is for position marking for the various fittings.


On reflection the bow end of the 'scope shroud is too vertical, it should have had more of a slant on it I think, I assume to shed any wires/nets it might entangle with.

Latest task is the catch mechanism for the hatch. I wanted something better than magnets, since if the dive bladder comes a bit loose it could try and lift the boat via the hatch, so I wanted a positive fixing.
Also, since I made the joint vertical it had to have 2 such fixings, as the hatch can't be hinged into place.
It the end I made 2 sprung loaded 'bolts' from 4mm brass rod that slides in 2 ply blocks and locks into a 3rd ply block. These are accessed via small hatches in the top casing, these will be held in place using magnets.
The springs I made from 0.7mm piano wire as I remembered my lathe has crew-cutting, not that I've ever used it for that, so made a set of springs with 3mm pitch wound on a 4mm rod. Took ages to set up up and about 30sec to make the springs!

by making the springs the same length as the gap between the 2 'fixed' blocks all I need is a single pin through the rod to act as the stop and as the thrust washer for the spring. In fact I used 3 washers so that the spring ends do not act directly on the ply blocks.
Originally I intended to use threaded rod so I was going to line the ply block holse with brass tube to take any wear, but getting all 3 holes to line  up was an absolute sod (no access when closed up) so the last hole (the catch plate) was drilled over size so I could pack the hole with small wood slivers (actually matchsticks) to get it lined up then all tubes fixed into the ply blocks with epoxy glue, with the bolt in place, and the hatch held to the main hull with cable ties whilst everything cured.
Some epoxy obviously oozed out as one had to be wiggled to free it up after it was all cured, but not too bad.

This shows the stern hatch latch from the underside of the hatch, with the access hatch vissible as the white rectangle. The spring pushes the rod to the left as far as the stop will allow, with about 15mm of the rod protuding to the left (into the main hull block).


The bow latch is similar, but the bolt slides in the main hull with the catch plate in the hatch.

As you can see there are various voids in the epoxy so I need to go round all these gaps and fill with resin/filler paste, and I think I will then coat the ply blocks with resin as well to seal.

After thinking a bit I decided I didn't like the way the rudder control rod exited via the 'pressure hull', historical accuracy is one thing but bad engineering is another!
I decide to add a 2nd shaft above the original with a new vertical link from a new bell crank, the the action now is through the rotating upper shaft.
Still not historically accurate, but at least it looks plasuible. I also added a guard to look a bit better in an attempt to match the 'plane guard on the other side.
It looks a lot better now, and when the hole is plugged and the hull painted who will ever know?

One thing I 've mulling over is the air intake tube (I think) that hinges on the side of the fairing. Originally I intended to simply fix it vertical using some wires to the bow & stern fairleads that could be unhooked to fold down. Hence the small flanges at the top. Eventually I decied some kind of spring catch would be better, so since my lathe was still setup for spring winding I made another. I pondered some kind of slot in the pivot tube, but in the end I decided on some historically inaccurate latch plates on the outside of the pivot plate.

The inside view shows the spring and a temporary retaining wire, and shows that some tidying up is needed to make the spring sit flat. This is actually due some of the expoxy resin glue fixing the pivot tube to the hull moulding.


I've also finished the adapters for the bladder hose and managed to use the WTC air pump to inflate the bladder, it takes ages but does work.

The reason for the 3-way hose connector is to provide a easy access branch, normally fitted with a bung, but it allows the baldder to be inflated at the start of a run, with ambient air pressure in the WTC. This means the WTC goes from ambient (surface) to higher pressure (dive) so hopefully minimise any water ingress.

Next step is probably secure the ends of the WTC and try it in the bath to see how much ballast I need to get the correct waterline etc. I get the impression that the bladder may be too small (220ml in volume), but it is a child size blood pressure cuff, so I can go up a size if needed. I intend to secure the bladder round the WTC which would allow me to fix position to get the correct trim.

I need to smooth of the filler I added round the wood blocks on the hull/hatch joints and round the fin/hull joints and then prime and paint, but its getting there.

well it hit the (bath) water, but only just, I had to take of  the rudder for it to fit & float


Initially it had a list and was floating high so I ended up adding about 100g to the stern and 50g in the bow and extra bouyancy (building insulating foam sheets) under the top of he huill. I assume the CoG is so near the CoB its a bit unstable. Adding the hatch on the top made it worse due to its weight, so I added more foam under the hatch at both ends.
Eventually it floated about right with the bladder inflated.


You can see the additional block under the stbd side wire frame to correct the list.

As suspected the bladder does not have enough volume to allow the boat to submerge, so a larger cuff is on order, but it may requuire a bit more ballast to equalise.

The next step is to varnish the plywood blocks to make a bit more waterproof and to cut some foam blocks to fit properly, I think a thin strip down both sides of the lower hull along the joint should be OK and under the hatch at both ends. I may add more foam than actually needed, but as I have some iron shot on order for ballast that can go right at the bottom and could be removed for transport anyway.

The other oops was I didn't check the seals to the WTC, and the battery feed-thro' had come a bit loose so it leaked a bit, but nothing too serious. I'm tempted to add the ability to presurise the WTC and check for leaks with a pressure gauge as a poolside check. I'm considering a car tyre type valve and cap but when submerged could the excess water pressure then enter via the air valve? Any one else does that?

a very common way to check for leaks is to have a brass tube approximately 1/8" diameter penetrating one of the endcaps with a flexible hose (12" long or so) connected to the brass tube, and a plug in the other end of the hose.  remove the plug, blow in the hose, look for bubbles.  the plug can be anything that seals the end of the hose.

A bit further, managed to prime & paint the hull, using standard acid primer and the then standard black (both Halfords standard range, so nothing special)
It seems to cover and look OK so far....




I used less than a single can of each so not too bad I think, and I also painted the rudder, 'plane, access hatches and the snort tube.
I blocked all the various bearings & control shaft holes, periscope tubes etc. simply with small peices of rolled up kitchen towel.

You can also see the additional bouyancy I've added along the inside of the (lower) hull to try and improve the stability.

Another mod was to make another vent bung (intended to allow the WTC to vent without a drastic dismatling) with a tube to which I could attache a balloon to provide a quick pressure test:


This used a balloon to create some pressure and a T-peice to allow it to be inflated once connected to the WTC. Result was it stayed inflated for about 24 hours, so no leaks, especially when repeated without the WTC I got the same result. I'm taking that to mean that he residual leakage is probably the porosity of the balloon itself, but it provides a quick pressure test in-situ.

However what should have been a quick software rebuild to invert both control servos has turned into a bit of a nightmare. The receiver software now simply keeps reseting every few secs.
The big mistake was to update all the libraries within the Arduino IDE without thinking, and some have changed a lot since I last built the code, about a year. I archived my code, but I forgot to note down the revisions of the library files, and the PC backup from way back when didn't capture those.
As a result I'm not sure if the cause is interference from the various relays/pump motor, so other loose joint or a software bug caused by incompatible libraries. The results is I'm going to have to strip the WTC inards out and check it. I supose it does mean I can replace the slightly bent servo control rods!
The one thing I find annoying about Arduino is not having a hex file I can simply donwload without having to re-complie the code, which is what I've always been able to do professionaly at work.
All annoying as it was working fine apart from the inverted control action. I probably could have swaped the X-Y joystick connections, but I though I'd do it properly.
Hey Ho.

I didn't realize that you built an arduino based receiver... is it your own design? what are the specs?

Yes, Simon has done a lot. (Shown in the 433/458 MHz thread.):

https://www.theassociationofmodelsubmariners.com/t83p225-uhf-radio-control-for-submarines-openlrs-lora-fsk-etc-on-458mhz-and-433mhz#11644

His software is here:

https://github.com/SHopper-prog/LoRaRadioControl

Just in case Simon doesn't update us on here.

Simon, you raise a very interesting (and disturbing!)  point about automatic library updates.

I suspect my software will have the same problems.

Any solutions to "freeze the software in time" ??

On a different topic could I recommend using a white putty filler to fill depressions and pinholes DURING the painting process. It sticks even between paint layers and you use it straight from the tube - no hardener etc.

The stuff I use is White Putty by Squadron Products in a tube. (A quick Google shows a lot of stockists say DISCONTINUED though. I see it contains Toluene which may explain it. Reading the safety leaflets could make your toes curl ! )

thanks for the tips on finishing the hull outer surface, I think the root cause is the seperation of the master plug from the mold, that and impatience to get it working!
As a recap, the radio is based on the DRF1278 LoRa module from Stuarts projects (Tindie), in turn using the SX1278 chip. Looking at the web it looks as if that adapter is no longer available- what a pain.
I never managed to get straight FSK working without Lora spreading so I run with the minimum spreading that takes about 2.5 RC channels, but since theres not much traffic and multiple channels should co-exist I'm happy with that. I can set the centre channel with a small rotary switch ion the master & slave, and the message format is 4 x analogue + 8 x digital (ON/OFF) channels plus 8 status bits/alarms back from the slave.

I use a PWM chip to drive up to 16 outputs based on the PCA9685 chip as this ofloads the PWM driving from the arduino.

The big trouble I had with the libraries was the IDE suggested library uploads that I accpeted without thinking, and once done I had no way of backtracking or finding out what the previous version was.
The main issue was that the LoRa library (RadioLib from Jan Gromes) is not only now incompatible (some functions now have additional parameters, which would just require some edits) but much bigger, so would not fit into the arduino memory with my other code. so, as a suggestion I would:
1. back up library code as well as source code (on my PC my code is on the D drive, but the libraries are on the C drive with the IDE code)
2. Take a note of what revisions the libraries are (each library has a file some thing like 'library.properties' that gives the revision)
3. look at each update, and see if you need the update

In theory the IDE is supposed to be able to handle with multiple revisions of libraries, but not with me, so I have a folder with the different download files, but only unpack one revision to the library folder at any time.
As I mentioned there doesnot seem to be a way of generating a downloadable hex file, so you always have to re-compile every time.

The RadioLib library is far more capable that I need, so I'm using v4.4.0 from April 2021 at the moment and it seems to be OK, so its back to the main build!

I was hoping to be at Papplwick as its not far from me, but it looks as if its not going to be ready, shame.

Sorry to hear that you are having so many problems with your home grown radios, but fingers crossed that you get it all figured out.

one suggestion vis-a-vis the hex/bin file:

turn on all the debugging in the arduino ide you can, and just do a "build" and follow the debugging output to see where temp files get stored.

then, do the same thing with a "build and run", to see what cli tool it is using to upload the hex/bin file to the mcu.


that should both give you a final bin file for uploading, and the tool and proper syntax for uploading it yourself.

I used to have to keep a copy of arduino for each sketch as a lot of the projects i was working on had modified libraries. I recently moved to arduino 2.x and it broke a lot of old code so i now keep a 1.x install and a 2.x install.