Big Gun PDN Mikasa

Discussion in 'Warship Builds' started by Kotori87, Nov 17, 2021.

  1. Kotori87

    Kotori87 Well-Known Member

    Joined:
    Nov 8, 2006
    Posts:
    3,522
    20220323_200523[1].jpg 20220323_201249[1].jpg
    Printing cannons, printing cannons! La la la!

    From left to right, a 2.5" twin .177" cannon, three layers deep, suitable for various cruisers. A 4" twin 7/32" cannon, one layer deep, perfect for that pre-dreadnought or dreadnoughts. A 5.5" quad 1/4" cannon, still being printed, for The French. A non-rotating tandem triple 7/32" cannon, for the stern of an Austro-Hungarian dreadnought. The WWCC's first battle of the season is coming up soon, so I am currently doing refits to my current lineup. New magazines in the same boats as before, with minor improvements throughout. Dished magazines to improve gravity feed when low on ammo. Better barrel alignment and gaskets, for improved efficiency. Many of the new designs incorporate a partially closed breech mechanism, further improving efficiency and eliminating double-feed issues. And of course a significant weight savings over traditional machined magazines, not that most ships are hurting for weight. I plan to publish printing files and Fusion360 files for these cannon magazines once they have been proven in battle. With the ease of manufacturing afforded by 3d printing, Big Gun cannons are easier than ever to make.
     
    Last edited: Mar 23, 2022
  2. Kotori87

    Kotori87 Well-Known Member

    Joined:
    Nov 8, 2006
    Posts:
    3,522
    Long time, no update. Being underwater will do that to you. Anyway, exciting new updates! First is some SCIENCE! I want to know how much room I need to set aside for lead, so I decided to sheet the prototype hull. Since this was only the prototype, I took the opportunity to try out a somewhat-new sheeting method. GregMcFadden has tested ironing with hot glue to sheet Fast Gun ships, which has proven quite effective. The heat from the iron rapidly soaks through the balsa, melting the glue and getting a fantastically watertight seal with little effort. So I decided to try the same thing with the Mikasa's 3/32" balsa skin. It both worked and didn't. The thicker balsa took much longer for the heat of the iron to soak through and melt the glue, which was somewhat inconvenient. More problematic, though, was the time required for the glue to cool down afterwards - it took several minutes, during which time the balsa consistently sprung off at least one corner. Eventually I had something functional and mostly watertight though, so I grabbed my scale and headed off to the bath- I mean SCIENCE room!
    20220815_210325.jpg 20220815_210256.jpg 20220815_210546.jpg 20220815_210554.jpg 20220815_210624.jpg
    This series of photos demonstrates the weight of the unballasted ship, the superstructure (so far), the ballasted ship after floating, and the final weight/location of lead added. I will add cutouts to the water channels for the correct amount of lead, to get the whole mass as low and centered in the hull as possible.

    OK that's very cool and all, but what about the Big Guns? This isn't any old Mikasa, this is a Big Gun Battleship Mikasa!
    20220818_220409.jpg 20220818_220434.jpg
    Here you can see a prototype bow cannon installed. I am not quite satisfied with the design, but it is very close to being finished. Well, for a given level of finished. I still haven't developed a satisfactory barrel depression system for this compact system yet, so I am debating whether to call it "good enough" or keep pushing it to be the ultimate in compact cannon technology.
    20220818_220611.jpg 20220818_220811.jpg
    Now lets break down how this thing works. The copper tee is the accumulator, somewhere around 3 or 4 cu.in of volume. Sticking out from the copper tee is the buna-ball valve. It's a pilot-actuated poppet valve designed for rapid opening and very high flow. You can see the buna-N rubber ball for which the valve is named. On the end of the buna-ball valve is the MPA-7 actuator. This opens the valve to fire it. Lastly, on top is the rotating magazine. It is held in place by several setscrews in the valve base, and rotates on a bearing in the magazine itself. More on that in a bit.

    20220818_220857.jpg 20220818_220906.jpg 20220818_220919.jpg
    The buna ball rides on a 1/8" brass rod. The rod is threaded on both ends with a #5-40 tap. Unlike the more common #4-40, the #5-40 can be threaded directly onto a 1/8" rod, so it is much easier to manufacture. On the actuator end, there is a nice stiff spring to re-close the valve. The MPA-7 actuator screws onto the end, ready to press on the rod and open the valve. This is one of the most difficult threads to get, since it is a Clippard special 15/32"-32 thread. Very unusual, and annoyingly, not something Fusion360 can innately model. I did recently figure out a way to model them manually in Fusion360 though, so once I do a few calibration prints for sizing I'll be able to print these threads directly onto the valve with no extra machining steps. You'll also notice a double O-ring seal. I sometimes had issues with earlier prototypes that used a single o-ring, but the double o-ring setup seems to be doing well on the current prototype.

    20220818_221436.jpg 20220818_221448.jpg
    And here's the magic of Big Gun. This is how the cannon actually rotates. For turret rotation, I am using the following parts:

    View: https://www.amazon.com/gp/product/B00ZVD22GG/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1


    View: https://www.amazon.com/gp/product/B07JKSGCMY/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1

    The bushing is drilled out to a 3/8" bore, then press-fit into the bearing. The bearing then is locked into the magazine, and the bushing fits into the cannon base. The hole on the magazine's bottom layer may require some tweaking based on the precision of your printer to ensure a good fit, but ultimately the bearing is sandwiched between a small lip on the bottom layer and the manifold layer. A similar setup can be used for larger cannons. It's worth pointing out that none of this setup is pressure-tight after the buna ball. Everything downstream from there only serves to direct the pulse of air towards the breech and down the barrels. While some sealing is important to improve efficiency, it doesn't need to continuously hold 150psig air. The manifold layer distributes the air pulse to the two barrels. The next layer holds the breech pistons, and then we get to the breech itself.

    Mikasa Cannon Cross-Section.jpg
    Lets use a cross-section to better understand what's going on here. The breech is a tee section. When loaded, one round rests on top of the piston, with later rounds in the ammo ramp waiting to get pulled down by gravity feed. When the cannon is fired, a pulse of air rushes up from the buna-ball valve. This high-speed air pulse, almost a shockwave, pushes on the piston and round. The piston slides up half-way into the breech, preventing double-feeding and directing most of the air towards the barrels. A small amount of air will leak out through the ammo ramps, but remember that these are mostly blocked by the other rounds waiting to be loaded. The round itself rides the air pulse up the barrel, around the bend, and then out the muzzle towards the target. Passing around the bend adds backspin to the ball, which probably has no effect on accuracy at normal engagement ranges but sure sounds fancy.

    Well that's about it for now. Time for sleep.
     
  3. Connor

    Connor Member

    Joined:
    May 30, 2018
    Posts:
    35
    Location:
    Merritt island Fl
    Legend lmao
     
  4. BigGunJeff

    BigGunJeff Well-Known Member

    Joined:
    Sep 26, 2014
    Posts:
    227
    Thanks for sharing. Very cool design. I especially like the piston that prevents multiple founds from firing
     
  5. TorpCruiser

    TorpCruiser Active Member

    Joined:
    May 16, 2022
    Posts:
    126
    Location:
    Everett, WA
    Looking good!
     
  6. TorpCruiser

    TorpCruiser Active Member

    Joined:
    May 16, 2022
    Posts:
    126
    Location:
    Everett, WA
    Looking fantastic! -Any plans for the underwater torpedo guns ? *side-eyes*
     
  7. Kotori87

    Kotori87 Well-Known Member

    Joined:
    Nov 8, 2006
    Posts:
    3,522
    Launching Ceremony for the battleship Mikasa

    View: https://youtu.be/sEkPTYBZejI

    Onboard battle video, sortie #2 (sole surviving camera from the event)

    View: https://youtu.be/U4hg3DlS0aY


    Well folks, the battleship Mikasa has been launched, and fought in her first battle. It was more successful than I expected, but less successful than I hoped. Here's a brief overview of the ship, and the results from the battle. Click the pictures, they get bigger!

    20230617_140333.jpg
    Assembled for combat, minus batteries. You can see how tight it is in there. 5oz CO2 bottle, dual 3000mAh LiPo batteries, two drive motors, and a brushless pump. Note that the brushless motor is not actually required for pumping capacity, but for reducing the height of the pump. It needs to fit underneath the regulator, after all.

    20230617_140352.jpg
    Closeup on the machinery space. Drive motors are RS365SH, already combat-tested in the battleship Teggethoff. Dubro fuel-line coupling to 1/8" brass propshafts. Standard-sized servos for turret rotation and firing. SMAVs for firing and a SMTV for safety switch. Forward is armed, aft is safe. The battery boxes are copied from my two Deutschland-class predreads, where they proved easy to work with. Also of note is the lead shot ballast. This is not permanently installed. Rather, I put down some saran wrap first, added the necessary mass of lead, then poured epoxy over the whole mix. The result is two custom-fitted lead bricks as low in the ship as possible, concentrated amidships for optimum turning and as outboard as possible for maximum stability.

    20230617_140745.jpg
    Front cannon, rotation servo, and recovery float. The float is a fairly standard design, using the forward-most part of the deck as a breakaway part with lots of buoyancy. No small bobber stuffed into a funnel and forgotten until it fails. And if the ship sinks stern-first, escaping air helps force the float out. Proven successful at the battle.

    20230617_140853.jpg
    And here is the aft end. It looks pretty open, but that's only because there isn't enough space for anything else back there.

    OK, now lets look at the battle damage and lessons learned:
    20230617_134443.jpg 20230617_134525.jpg
    First off, the superstructure took an absolute hammering. It felt like every shot someone fired in Mikasa's general direction resulted in pieces flying off the boat. This was almost entirely due to the superstructure being printed with a bad batch of PETG. I used up half the spool trying different settings, but the plastic always came out weak and brittle. I am told this is NOT normal for PETG, but after burning half a spool with tests and not getting any better, I said "screw it" and just printed everything. It did hold up better than balsa wood, but not as well as lexan-plated plywood would have. Note how the armored conning tower, in white PETG, was easily penetrated by at least one heavy-caliber round. The flying bridge above it, printed in my usual grey ABS two days before the battle, resisted multiple impacts without failing. I believe I will be re-printing the entire superstructure in "the usual stuff" and giving it a thorough acetone vapor bath for strength. I am quite happy with the performance of the hull-level casemates, though. These are the add-after-sheeting type, printed in ABS and given a 1/8" brass rod to affix them to the hull. The hull itself has 1/8" holes to receive them. They definitely helped, as you can tell by the multiple dents they took.

    20230617_134230.jpg 20230617_134306.jpg
    Port side: 25 aboves, 4 belows. Stbd side: 29 aboves, 1 one on, and 3 belows. If you're curious whether armor matters, the answer is YES. In addition to the 54/1/7 hits, Mikasa also had at least 82/0/16 dents of various levels. I also counted at least 52 unique superstructure hits. No idea how many hits it actually took, as many of the damaged sections were blown away by later hits.
    20230617_140521.jpg 20230617_140613.jpg
    That's funny, the damage doesn't look so bad from in here!
     

    Attached Files:

    diceman42, notSoGnarly and WillCover like this.
  8. Kotori87

    Kotori87 Well-Known Member

    Joined:
    Nov 8, 2006
    Posts:
    3,522
    After Action Report and Lessons Learned
    What worked. The ship proved incredibly stable and maneuverable. Electronics were reliable, and did not have any failures during the entire event. The balsa was effective, both protective and penetrable. The paint job was surprisingly good. I've heard from other battlers that black paint hides damage when patching, but I found this was not the case. This may be due to the red primer underneath the black paint, which highlighted every hit with a thin ring of red.

    And now for the longer list. What wasn't great, and how to do better.
    First, damage control was inadequate. Specifically, the bilge pump was unable to ingest sufficient water. Bathtub tests prior to the event showed a solid 1/4" stream of water jetting all the way to my bathroom ceiling, and splattering off the ceiling with decent force. Once installed in the ship, however, the pump cavitated heavily. Bathtub tests after the fact, involving the entire ship to validate installation, showed that the structure of the water channel only allows the pump to suck from a small gap between the motors, which is unable to flow enough water to meet demand. Raising the pump helped somewhat, but not enough. Ultimately resolving the issue will require modifying the water channel in that area, widening it so the pump can suck from three sides.
    Second, I observed a bow-first sinking tendency, which is undesirable in this ship. I believe this is due to the front cannon acting as a barrier in the forward water channel, preventing water from flowing aft. Since the front cannon cannot be relocated, my only solution is to modify the forward water channel to fill in the areas that are currently floodable. This will be updated in the CAD model, and I may be able to add lightweight foam in the Mikasa herself to accomplish the same task. I can also design in an attachment point for the recovery float string.
    Third, cannons. The cannons themselves suffered multiple minor issues throughout the weekend. First, the forward turret barrels, affixed by clear PVC hose and zip ties, began to push themselves free as the cannon fired. You can see this in the combat video, as I had to return to port several times to push the barrels back down before resuming the fight. This was significantly reduced on Sunday by replacing the zip ties with hose clamps, although a more long-term solution is to swap out a different hose with better grip.
    The other problem the cannons had was a catastrophic failure of the rotation bearing joint. I knew this was a possible failure point, since I had thinned it out a lot during the early design phase in an attempt to save height. My mistake was in leaving it thin even after I'd settled the rest of the design, and knew how much room I'd have to reinforce it. I will be updating the CAD model with a thicker rotation bearing joint, and updating the cannons in Mikasa as well.
    Speaking of being too thin, the magazine lid was also proven too thin. When the depression servos were cranked to full depression, the magazine lids would flex quite distressingly. Although neither lid failed, nor was close to actually failing, this flex adds a level of unpredictability to the depression system that can easily be resolved just by thickening the lid from 1mm to 2mm.
    Another modification I'll be making to the magazine lid is a holder for the turret covers. I tried to have the turret covers ride on the barrels, with a plastic bushing on top of the barbettes. This proved unreliable, sticking at inopportune moments and flopping over. I didn't actually miss many shots because of it, but that was more because I was missing all of my shots due to user-interface issues. By affixing the turret covers to the magazine lid, I completely separate them from the barbettes, eliminating any sticking. This will be a return to Big Gun standard practice. The cannons are centered in the barbettes with CAD precision, so I don't have the usual concern of cannon rotation and barbette center not lining up.
    Speaking of user interface, that was the single biggest issue I struggled with. On Saturday, I was trying to use a cheap FlySky radio for positional turret control. This proved to be a terrible mistake, since I had to move my hand to rotate the turrets, then move my hand again to adjust depression, then move my hand a third time to actually fire. This was horribly inefficient, and I missed a LOT of opportunity shots. Overnight I swapped radios to my FrSky Taranis, which had some very convenient sliders. No more moving hands to aim, elevate, and shoot. I did much better, but I still had some issues. The turret rotation was still not fast enough for a proper position-control system. Worse, the forward rotation servo had somewhat slowed down, throwing off my aim even more. I think, going forward, that I will switch back to a velocity control system. This will allow me to switch back to the much cheaper Flysky radio, simplify the user interface, and eliminate any dependence on expensive and hard-to-find multi-turn servos. Continuous-rotation servos are much more common, and easy to waterproof.
    My final problem was propulsion. Simply put, the propulsion was too powerful, so I had to end-point-adjust the throttle down to about 40-45%. I say "about", because at such a low throttle, tiny changes had a big impact. At 45% throttle, the ship ran 28 knots. At 40% throttle, the ship ran 20 knots. At 42 to 43%, it would sometimes run slow and sometimes run fast, and the motors had so little torque that they'd run at visibly different speeds. During battle, I ended up just being careful to catch the ships I'm supposed to catch, and not catching the ships I'm not supposed to catch. Not a good solution though. What I really need is some way to run at, say, 85-90% throttle for 25 knots, so my motors are getting the power they need. 2:1 or 1.5:2 reduction gears would be the most ideal solution, but the ship was designed for direct drive and it's far too late to change that. So instead, I'd like to do the next best thing. Flatter pitch props. Now I just need to model the props, print several different pitches in resin, and try them all. Once I find a satisfactory solution, I can order proper bronze versions from Shapeways. or maybe not, modern high-strength resins are quite strong and Big Gun doesn't put nearly the stress on the drive train that Fast Gun does.
    Other than the aforementioned corrections, I have a few more updates to perform on the hull before building the next one. I need to fillet in all of the sharp corners for strength. Then the SMAV mounts and depression servo cable controls need to be incorporated into the hull proper. After that, it's off to the printer for the next hull. This next one is going to be HMS Majestic. Virtually identical hull form, within 5% on all major dimensions, and beautiful side-by-side funnels.
     
    diceman42 likes this.
  9. Kotori87

    Kotori87 Well-Known Member

    Joined:
    Nov 8, 2006
    Posts:
    3,522
    A little more discussion of the user interface issues. The first radio I used on Saturday was the FlySky 6-channel radio:
    rm2327-9-18a7.jpg
    I replaced the two plastic dial caps with collars and extended setscrews, similar to other position-control setups I've used, so I could feel the precise position with my finger. In this case, the right dial was for rotation, the left dial was for depression, and the right stick (up and down) was for firing. You can see the problem. Three different hand placements required for every attack. Not ideal.
    By Sunday, I'd swapped to the Taranis:
    frsky-x9d-plus.jpg
    This one has a pair of dials in similar locations to the FlySky, but it ALSO has a pair of sliders. You can see the right slider on the right outer edge of the radio, below the switches. Very convenient placement for my index finger, when my thumb is on the stick. The left slider is the same. Needless to say, these were my preferred method of control on Sunday. It was a significant improvement, but it still wasn't perfect. The slider allows very fast response, but I cannot feel the precise end position of the turret. For me to use this setup effectively, the turret rotation has to be very fast, essentially mirroring my movements as quickly as I can make them. With a slower turret rotation, I could neither feel the final position nor see the final position of my guns, and thus I was often fumbling my aim.
    Ultimately I will be returning to a velocity-control system with the FlySky radio, with all major controls on the right stick. Stick right, rotate right. Stick left, rotate left. Fire front, fire back. Elevation will be a three-position switch on the left, giving me three options for range. I could potentially modify the FlySky by replacing the right stick with a dial and buttons, but that's a lot of work that just isn't really worth it. Velocity control is a simple, intuitive system for anybody to figure out.
     
    Seventy Two likes this.
  10. JustinScott

    JustinScott Well-Known Member

    Joined:
    Oct 20, 2006
    Posts:
    2,203
    Location:
    Dallas
    Good lessons there!

    The cannon user experience is challenging. @Huru MorDae and I hit the lake yesterday & did a little bit of dog fighting; which was incredibly useful for me.

    Cannon rotate will take ME a lot of practice to learn. I’m happy with the positional system overall; but it does still feel a little awkward. Never having one before; I’m sure any canon rotate system would feel awkward

    I’m curious… Forgetting how difficult something might be to build, do you have any thoughts on what you would expect an ideal UX to feel like?
     
    diceman42 likes this.
  11. Kotori87

    Kotori87 Well-Known Member

    Joined:
    Nov 8, 2006
    Posts:
    3,522
    Battleship Mikasa is getting put back together for sea trials tomorrow. Features to test:
    1) New resin-printed props. These are a shallower pitch than the brass props, so I should be able to reach rated speed at a higher throttle%. This should also result in a more consistent speed. If they prove successful, I will send them to shapeways for manufacturing in brass.
    2) Improved pump sump. This should resolve the low flow/cavitation issues I was having before. Tomorrow is non-combat testing, so I don't expect to pump much though.
    3) New velocity control rotation system. Basically swapped the sail winch servos with continuous rotation servos, but I've also got a clever mix going on. I'll try to post a short video about it in operation.
     
    TorpCruiser and diceman42 like this.
  12. darkapollo

    darkapollo Well-Known Member

    Joined:
    Mar 26, 2008
    Posts:
    809
    Location:
    Carlisle PA
    So how I set up the flysky…

    Right stick is elevation and depression and rotation
    I replaced the flip switches with push buttons mounted to the back of the TX.

    Just a thought.
     
  13. Kotori87

    Kotori87 Well-Known Member

    Joined:
    Nov 8, 2006
    Posts:
    3,522

    View: https://youtu.be/ITMEVLFFsRA


    Didn't get good speed data from the radar gun due to the heavy seas. What little I could tell was that I'm a little bit closer, but still have a ways to go. Gonna get some further data on flatter water, then make some new test props. Pump sump worked well, although I didn't have any battle damage to really test it.
     
  14. TorpCruiser

    TorpCruiser Active Member

    Joined:
    May 16, 2022
    Posts:
    126
    Location:
    Everett, WA
    That's SWEEEEETNESS! Well done!
     
    diceman42 likes this.
  15. JustinScott

    JustinScott Well-Known Member

    Joined:
    Oct 20, 2006
    Posts:
    2,203
    Location:
    Dallas

    Ok so tell us about this? Looks like you are using a switch to change between the different modes?
     
  16. Kotori87

    Kotori87 Well-Known Member

    Joined:
    Nov 8, 2006
    Posts:
    3,522
    It's all on the right stick, just two different mixes. Forward and Aft turret rotation is plugged into channels 1 and 2. Left/Right mixes channel 2 with negative channel 1, resulting in convergence towards one side or the other. Forward/back mixes channel 1 with channel 2, resulting in traverse CW/CCW. I then (still need to install) buttons to replace one of the switches, for firing.
     
    TorpCruiser likes this.
  17. JustinScott

    JustinScott Well-Known Member

    Joined:
    Oct 20, 2006
    Posts:
    2,203
    Location:
    Dallas
    i think you should post a video…
     
    diceman42 likes this.
  18. TorpCruiser

    TorpCruiser Active Member

    Joined:
    May 16, 2022
    Posts:
    126
    Location:
    Everett, WA
    Kinda reminds me of how people do the mixing for ruddervators on RC V-tail aircraft. Repurposed for turrets! Love it!
     
  19. Kotori87

    Kotori87 Well-Known Member

    Joined:
    Nov 8, 2006
    Posts:
    3,522
    TorpCruiser likes this.