This gun was inspired by Rochester Air Show in 2002, where I pretended to be a stupid foreign tourist not speaking any English, and climbed through the fence to to touch and photograph 30mm 7 barrel rotary cannon of A-10 put on static display in live:
(Yes, that is my finger there)
I already designed earlier a joystick-controlled nose gun turret with non-working 4 barrel rotary cannon (see here) for one of my MOC helicopters:
At that time I was pretty convinced that Lego Technic is not suitable to build it with working gun.
188.8.131.52.Basic concept of current model
Finally, it was possible to design a working rotary gun, but only with strong compromises. In any Gatling-system guns, there are separate bolts for each barrel moved by a helical groove in gun's rotor, and they are locked to breeches during firing. You can see an excellent animation about its working here:
It would be very hard to build it from Lego Technic in reasonably small scale (1:18..1:20 Lego Technic Figure scale), as weak point of Technic is the lack of rods/tubes/barrels with multiple diameters, which can be build together coaxially, as Gatling-system heavily relies on these type of components.
Moreover, in Lego Technic, we cannot use chemical propellant to launch projectiles, as ABS material would not tolerate high pressure and temperature. We have to simulate firing with mechanical launching projectiles.
So I decided to build the rotary gun with simplified operating system compared to Gatling: single bolt with delayed blowback by toggle, where toggle is driven electrically by chain. This is a mixture of classic recoil operated, toggle-delayed blowback system (e.g. Schwarzlouse machinegun of WWI) and the chain gun system (e.g. M42). Don't confuse chain gun with Gatling, which is erroneously referenced as "chain gun" in most FPS computer games. The real chain gun has one bolt controlled by an electrically driven chain running rectangular cycles on 4 sprockets.
(Note 1: If you do not understand technical details of operating systems of automatic weapons, see the excellent and easy-to understand article of Wikipedia about it.)
(Note 2:In the forthcoming technical description, parts of the gun will be referenced by code numbers can be found on attached drawings)
184.108.40.206.Operating cycle of the gun
- G29 M-sized Technic PF motor mounted on gun drives G13 and G14 ammo belt receiver gears through G7 chain and G8, G10 chain drive sprockets: the upper half of the chain is tensioned, sliding on lower edge of G3 rounded breech block, while the lower half of chain is loose. Chain rollers G9, G11 prevents chain to get loose from sprockets. This chain solution was necessary, because the motor is quite sizeable compared to the gun, and it would not fit on the same side of the turret as gun gearing. Moreover, I wanted to consume minimal space for chain under the gun to reduce drag of gun turret.
- G29 motor is geared to G13, G14 belt receiver rollers 1:1 through chain. G16 roller shaft is extended backward, and it has G17 worm gear driving G18 crankshaft gear at 8:1 ratio.
- G19 Crankshaft has a G20 crank with 1 studs radius fixed on that. Crank has a stud catching and driving G21 toggle. Toggle is a freely rotating arm on crankshaft with a radius of 2 studs. It is driven by motor ONLY IF crank is catching it.
- G21 toggle moves G22 pushrod forward/backward, which moves G24 bolt 4studs forward/backward cyclically through jack G23. When bolt moves backward - forced by crankshaft - it compresses G6 recoil springs. Then, after reaching rearmost position of bolt, recoil springs expand and push bolt forward fast. This is how I solved that bolt should press projectile quickly exactly at uppermost position of actual barrel. If G20 crank and G21 toggle were rigidly connected, it would be impossible, as barrels rotate 30 degrees at 180 degree turn of crank.
- In my system, rotation of barrels is not really necessary to get the gun working: it would work as well with single fixed barrel. As there is no explosion of chemical propellant and heat, and there is only one bolt, it does not make any sense to rotate barrels. So it is clearly a decoration (but working decoration...) to get the illusion of working Gatling-system. This is how I solved rotation of barrels: G19 crankshaft drives a chain of three G25 16-teeth gears, then a 12-teeth double conical gear fixed on G27 barrel transmission shaft. So this shaft is geared to crankshaft 4:3. Barrel transmission shaft has a worm gear driving G28 8-teeth gear fixed on G4 barrel rotating axis, so they are geared 8:1. This way, crankshaft is geared to barrels 3/4 × 8/1 = 6:1. As we have 6 barrels, this is just the correct ratio! This way, motor is geared to barrels at 48:1 ratio.
220.127.116.11.Ammunition and feeding system
- Projectiles are 3-stud long rods, which are linked together into G15 flexible belt by 2×1×1 stud rubber tie parts, echeloned alternatively in 2 strips. Forward moving G24 bolt first pushes the projectile into G2 barrel from rubber ties, then it retreats, allowing rubber tie parts fall free out of gun, so the belt is disintegrating type. As G13, G14 belt receiver rollers geared to crankshaft 8:1, rollers rotate 8 in one firing cycle, so they constantly slide on rubber belt links with mild friction. I chose this solution enbling belt quickly advance one position in belt pathway when bolt is retracted backward. Advancement of belt is stopped by the actually loading projectile hitting an adjustable projectile aligner. The aligner ensures that projectile will be in one line with bolt and upper barrel when bolt moves forward.
- I omitted cartridges and cartridge extractor - as they would require very small sized catches and hooks compared to other components - and I assumed usage of disintegrating cartridges. With the lack of chemical propellant, also there is no sense to use cartridges. Moreover, in the reality toggle-delayed blowback system has a problem with cartridges, as it delays bolt too good compared to other delayers: the cartridge moves only a small distance backward when gas pressure in barrel already rapidly decreases. It makes cartridge extraction hard and prone to jamming. So toggle-delayed blowback guns always used wax/or oil-fluted cartridges to prevent it. With cartridgeless ammo, this difficulty can be eliminated.
- Belt is fed from G37 rotary magazine (ammunition drum) with a capacity of 45 rounds. Of course in the reality it would be ridiculusly small for a 6 barrel Gatling gun firing 6000 spm (100 shots/sec), emptying the whole magazine in a 0.5 sec burst. But the Lego version is much-much more slow, so magazin capacity is pretty enough. G37 belt spool can be loaded with belt wound up in one move from downward. Alternatively, belt can be wound on empty spool fixed in its place with the help of reversed working of G38 belt extractor gear.
- But - besides loading - why we need a powered belt extractor? Why can not the receiver gear of gun just pull out belt from ammo drum? The reason is laying/training movement of turret: as gun moves with that, its distance from fixed ammo drum can rapidly change - also the necessary belt lenght between ammo drum and gun. As gun pulls belt with constant speed, the only way to adjust free belt lenght is the G38, G39 powered belt extractor gear.
18.104.22.168.Turret laying/training system
- In the reality, training and laying of powered gun turrets is performed by two separate servo motors. Laying servo can be placed besides the turret, but training servo should be on that. One big shortcoming of Lego Technic is the lack of small sized servo motors. In 2012, they introduced a new servo motor at a rock crawler model, but it is even bigger than M-sized motor. (Mindstorms servo motors are also bigger). In the turret, there was very clearly no space for another M-sized motor besides the gun driving engine and gearing to keep it reasonably sized. So I had to solve all power functions of the turret and magazine using one G44 M-sized central servo motor with complicated transmission and direction reverser gearing. This is how it is done:
- On a main shaft - driven by electric motor through gears G43 - opposite facing half conical gears are fixed in 1.5 stud-intervals. Axises peripedicular to main shaft have also a half-conical gears and universal joints, and they are mounted on 3 stud swingarm. Swinging the arm it connects half-conical gear of swinging axis to the half-conical gear fixed on main shaft to the left or right. This way, it performs both clutch and direction reverser functions. Swingarms rotate on 3 stud long rod with considerable friction. It means that they hold half-conical gears connected if the torque load is not too high. At overload, swingarm jumps out to neutral/disengaged position, preventing the mechanism to break up. There are 3 swing axis direction reverser/clutch units:
- Unit G45 drives G38 belt extractor gear through axis G41 and chain G40, controlling belt extraction speed from magazine, adjusted to gun firing and turret movement.
- Unit G46 drives G35 laying transmission shaft, then G32 laying conical gears, then G31 laying worm gear, which rotates turntable of the turret left/right
- Unit G42 drives G36 training transmission shaft, then G34 training conical gears, then G33 training worm gear. Wormgear drives an 8-teeth gear, then combination of 3 connical gears, transmitting drive through the cetral hole of turntable into the rotating part of the turret. There G47 training swingarm is actuated, which slides G48 training slide on G49 training lever, which is fixed to training axis of the gun. So the gun barrels can be moved up/down.
- As turret is controlled from a remote position, it should somehow designate expected trajectory of the projectiles for exact aiming. In the reality it is done by a complicated computerized system considering number of factors: speed vector of own/enemy aircrafts, aerodynamic drag, projectile drop by gravitation, turret delay and vibration, etc. The output is usually a virtual expected projectile trajectory projected on pilots/gunners Head Up Display or Helmet Display. Of course, in Lego Technic, it cannot be done, and softness of ABS material makes turret very-very unexact.
- But, to enable some limited targeting at pointblank range, we fixed a laser target designator on the gun: R.J.McNamara manufactures G30 laser range finder built in Lego Mindstorms Light Sensor part, which is fixed on gun training axis, so it is layed/trained parallel with the gun
22.214.171.124.Unsolved problems and shortcomings
- Training and laying drives are interlinked: if laying is changed, training also has to be adjusted to keep the given training angle. This makes computing correct firing solutions even more difficult. This is a forced solution by the lack of reasonably small servo motors.
- Barrels should be drilled up. 3-stud rods used as projectiles have very strong friction in connector pegs forming inner part of barrels. So, get the gun working requires to drill up those connector pegs to 0.2mm bigger inner bore. Of course this is an illegal movement by Lego principles, but re-usability of connector pegs is not lost totally by drilling upp, just they became somewhat weaker.
- Of course these were only the known problems of the design. In the practice, there can be range of new problems (eg. lack of lubrication, etc.). We intend to build the gun in the reality and refine the design with practical experiences.