In my world I have a genius inventor like DaVinci and my inventor has been asked to create a hand-held semi-automatic crossbow. The crossbow reloads and cocks itself without human input.
What is a believable design for the re-arming mechanism for such a crossbow? This is the mechanism that re-pulls the bowstring back into firing position.
An answer and a frame challenge
First, the answer: Yes
My working theory is springs. The concept of a spring has been around for thousands of years. In fact, a crossbow is based on a spring (the bow limbs). But to make things work the way you want them to, we need to invent something earlier: the coiled mainspring. But coiled mainsprings were invented in the early-1400s. That barely meets the Medieval requirement. But could we have it earlier?
To make this happen we need to assume some advances in metallurgy and manufacturing, but none are incredible save for the possibility of a hardened spring metal. Let's handwave that! It's thus believable that a coiled mainspring could exist, perhaps, as early as 750-1000 ce, if the right people thought the right ideas quickly enough. So, we have mainsprings. So say we all.
What the coiled mainspring does for you is allow you to wind up the crossbow like a mechanical clock. You'd obviously be limited in the number of bolts you could throw (the idea of the weapon in the movie VanHelsing in the Medieval era isn't possible or even plausible), but I suspect 3-6 bolts is believable. The energy of the spring combined with a few gears and latching mechanisms would allow for the bowstring to be re-pulled after each shot.
Keep in mind that we are talking about the Medieval period. I would hate (and I really mean that... I'd hate) to be the guy holding a crossbow when the latches give way. The physics of an uncontrolled unwinding spring are... impressive.
And now... the Frame Challenge
While it's believable that such a weapon could be created (so long as we handwave the necessary innovations and investigations that brought coiled mainsprings about at the very end of the Medieval period), the re-arming process has major problems.
It would be heavy. A Medieval crossbow is already heavy compared to even a longbow — but it's advantages well outweighed this problem. But that mainspring and the additional mechanics needed to manage it would seriously add weight. Yes, you could launch bolts as fast as you can pull the trigger — and all the while you're lowering the weapon due to its weight.
As mentioned, the harm that could (would...) be caused by catastrophic spring failure cannot be ignored. There's not a way to prevent this without adding even more weight to the device. Suddenly we're talking about a light artillery weapon that only shoots short-range bolts. Not real useful.
But your biggest problem is... what's the point? And here we need a longer explanation.
Firearms can easily be semi-automatic because the energy needed to eject a cartridge, load the next cartridge, and cock the hammer is infinitesimally small compared to the energy contained in the cartridge. Remember, it's the powder in the cartridge that's pushing the bullet — not the hammer, which requires a lot less force to push back against its spring than the bullet needs to do any damage.
Compare this to a crossbow where the energy to move the bolt is equal to the energy needed to cock the weapon. This is a consequence of Newton's third law: equal and opposite actions. The force provided by the spring (bow limbs) to push the bolt is equal to the force needed to set the springs (bow limbs) to do just that.
That means there's no "leverage" like you find with firearms. If you want to fire six bolts, you have to turn a crank and wind a spring to put six firings worth of energy into the spring. That's tiring! But it also begs the question....
If you have the energy to throw the bolt wound up in the spring, why waste time pulling back the bowstring? Why not simply launch the bolt directly from the force of the spring? A basic engineering axiom is to keep everything as simple as possible to reduce the potential for error and failure. And if we run with this realization, what do we get?
The spring-driven Maula Pistol from Frank Herbert's Dune.
Conclusion
Yes, I believe it's believable to have a semi-automatic crossbow using Medieval technology.
No, I don't think it would ever come to pass. It's too obvious to pursue a simpler solution — the spring-fired pellet gun.
There's a recent design from a film that might interest you.
The film "Gladiator 2" had a four shot rotating crossbow that should work fine'
The proviso is that the strings must all be cocked in advance and reloading would involve cocking all four potentially.
The propmakers at Zorg Industries have created this one (though the company website is now 404):
Copyright Zorg LTD, 2006. Fair usage. From thearbalistguild.forumotion 2026.
The steel may well have been out of place in the film but would work in the middle ages. Making carbon steel (1% in this case) is simple with charcoal. As to the springiness of the steel, that's all in the tempering.
As you can probably guess, the stock stayed still whilst the barrel and all the strings, bolts and bows rotated by 90 degrees between shots. A couple of stills can be found here from the feature film itself.
It should work.
The automatic reload systems were put in service quite long after gunpowder outclassed crossbows and bows and made them style or sport accessories.
Bows and crossbows store energy mechanically in the bow (spring) and the frame (or the archer) must hold the energy before the shot.
For reloading mechanism without input, the energy for n-shot burst must be at least n-times higher, the frame must be n-times stronger and the preloading will take n-times longer, plausibly demanding a horse or an ox to prime it.
On the other hand, the gunpowder stores the energy chemically. And even that needed the invention of sealed bullet-cartridge assembly to allow the sipler reloading and later the automatic reloading.
You would need to redesign the bow-arrow propeller-projectile pair. If you assemble the arrow in the spring-loaded cartridge with a durable latch mechanism then the frame does not need to hold the propelling force, it is stored in the cartidge-arrow assembly.
Then you just need a mechanism that releases the latch (fire), then expell the discharged cartridge, places the new cartridge in and primes the release mechanism for new shot.
This reloading mechanism can be powered by the bow and set of pulleys and meahcnical-clock-like mechanism to release the spring (bow) just a little bit.
The closest that you could reasonably get with medieval technology would be the Chinese Repeating Crossbow.
However, the low power and lack of fletching on the bolts meant that due to their low power and inaccuracy, they would only be of use at short range and/or against massed attacks, and the bolts would need to be poisoned in order to make them sufficiently dangerous to warrant being used.
In a TTRPG setting, I introduced a device like a crossbow that was recocked using a gunpowder charge, but that device fired missiles like a spinning circular razor-edged disk for... reasons... that couldn't have been launched from a gun. If you're trying to shoot a penetrating missile, a gun would do the job better, so there would be no need for an unnecessarily-complicated gunpowder-recocked crossbow.
Give him a sufficiently motivated hamster running on a hamster wheel with a sufficiently large gear ratio, and Archimedes shall shoot the world... very, very slowly... with technically semi-automatic crossbow bolts.
More usefully, if you connect a fixed crossbow or battery of crossbows to an off-site power supply (something like an ox team or a river turns a drive shaft which turns a drive gear), you can power that way. Loading the bolt in from a magazine once you have the bow bent is a historically solved problem; repeating (muscle-powered) crossbows existed, they just weren't very good.
Best cheating option: train a bonobo to do it. Notoriously athletic little guys, plus they have those foot-hands, bad for powerlifting, but excellent for push-pull leverage! No human input, win condition met!
Second-best cheating option: be so sexist/racist/caste-obsessed that you think that women/other races/other castes don't count as people, and have them do it!
It is possible, but would be almost useless in practice due to the tension. This was the drawback with automatic crossbows which used two hands. Crossbows suitable for war needed to be cranked or use foot levers to create enough tension. In your case you can only tension it with the strength of your finger pulling a trigger.
You would have to be so close to an opponent that you'd be better off attacking them with a knife. But if it was a toy, then sure, you could shoot at cockroaches and things like that and have a bit of fun.
There is no way to keep such thing hand-portable and powerful enough for military usage because no matter what solution you use, you will need add a whole bow's weight to the design as a theoretical minimum as somewhere to contain the extra potential energy... but many historical crossbow like weapons were not hand portable anyway; so, this may not be an issue. The Ancient Greeks and Romans often used heavy crossbow like weapons that they could fire from chariots, wagons, ships, or city walls. Many of these were designed to be powerful siege weapons, but they stored many times as much potential energy as a standard crossbow meaning that you could modify this concept to make a multi-shot crossbow.
Once you eliminate weight as a major constraint, you can begin to consider designs that would otherwise be impractical. Crossbows with multiple bows would be one option, but too many bows bolted together and they will start to get in the way of aiming; however, one possibility would be to use a ratchetted drop weight mechanism. With this method, you'd have just one bow that redraws itself using the potential energy stored in a heavy drop weight. The way it would work is that it would fire the bolt, and when string hits the end of its stroke, it will hit a release on a ratchet that will allow a weight to drop. This dropping weight would then pull the resetting mechanism on the bow and the ratchet wheel would relock when the string falls into the setting mechanism. You could then use a magazine to set a new bolt allowing the weapon to fully reload itself. Unlike with multi-bow concepts, this weapon could fire many shots without making the bow itself overly cumbersome to work with. A 20 shot bow would be just as simple of a design as a 2 shot bow, it would just require more drop weights.
Loads of good ideas up there- about how to transport and store that energy- my two cents on how to produce it are on legs. The legs are the most durable muscles in the human body.
Now how to do that? The answer is one huge burst, in one hard motion against a mechanism you carry with you. Or continuous mechanical energy produced, stored in "cassettes". I would go with the second option here- and hide the windup mechanism in two kneepads of armour near the knee. Running, kneeing you wind up 4 of the cassettes, who then transfer that energy to another storage that is able to produce that sudden burst of energy you need to accelerate the bolt because springs suck at producing fast, linear motion. They are great at producing continuous small forces.
Why the knees? Because resistance there is still the most ergonomic thing to handle as a soldier- and the springs and the winding gearing provide additional armour while not impeding motion as any other variant of energy production would. You could use the arms, but they would not be strong enough to load many cassettes - and the mechanism there would reduce the aiming precision.
I would also encourage everyone to see this as a measure of last ressource. Humans hate exercise, especially when it comes on top of exercise like marching, fighting and surviving wounds. So, if the energy can be stored and transported it would be gathered elsewhere- watermills, windmills, horses and donkeys walking in circles. Transport it wound up in your backpack and reuse it to reload your crossbow via a https://en.wikipedia.org/wiki/Flexible_shaft .
Have the recharge mechanism only for dire emergencies and cozy nights near the campfire, struggling to reload the backpack because the horse died. Zoom out. Distant curses. The end.
Making such a device "hand held" is probably not possible with medieval technology, because the sorts of energy storage and transmission systems available in the Middle ages, or even in late antiquity, are going to be large, cumbersome and low powered.
There was a form of repeating catapult in antiquity, which could be adapted - the Polybolos (Multi thrower).
Polybolos Mechanism
As you can see, turning a horizontal windlass mechanism drove a chain drive which spanned the bow, cocked the firing mechanism and allowed another bolt to drop into position to be fired. Since you are actually wanting to do this without human input (the weapon is aimed, but the mechanism isn't worked by the soldiers), then the windlass or cranking mechanism needs to be replaced. A powerful wind up spring might work, although springs of this nature were not invented until the 1400's, very late into the medieval period. However, the crank mechanism could be replaced with a flywheel, which is cranked to speed prior to bringing the weapon into action. Once it is spinning fast enough, the soldier takes aim and engages a clutch to couple the flywheel to the chain drive (probably through a reduction gear mechanism), resulting in a burst of fire so long as the clutch is engaged.
In effect, you have a machine gun like mechanism, and the soldier operating the Polybolos can control the rate of fire by engaging and disengaging the clutch. With sufficient practice and skill he could indeed fire single aimed shots, so long as the flywheel was spinning with sufficient speed.
The question is what would impel anyone to build such a thing? The main reason would be to conserve manpower - castle walls and ships at sea would need fewer fighting men with this device, and it can unleash a flurry of bolts when needed, making it particularly dangerous against storming parties trying to enter a castle or a boarding party on the deck of an enemy ship at sea.
A Cog. A repeating Polybolos mounted on the castle would make a formidable weapon at sea
So long as the operator or sergeant overseeing the device maintains the speed of the flywheel and can serve the weapon in action (dropping new bolts into the hopper, and maybe lubricating the moving parts between bouts of firing), then you have a formidable, although not "hand held" weapon.
