For an educational project, we are considering building an onager or two. I have only one drawing, and would really love any and all additional info on things like trigger mechanism, the size of the prawl, the way you think the winding mechanism was constructed and the size of the drum. Also how the arm is set in the torsioncords. We would like to make it a takedown model for transport and storage, so does anyone have experience with that?
Would really appreciate input, mostly from those who have already built one. After the project, we hope to donate it for the use of Legio X Gemina.
Robert,
I don't know which style of onager your drawing depicts, but I imagine it follows one of the three traditional designs that differ mostly in how the buffer that stops the arm is attached. Most build either the Payne-Gallwey with it's vertical four-legged railroad style buffer frame, or the Schramm type with a tripod buffer frame. Both are equally wrong since there is no evidence that the onager ever had a buffer frame to begin with.
The oldest and most historically accurate interpretation, designed for Napoloeon III by DeReffye, had no buffer frame. It used a wooden ramp to substitute for the pile of turf blocks that is supposed to hold the buffer. I've worked with all three types and from my experience the DeReffye is the best. We had one that was originally a Schramm-type. Every time we shot it either the arm or the frame would break. Once I'd convinced it's builder to ditch the buffer frame, extend the arm, and let the it go forward like Ammianus said, the range nearly doubled and the machine stopped trying to demolish itself. While researching the problem for him I realized that the arm should actually be bent. This solves nearly all the problems that many see as inconsistencies in the text.
P. Clodius Secundus (Randi Richert), Legio III Cyrenaica
"Caesar\'s Conquerors"
Wow, thanks for the input, guys. I wonder how I missed that thread, though, as I did do a search on "onager" to begin with. Now, having resurrected the interest, this could be fun to share the pictures as we build it.
There is something puzzling me, though, as the pictures of the De Reffye style onager do not seem to confirm with the description of Ammianus. The later speaks of the machine being basicly constructed of two oak beams which have been bent to a humped shape. If a beam is thus bent, this would leave ground clearance in the middle, the beam only touching the ground at each of its ends. This is not consistent with the model shown, here the machine rests on the ground its full length and the humps are created by carving out part of the beam.
We are to build the onager at the Liburna shipyard, and they have bent the ships planking using heat and tension. So we should be able to build it just as described. In the previous thread, Randi gives an excellent translation, which we could indeed take at face-value. We could even turn the yard into a onager production line, as we will be making the Roman flax rope ourselves as well and being a blacksmith, I could build a small smithy on site and forge the nails and parts on the spot.
Would the washers have been bronze or iron, I wonder.
I must say I am very enthousiastic about the design that is coming to mind, as it solves quite a few questions I had from the simple mechanics of being able to put this together in the field in a short space of time.
For the educational project, I think we will go as far as stacking turfblocks for its trial firings!
Robert,
Most interpretations have the curvature being carved rather than bent into the beams. If you have the capability to steam and bend such substantial timbers it would be interesting to see the result. It would certainly give more ground clearance for the spring and the end of the arm which extends through it. An added benefit is that the beams could be somewhat lighter. You will just have to be very careful with the placement of the crossbeams which will act as stanchions do in other torsion engines.
Steam bending is just the thing for bending the distal portionn of the arm. Other than finding a grown-bent timber of the right dimensions, it seems like the most practical method. I would suggest using a section from the outer edge of a log with the bark carefully stripped to avoid grain violations. This is how bowyers make bow staves that don't split under tension.
When it comes to the choice of iron or bronze for the washers, you might consider the use of wooden washers like on the larger two-armed ballistae that the onagers seem to have replaced. Since there is only one spring there is no need to balance the tension by adding or reducing twist. The real power comes from the initial tensioning. At very high stringing tensions additional twisting becomes a self-defeating exercise. Supposedly, it causes the bundle to bunch up and reduces the springs responsiveness.
The rest of the hardware should be very simple since no fasteners other than the spring rope are mentioned or needed to hold the frame together. The only iron parts I see necessary are the collar and pin for the sling, the trigger/latch, and maybe a bent-hairpin shaped piece lashed to the arm at the bend to reinforce it and froming a ring where the trigger can be connected (see attached drawing).
Good luck with your project. After figuring out the bent-arm solution for this machine, I've gotten side-tracked working on similar revisions or all the other Roman weapons. I really need to get back to this though.
P. Clodius Secundus (Randi Richert), Legio III Cyrenaica
"Caesar\'s Conquerors"
This is what I had just done for a first attempt. I am looking at things like easy transportation and the cord bundle keeping the whole thing together. To be able to fit it all together, the cord bundle is inserted through the washers last and will then need slight tightening to best grip the beam. This design would be very easy to disassemble. Thanks for your drawing, though, it does seem the track we are on is correct!
Looks good Robert!
I like the look of the bowed beams. It allows the frame to be made from a thinner orignial log and there is no need to build it up by stacking and riveting logs on top of each other.
Two things you might want to consider are ditching the ratchet in favor of simply using the winch poles backed against the rear crossmember and turning the crossmembers about 45 degrees to line up better with the arc of the arms travel.
P. Clodius Secundus (Randi Richert), Legio III Cyrenaica
"Caesar\'s Conquerors"
OK on turning the crossmembers, I see your point. On the ratchet, I do not really envision how that would work. You stick the last pole in, allow it to rotate back a bit and leave it to hold the drum, is that right? But that would mean that when winding it, there should be constant pressure kept on the drum when someone is resetting his pole in a new slot. If it slips, that could be hazardous. I think I will keep the rachet as a safety feature (the only one), it makes life a whole lot easier on the gun crew. And it will be an educational project, so we will be having 11 and 12 year olds firing the piece :errr: :woot: hock:
The only addaptation then will be using half the amount of torsion cord and replacing the sling with a cup to hold the wet clay ball they just made. We do not really want to sink the target :wink: When the kids have had their fun, we can crank up the amount of rope, install the sling and really let fly :-)
Perhaps they did not even bend the log, just went and picked a tree with a suitable bend, dressing it and halved that, turning it into two logs, just like a shipwright would do.
I read your comment on the sling being placed on the ground loaded. I can see the logic, but what of the previously attached sling? There would have to be a sort of easy release fastening to allow the sling just used to be detached. Could be 3/4 closed hooks, I suppose. This is going to be fun building and testing!!
@agrimensor No need to buy it if you have it there. If you help in the construction you will even get to fire it full power :-) Are there some more pictures of the actual mechanisms, this is a bit too general for me. And I am not very fond of a triggermechanisme which needs operating by standing next to a full cocked torsion arm with a hammer. Need to re-think that one, as I prefer a trigger that works when yanking a cord.
Robert,
Since you are having children operate it I agree that using a ratchet for safety is a good idea. Historically speaking, I don't think the Romans were all that concerned with safety.
If you want a very safe trigger mechanism I would suggest purchasing a properly rated release from a company called McMillan Design at http://www.seacatch.com/. Their seacatch line of release toggles is very popular with the groups here who buiild the massive Punkin' Chunkin' machines. They are fully tested and they are proofed way beyond their rated strength. They will also allow for a nice long trigger lanyard.
One problem with using the proper ramp buffer instead of a vertical frame is that it might not work with a mangonel type cup. The bent arm may allow it, but you won't know until it's built. I'd just get it set up to fire properly with a sling at reduced power and use the long lanyard to keep them safe. Any type of cup or basket is just perpetuating one of the worst anachronisms. Probably the only thing worse would be putting wheels on it. If you have the means to build several I'd reccomend buiding a separate medieval mangonel to show the differences.
P. Clodius Secundus (Randi Richert), Legio III Cyrenaica
"Caesar\'s Conquerors"
Ah yes, I see what you mean. Without the sudden deceleration of the arm from hitting the buffer, when using a cup in this design the projectile will stay in the cup and be launched into the ground in front of the device. The sling motion however will slip the restaining ring from the pin, allowing the projectile to come free at the height of the arce.
I was now thinking of using the Roman style trigger mechanism found on the ballista, with the lever/trigger holding it firmly shut. Those Seacatch releases look awesome, though. I suppose we should best build it well before we actually need it and test a Roman style trigger for safety. Should that not be safe, the Seacatch would indeed be a great alternative. Thanks for pointing it out!
Quote:Ah yes, I see what you mean. Without the sudden deceleration of the arm from hitting the buffer, when using a cup in this design the projectile will stay in the cup and be launched into the ground in front of the device. The sling motion however will slip the restaining ring from the pin, allowing the projectile to come free at the height of the arce.
That is exactly how Sir Ralph justified creating the buffer frame on his first version of the onager in his book on crossbows. He later published a corrected appendix with "new" information that had come to his attention. Considering that the only descriptive source we've had for the onager is from 368AD it's hard to call the idea of a sling new. It's a shame that his version has become the most popularly disseminated. From wheels to cogs to the buffer frame it's pure fantasy.
P. Clodius Secundus (Randi Richert), Legio III Cyrenaica
"Caesar\'s Conquerors"
Right! Now, as to the boom or pole. If you say the rope bundle was not pre-tensioned, that makes sense as in doing so the boom would not remain upright in the weapon. Any tightening rotation of the bundle would tilt the boom front. This would also imply the boom would have to have had a larger contact area to the rope to prevent lateral movement (falling to one side). The boom would need to be wedged between the ropes tightly. This action may have the effect of tightening the ropes enough to be able to hold it in an upright position. This would also imply the boom having a larger/wider base and tapering quickly to a more uniform diameter. The more weight in the upper stretch of the boom, the slower it would fire due to the inertia of the boom itself. Has anyone made a working model of this design, I am wondering ??
The above would rule out having to use a washer, in my mind. To fix the rope bundle to the frame on assembly, pieces of wood or iron would need to be inserted through the eyes of the bundle after passing them through the frame. These could be wedge shaped to allow them to pass through and then tighten the rope bundle. Two pins would be enough to stop the wedges rotating, although even friction could be enough. In my evolving theory, the bundle of rope would have been pre-wound and lashed to hold its shape. It would be a proper pain to have to thread the rope- bundle through the holes in the frame each time the piece was set up for action. Now, I am working on the assumption the onagers were carried in parts on a cart, and when the need arose, quickly assembled to be able to fire as soon as possible. One cart could hold one or two onagers and ammo, depending on the size. I wonder how big or small one of these had to have been to be both practical and effective? Is there any data on the weight of rock/missile these would throw?
Rock like granite or hard/dense limestone has a specific weight of about 2.5 kg (five pounds) for a cubic decimeter (10 cm or about 4 # squared). A ball of 15 cm diameter would weigh about 3 kg or 6 pounds. With some speed behind it, that would have quit an impact on the receiving end.
I hope this thread inspires a lot of onagers being build, as they would seem really handy weapons and relatively simple to both make and transport. With the design that is beginning to emerge, one of these could be assembled in less then half an hour by four men. Any group could have their own artillary cart :woot:
Ahhh, and I have just found a 2 meter long ash 1/4 section of a log waiting to become a bow. Just the bark removed. I just hope 2 meters will be enough for the boom. That ash could work like a charm for this, as it is straight grained and has been drying for a few years now. Would that be suitable?
To clarify Robert,
I meant to indicate that the spring bundle was probably heavily pretensioned using a stretcher machine similar to that used for other catapults. That is where the real power comes from. Twisting the washers to build tension like many people do is a case of diminishing returns. The more you twist the less responsive the sinew bundle gets over the long arc (90+degree) of its travel. Even though the initial load is greater the arm won't continue to accelerate as well. At least that's the theory. Pre-tension good, Twist bad!
For the arm's shape I'd go with a gradual taper. Lighter/longer until it almost breaks is better. A gradual taper also causes les grain violations that encourage splitting. From what I've seen you don't need to worry too much about the spring gripping the arm or holdinng it upright. At high tensions it will soon crush the fibers of even hardwooods like oak, ash, or hickory enough to make it's own gripping surface and it will center itself up just fine under load.
P. Clodius Secundus (Randi Richert), Legio III Cyrenaica
"Caesar\'s Conquerors"
hock: 
