Artillery Safety

[P. Clodius Secundus]

I don't know if anyone has opened a thread on the subject, but here goes..
What happens when ballistas go "ballistic" and the genie named Tormenta gets let out of the bottle when you're not expecting it?
There are two main reasons to discuss this, safety and academics.
"Safety first" as they say.
I was able to find one reference to a weapons failure in this forum. There may be others.

A few years ago I ended up with quite a prominent scratch on my armour when one of our manubalista's arms broke under tension. The broken arm moved sufficiently fast that I did not see it move, even though I had been looking at the arm at the time. One moment it was there and the next moment there was a loud bang and it was half the size and in a different position. It was only then that I noticed the impact and the scratch on my armour. Had I not been wearing armour at the time I am fairly sure that I would have been looking for a new kidney in a hurry!".

Safety has been described as a pyramid. For every catastrophic failure there are likely to be 100 near-misses and 1000 unsafe conditions. The best thing is to anticipate problems. Next best is to notice unsafe conditions and correct them. Near misses are red lights indicating a major screw-up and the need for a serious rethinking of procedure. Failure? Like they say at NASA, "failure is not an option"! Admittedly, the Romans didn't have to worry about government agencies like OSHA here in the States breathing down their tunics. Re-enacting, especially artillery, involves a degree of risk. That doesn't mean we can't reduce it by analyzing past incidents and thinking ahead. The incident above makes me wonder where is the safest place to stand while working a torsion weapon? Has anyone examined this? I'm sure it will depend upon the configuration of the weapon. What are your Legion's policies regarding crew safety? Our Centurion is adamant about crew drill and wearing full armor and helmets. Sorry if I've opened Pandora's Box, but I can't "unthink" these questions. I've spent too much of my youth working safety detail on weapons ranges to ever turn it off. Input from you artillery veterans out there will help make everyone safer. Please contribute.

Academic Questions;
What are the physics of failure in torsion artillery? Does anyone have access to dynamic modeling software that can compare the different types of engines and scenarios? I am willing to try some small scale tests, but will gladly defer to the pocket-protector guys on this one. Any Mech Eng majors out there looking for a cool thesis? Broken arms and slings/strings are the most readily apparent disasters. Any others?
What does this add to the in-swinger vs out-swinger debate? A quick look at diagrams tells me that, for a given arm and washer size, out-swingers have slings about twice as long. Does this mean twice as much ability to absorb the shock? Twice as much sling to fail? Twice as far to stand back to avoid being whipped by a broken arm? Is one type or the other safer for the crew? Having confidence that a weapon won't break your shoulder or take you head off certainly improves accuracy. Even if commanders didn't place the same value on solders lives as we do I don't think the troops themselves were suicidal or stupid. Opinions or anecdotes anyone?

[Gaius Julius Caesar]

Gee Randi, do you work for Total too? :lol:

I can appreciate your views on safety, just starting my carreer as a safety advisor this week! D
The legionaries who operated these devices would undoubtedly have had an appreciation of the damage they could do if thay failed, and would probably been on top of maintaining their equipment to minimize failure. But as we all know, SH&T happens! :?

Give me some time in my new job, and hopefully I will meet some young engineer who finds the topic interesting, and perhaps get some answers to your question! Will need access to detailed info on the machines tho!

[Robert]

Hi guys,

as someone who had training as a safety engineer (but not a native speaker of English), just some thoughts based on analytical review and not on training (you would surprised at the lack of attention paid to torsion weapons these days :lol: ). Below is a incident analysis based on the weapon failing in cocked position (under full tension).
The main power/energy of a torsion weapon is aimed towards the front, stored in the wound up ropes. In the case of an arm failing (breaking), two different forces come into play. The wound up spring carries the greater amount of energy. The part of the arm attached to this spring will be violently catapulted forward, smashing into the rest. The other part will be released from this forward facing energy and be propelled back by the tension build into the bowstring and possibly the other arm (this will often be prevented by the hook of the trigger device). The force of the backlash will be nowhere near as powerful as the forward directed tension of the spring and will depend largely on the elasticity of the bowstring UNLESS the other arm comes into play.
This later scenario is dependent on the construction and strength of the trigger mechanism. Should the other arm come into play when the trigger mechanism fails to hold the bowstring on the failing of one of the arms, this changes the situation, as this will violently pull the free end sharply inward, perhaps causing the freed part of the limb to swing/whip towards the rear of the machine. The main vector of this force is however directed right to left or vice versa.

Looking at the forces involved, I would recommend not to stand within an arms length to the side within the arc of the arms. Standing behind the machine to observe, aim and operate the trigger mechanism should be quite OK, observing the position of the triggermechanisme as the most forward position, drawing a line diagonally at right angles to the trigger and operating the trigger with an extended arm.

There is a safety feature I dreamed up analyzing this situation. A length of rope or sturdy linen running from center to tip of each arm, securely fastened to center and end of the arm and bound to the arm itself would stay intact on failure of the arm and hold the pieces together long enough to prevent the violent back-whip which COULD occur if the other arm where to come into play on the secondary failure of the trigger device. Best to use this on the inside of the arm as not to greatly stress the points of fastening. Perhaps even wrapping the arm in a single swatch of sturdy linen would achieve the same effect without greatly diminishing performance. In that case, do not glue the linen on, as this will make it an integral part of the arm and prone to failure when it breaks. Leather is not recommended.
Hope these thoughts help

[P. Clodius Secundus]

Gaius and Robert,
Thanks for your responses. They are helpful and encouraging as usual. "Just the ticket" as they say. I am not a "safety guy" by trade, but my employer takes it very seriously. Attendance at safety meetings is a closely watched metric and each meeting leads off with a breakdown and analysis of all recent incidents before the main topic. For each process step or proposed change we are required to conduct a Job Hazard Assessment (JHA). It must be rubbing off on my hobbies as well.
Robert, your analysis would appear to indicate that in-swingers have a greater safety zone to the sides since the primary forces are directed to the front where no one should be anyways. Is that an accurate statement? I mention this because we found it was much easier to load the weapon from the left rear while the firer took aim. This made me nervous during our trials (with an out-swinger) since I felt like I was within the "Whip Zone" of a broken torsion arm. I am planning to place the ammo baskets in the cart bed on either side of the stock of the catapult, closest to where they are needed. Considering that the arms are at chest/head height you can see my interest in getting this right. I am thinking of putting together a safety handbook for ballistriarii. Any suggestions on what points should/must be included?

[Robert]

I would need some further data on the inswinger-outswinger issue. I am not a ballista expert and am contributing on basis of understanding of vectors and the way forces work on objects under tension. Please give me a link to the approriate chat on the forum or point me to a diagram of both. Like I said, much does depend on the direction force is applied. The string carries little energy (the bow/torsion arms wouldn't work if it were able to stretch to any great extent, thus building up energy there), even when the engine is cocked.
Loading should be done from the rearside, left or right. As these things resemble crossbows, there is no nock, so dropping in the bolt should get a good position if you reach over, place the bolt and slide it back a bit. The winding of a linnen cover over the arms and avoiding a bowstring with to much stretch should keep you safe on outswingers. The loading sequence of most torion or spring opperated weapons with a fixed trigger assembly is first cock, then load and then aim, allowing the loader to get well clear before anyone gets near the trigger. Of course, the sequence with a bow is load, cock (draw) and fire, but then it is just one opperator.

[Crispvs]

Just to clarify, when the incident referred to above happened I was in the action of loading the weapon. The bowstring and arms had already been cranked back and I was standing to the side of the manubalista, as we generally load the bolt with both hands to ensure that the bolt sits correctly in its track so that it cannot misfire. I was therefore reaching over the drawn arm and bowstring with both arms. When the metal arm snapped, it behaved exactly the way Robert suggested, with the piece of the arm still attached to the spring springing forward through its maximum possible arc, whilst the broken portion of the arm, still attached to the bowstring shot backwards, presumably through some sort of arc, as it hit my armour on the way past.

As a matter of safety we have always insisted that artillery crews operate in both body armour and helmets. This is the only example of a dangerous failure I am aware of but the practice of always weraring armour when operating artillery was fully vindicated on that day.

As loader I would normally place the ammunition basket where it was most convenient to draw from. We always take up the new bolt before the arms are cranked back, although we do not load until they are in position.
It should be said here that our manubalista incorporates side rachets as the depiction of the engine shown on the Cupid gem would require rachets in order for the crank handles to operate effectively in the position they are shown in.
http://www.unc.edu/courses/rometech/pub ... pidgem.JPG

Our normal shooting sequence is: draw, aim, load, release.

Incidently I am in the camp which holds that the inswinger catapult is a fantasy based on an incorrect reading of the report on the circumstances of the Hatra find, so I don't believe that there is much point in comparing inswingers to outswingers. :wink:

Crispvs

[Paullus Scipio]

"Incidently I am in the camp which holds that the inswinger catapult is a fantasy based on an incorrect reading of the report on the circumstances of the Hatra find, so I don't believe that there is much point in comparing inswingers to outswingers. :wink: "
Crispvs[/quote]
Hear ! Hear ! I have expressed a similar view on the "Aitor alert" thread and await his response with bated breath......

[D B Campbell]

... an incorrect reading of the report on the circumstances of the Hatra find ...

Please elaborate!

[Crispvs]

Duncan,

Alan Wilkins pointed out to me a few years back that according to the excavation report and photographs taken at the time the remains of the catapult were positioned in such a way as to show that the catapult could not have had the forestock that Aitor has demonstrated an inswinger would require. He suggested that the stansions should be positioned differently to the conventional positioning and in doing so suggested that the cuspings that suggest an inswinging motion may actually have been made to allow outsinging arms to travel through a wider arc by accommodating the foot of each arm further round than would be possible without the said cuspings. He also pointed out that the arms on an inswinger would need to be very short not to strike each other as they thavelled through their arc, whereas the Hatra frame could indicate a maching which functioned as artillery also had done in the past but which allowed for longer arms and thus an increased level of kinetic energy in the weapon.

Crispvs

[Robert]

From a mechanical point of view, inswingers sound both unlogical and cumbersome. I can't see any advantage, but perhaps that's just ignorance. I'll try to find the inswinger thread or artical but for now will side with the "outswingers rule" camp.

[D B Campbell]

... the remains of the catapult were positioned in such a way as to show that the catapult could not have had the forestock that Aitor has demonstrated an inswinger would require.

I think this is the argument that says: if there had been a projecting "case", the spring frame wouldn't be lying face-down. I'm not sure that this is valid, though. We really don't know the precise chain of events that led to deposition. I can imagine a scenario where a projecting wooden component could've snapped off as it crashed to the ground.
(Remember the old adage that absence of evidence needn't be evidence of absence.)

He suggested that the stansions should be positioned differently to the conventional positioning and in doing so suggested that the cuspings that suggest an inswinging motion may actually have been made to allow outsinging arms to travel through a wider arc by accommodating the foot of each arm further round than would be possible without the said cuspings.

I don't really follow this, Crispvs. The back-stanchions of the Hatra frame obstruct the forward swing of conventionally-positioned arms. You seem to be suggesting that the rearward range of motion of the arms was increased, by allowing the butt-ends to move into the cut-away space, as the arms are drawn back.
But there's still the problem of the limited forward range of motion. Conventionally-arranged arms would only move through about 30 degrees. Not ideal.
(You can see the rear-stanchions in Plate F of [amazon]Greek and Roman Artillery 399 BC-AD 363[/amazon].)

The arms on an inswinger would need to be very short not to strike each other as they travelled through their arc, whereas the Hatra frame could indicate a machine which functioned as artillery also had done in the past but which allowed for longer arms and thus an increased level of kinetic energy in the weapon.

I don't see a problem with short arms, Crispvs. The inswinging theory allows them to travel through about 110 degrees, with the consequent increase in torsion energy.
The longer arms theory is, to my mind, killed by the rear-stanchions of the Hatra frame. I'm sure Aitor explained this in his Gladius article, but I don't have it to hand just now.

[A_Volpe]

Why not load the bolt before cranking?

it ought to keep the "loader" clear of the weapon while cranking, and, if one is reaching over and placing the bolt while it's at full tension, I always fear an unexpected release, meaning your arm, hand and head are directly infront of that bowstring.

If the latch released "early", the bolt would be going downrange anyway, if not very far if little tension applied.
(that being said, I think it would be a good idea to have a "spotter" to make sure the range is clear during the load sequence....Maybe the crankers could try to look up/downrange after they crank and reset?)

It also means that if the string failed, as had been mentioned, the loader would not be in that arc area.

I think standing as directly behind the engine as possible ought to be the "safest" position, or perhaps adjacent to the crank pulleys, as from what I can think of, that's where the least amount of flying objects are located.

However, with an Ongaer, like the one our Legion has, I'm nervous being near any part of that engine. Although I think being diagonally behind the rear corner/wheels of the engine is probably the best bet; as directly behind there is a risk of the sling failing in mid-flight, so the ball and/or sling comes backwards; and standing at the side of the engine isn't much better, as when that arm does release, there is no time to blink when it goes.

just some of my thoughts.

[Robert]

Actually, a force vector analysis of an onager shows it to be a pretty safe machine. It's forces work in one plane in a rather uncomplicated manner and like you said, standing to the right or left of this plane should leave you pretty safe. Having said that, there is always some danger in projectile weapons of any era, but if you are nervous about weapons, just stay well clear (and I do mean several yards to each side). People afraid of weapons often do not fully understand what makes these things tick and ignorance can get you killed very quickly when dealing with weapons. This is not a rebuke, just an observation. :wink: Stay alert, stay alive

PS your idea on cranking the machine loaded does have a lot of merit, as it will prevent a "dry fire" if the trigger assembly fails and releases the shot prematurely. This will prevent damage to the string and arms, as any archer will concur.

[Kevin J. Aschenmeier]

I always load the bolt/arrow into my scorpion before I crank it up. It just seems safer. I get a bit nervous when I hear the machine groan and moan when it is fully cocked. So I prefer to load first, then I do not have to reach over the string.

Kevin

[P. Clodius Secundus]

... the remains of the catapult were positioned in such a way as to show that the catapult could not have had the forestock that Aitor has demonstrated an inswinger would require.

I think this is the argument that says: if there had been a projecting "case", the spring frame wouldn't be lying face-down. I'm not sure that this is valid, though. We really don't know the precise chain of events that led to deposition. I can imagine a scenario where a projecting wooden component could've snapped off as it crashed to the ground.
(Remember the old adage that absence of evidence needn't be evidence of absence.)

Good point. It’s nice to see that you’re willing to play the Devil’s Advocate as well. I don’t know the exact positioning and context of the Hatra artifacts but my answer is that the machine was probably disassembled and in storage at the time. The cheiroballista as a hand- held weapon would logically have an integral stock. Larger machines probably spent much of their time reduced to more manageable pieces. One of the carroballistas on Trajan's seems to be in transit up to a field fortification. Unlke the others there is no slide evident. Unless it was an oversight or part of the missing metal supposed to have adorned the reliefs, it would appear that they were not assembled long before going into action. I find it hard to imagine a scenario where a fully assembled engine that size was lost in the field and no one bothered to at least reclaim the metals. Does anyone know if the epyzigis bars were found in place, indicating if it was even strung? What about any other metal bits? Either way, even if one insists that the Hatra frame was an out-swinger it still demonstrates that the old Vitruvian machines could have been made wider without switching to metal. So, can someone give me another reason for the big change in design and tactics?

Regarding the safety question, I think all can agree that full armor is a must for ballistriarii. Pre-loading of projectiles would seem to be a good idea as well. As a rule, dry-shooting is usually bad for any weapon. In torsion weapons where all the energy is expended without any resistance from the projectile it is probably even worse.
I have only one concern.... We, and hopefully everyone else, only shoot live ammo without anyone between the engine and target. This is pretty much a no-brainer. The Romans fired over the heads of friendly troops. Those troops are there to protect the weapon and crew. Pre-loaded rounds which go off “half-cockedâ€