Pilum performance test

[Dan Howard]

That is my point. Can you name someone other than Erik who can pick an extant example and make a replica close enough to validate performance tests? I can't. None of the commercial examples I have seen are close enouigh to be useful. The riveting isn't the same, the overlap isn't close enough, the shape of the hole, the heat treatment, degree of flatening, etc, etc. Not to mention the great differences between wrought iron and mild steel in mechanical properties. Just because you use links that are of a similar size and rivet them closed doesn't make it "authentic". It is fine for re-enactment purposes but hardly suitable for performace evaluations.

On top of that we still have little idea regarding the type of padding worn underneath. For performance evaluations against various weapons the padding is just as important as the mail itself IMO. A serious test would need to pick a specific extant example, make a replica as closely as possible, and then test it against various types of padding. Then you start again with another extant example to eventually enable some general conclusions to be drawn about the performance of mail.

An example to make my point: Late medieval Germanic mail often used wedge-shaped rivets rather than round rivets. Modern reconstructions made links with rectangular holes to accommodate the rivets. Closer examination of originals revealed the holes to be ovoid rather than rectangular. Performance tests indicated that any stress on a rectangular hole cause it to tear at one of the corners. This didn't happen in ovoid holes. So if someone decided to test weapons against the mail with the rectangular holes, they would conclude that the mail provided far less protection than someone who did weapon tests against the more accurate reconstruction.

[Gaius Julius Caesar]

Well, once I get a new one made, I will donate my subarmalis for reasearch purposes. You can find your own mail tho' :wink:

I would be interesting to see results based on actual riveted mail,
but how far out are the examples made by Ashok Rai, DSC, Deepeeka, and find-it -armoury? The examples I have seem to have round rivets,
is that what you are referring to as to the wrong shape of rivets?
As to the quality of the steel, what evidence is there as to the metalurgical qualities regards the make up of the wire the romans used?

I have seen the examples at Vindolanda, South Shields, and Edinburgh, but has anyone done analysis on the metal they are made of?
Erik has seen these too, but was he able to do, or recieve any reports of, analysis made on these samples? I know he is the top dog on mail! 8)

[Matt Lukes]

But Dan, do you have any proof that "None of the commercial examples I have seen are close enouigh to be useful" and that "It is fine for re-enactment purposes but hardly suitable for performace evaluations."? Those are the kinds of statements that require significant backup- and unless you have performance data for both an exact reconstruction and these commercial reproductions that prove there's a significant difference, they're invalid, yes? For all any of us knows, the differences are inconsequential.

What I said is that there was likely to have been some significant difference between quality and manufacture across the population of original mail made, therefore perhaps nearly any reconstruction made to the correct dimensions and of at least similar materials might just be fine- not quite the same thing you're saying.

Just looking at tensile strength and ultimate shear data for wrought iron and mild steel, I found there could be as little as a 5% difference in the former, and 3.5% in the latter. Heat treatment ought to be a non-factor since without much carbon, no real hardening is possible- the metals might be made softer, although the hammering to form them would re-harden them some. The rivet joint might not be much of a factor either according to some preliminary tests I've recently run with some commercial mail, most of the time riveted rings broke at some point along the wire, only rarely at the rivet, and just as many solid links tore as well. So that leaves the dimensional data, which might also be not much of a factor since at the level of 8mm OD (so-called "6mm") mail the dimensional differences will be very minor.

To me this all suggests the difference may-well be insignificant. Certainly I agree that an exact replica would be the best test material, however in the absence of sufficient amounts- testing requres a LOT of mail since it's destructive- using commercial stuff should at least provide some good model data. I'm going to continue using it and I'm sure not going to discount the results out of hand It's hardly for publication in Nature, but it ought to be interesting. If someone wants to get some fully-accurate samples and try it out too then we'll have an interesting comparison...

[Dan Howard]

How can I prove a negative? You find an extant example that even remotely resembles the links coming out of India and sold by Western retailers. There isn't one. The difference is definitely not "insignificant". The only way performance test is going to be useful is if you pick an extant example and make a replica of it. Riveted does not equal historical. The shear strength of mild steel isn't the only difference. The ductility of wrought iron far exceeds that of mild steel. Slag impurities also play a large role. Have you got any details about the way the Romans heat treated their mail both during manufacture and afterwards? The degree of flattening also plays a part in the links effectiveness. If you don't use links made by Erik or someone of equal experience then your tests won't demonstrate anythying exscept how modern Indian mail performs against replicas of Roman weapons. They won't tell us anything about Roman armour.

Here is a photo of one of Erik's links compared to the stuff coming out of India. Find an extant example that looks like one of the bottom two links and I'll conceed the point.

[M. Demetrius]

To lighten the tension here, it's simple enough to conclude that if Roman scuta and hamata/segmentata were utterly immune to pila, then in the Civil Wars, the people who knew all there was to know about using their own weapons would not have bothered to throw volleys of them at each other. We know that they did, as there are various recorded instances of their having done so.

Tit for tat, there's no way to prove whether our stuff is exactly like their stuff, and there's no way to prove it is. Best we can hope for is to make as close as we can, then draw tentative conclusions based on what we have. I'm no historian, metallurgist, engineer, or technical advisor, but if the armor didn't work, they wouldn't have used it, and if the weapons wouldn't work, they wouldn't have used them. They used both. So somewhere, there's a balance. Exactly where is probably the rub.

[Dan Howard]

There are plenty of targets that can incapacitate a soldier without the weapon having to punch through his armour. An arrow or javelin through the leg or arm will take a man out of the fight just as surely as a shot through the heart.

[M. Demetrius]

You're right about that, of course, Dan. But the whole purpose of the scutum was to make almost all the body covered, right? The typical battle plan would only have pila or other javelins being thrown before the lines had closed, so the scutum would be held in a protective angle to cover all. One could sneak in and get a leg or a foot, or maybe slip over the top and hit a face (that would be bad!).

I wonder if a pilum would penetrate a brass or bronze helmet? It might.

[Dan Howard]

Yep. That would be bad IMO a pilum would stand a decent chance of punching through a Roman helm but I don't think an arrow could.

[Gaius Julius Caesar]

All very valid points, and that is how I imagine it worked.
As far as the actual mail goes, have to admit my sample look like the ones posted, but not like Eriks or or the original you posted Dan.
Guess us plebs will have to make do with the dream, not the reality.

[Matt Lukes]

How can I prove a negative?

But it's not a negative Dan, a negative would be something like 'gold mail never existed'- that's impossible to prove- but you wrote that commercial mail is useless for testing purposes because would not perform sufficiently closely to a closer recreation, which is an easily and necessarily testable fact. If you compare Erik's and the commercial type in physical damage testing, and they perform significantly differently, then your statement is correct. If you haven't done this testing, then the statement is simply invalid, right?

[Dan Howard]

The negative I was talking about was trying to prove that there doesn't exist an example of Roman mail that resembles the Indian items. It is up to you to find an example of Roman mail that DOES resemble the Indian pieces. I doubt one exists.

It has already been demonstrated that Erik's medieval reconstructions perform significantly differently to Indian pieces. The same thing will happen with the Roman examples since the same people make both.