Iron Age Swords
Friends,

I have a rather broad topic for discussion, which then has the potential to evolve into a project... maybe.

I am curious about "iron age" swords, namely early ones (think "BC"). I am not sure exactly where to look for examples of sources, but I figure someone here might have some suggestions.

Next, I am curious about their metallurgy, and I'm wondering if anyone can offer some insight into that metallurgy specifically. Granted, that is a subject matter on a case-by-case basis, but I am wondering if anyone has any particular points to bring forward. In general, I am contemplating weapons which are considered to be "iron" rather than steel. As per the reason, note my points below.

Concerning a potential project (which will be a ways out, for sure), I occasionally have access to steel scrap from work. I am under the impression that most of this stuff is going to be A36 mild steel, and that is available in the form of flats 3/16" to 1/4" thickness - all I generally have to do to get something is ask, really. I am wondering if there are any examples of historic weapons with similar metallurgical properties to what I have access to, and if so, what might they be?

...The idea of a project like this is to better understand the performance of an iron weapon which is not steel - something which, unless I am mistaken about, is probably something that is fairly alien to us today. Namely, customers of reproduction weapons are probably not willing to spend a good deal of money on a weapon made of "substandard" materials, but I am generally under the impression that such materials were those which made up the bulk of early iron swords (aside from reports of the occasional smith/cutler who knew how to make steel). And then, if I can simulate any of that with resources I can source, why not?
Start with "The Celtic Sword" by Radomir Pleiner.

IMO the practical differences between "iron" and steel swords are minimal. Iron swords are more likely to bend and the edge dulls more quickly but how does that stop them from killing an opponent? They only have to remain functional for one encounter. Afterwards they can be straightened and sharpened. The best advantage of a steel sword is one of simple longevity.
I can't tell you very much as far as specifics, since I'm only picking this up from a variety of sources and folks who know a lot more metallurgy than I do. And so far I have found precious little about the metallurgy of early Greek iron swords, where my interest currently lies. (You can find a LOT about *bronze* swords...) Bottom line, wrought iron seems to have been good enough for early swords, because that's what many of them were made out of, right through the Roman era.

Sure, there's always some carbon content. And we also find things like sulfur or phosphorous, which may be contaminants but they do add to hardness. We also know that basic wrought iron can have approximately the same carbon content as modern mild steel, but it still isn't quite the same animal. As I understand it, heat-treating won't have much effect below a certain carbon percentage, but apparently it was sometimes tried anyway. And we know there was some steel, though I couldn't tell you how early or how often they were welding on steel edges, for instance. Or when they started welding on a lower-carbon *tang*, which apparently was quite a common medieval technique (probably Roman, maybe earlier?)

With all that in mind, I've made a number of blades from scrap 1/4" steel, with little more than an angle grinder. This is still my favorite:

[ Linked Image ]

Got a couple Roman ones in progress, and enough metal left for a Celtic sword, too, which I have always lacked. Will they perform like spring steel or anything with a higher carbon content? Heck, no! Will they be something like what the ancients may have used? Quite possibly. Couldn't tell you for sure, because I just haven't seen anyone make a bunch of ancient-style blades from bloomery iron and test them.

It *is* pretty clear that the first iron blades were highly refined weapons, being straight translations of the bronze ones already in use. So the smiths didn't have to bother figuring out blade geometry, since they already had a thousand years of experience with that.

Of course the big benefit to us reenactors or collectors on a budget is that angle grinders and scrap steel are CHEAP. And I've never done cutting or combat with my Greek and Roman swords, so I'm not worried about them folding up under abuse. I'm content that they are reasonably accurate otherwise, and probably perfectly lethal. (Though I don't tend to put a real edge on, too many kids are handling them!)

Sure wish I had more sources of information to throw at you!

Matthew
Thanks, guys. Dan, I had forgotten about that book, but know you've posted links to it before. Matthew, did you do any heat treatment at all? Furthermore, do you know of what type of steel you were working with? Certainly, for consistency, the modern low-carbon stuff has to be better than what was better then. Only the question of hardness and cutting power remains.

...What of carburizing? How well does evidence of this process hold up in archaeological finds, or does this just tend to oxidize away?
Nice video here: https://www.youtube.com/watch?v=4iANEih6SFo Metal info at 4:30
Hi Michael,

The history of metallurgy in swords is something that I love to dig into. If you have't found it already I would highly recommend Helmut Föll's site on the subject. He is a professor at the CAU Institute of Materials Science in Schleswig Germany. His site, which I gather is simply a hobby of his, has some excellent information. If you don't want to read the whole thing (there is a lot) start in chapter 11. Of particular interest to you may be the subsection on Celtic sword metallurgy. Enjoy!


Link to chapter 11 index page: https://www.tf.uni-kiel.de/matwis/amat/iss/index.html

Link to main index page: https://www.tf.uni-kiel.de/matwis/amat/iss/index.html

Link to the section on Celtic sword metallurgy: https://www.tf.uni-kiel.de/matwis/amat/iss/kap_b/backbone/rb_2_2.html
Michael Beeching wrote:
Matthew, did you do any heat treatment at all?


Nope. It's quite likely that some smiths experimented with it (I honestly don't recall in detail what has been found), but it's just not something I've ever tried so I wouldn't want to give it a whirl on a blade I sweated over for several days! Plus it probably wouldn't do much good on the metal I have.

Quote:
Furthermore, do you know of what type of steel you were working with?


Nope! I'm assuming it's garden-variety mild steel, but it could be a little better than that. So far I've used 3 different pieces of 1/4" angle iron pulled out of the dumpster. If I did this a lot more frequently and paid closer attention to how each piece reacts to my tools, I might be able to tell you that one piece was harder than another, but honestly I seem to go through part of my learning curve every time....

Quote:
Certainly, for consistency, the modern low-carbon stuff has to be better than what was better then. Only the question of hardness and cutting power remains.


Agreed! Sure wish we knew more.

Quote:
...What of carburizing? How well does evidence of this process hold up in archaeological finds, or does this just tend to oxidize away?


Don't know. If you want something to keep you awake at night, I have heard that in the 19th and early 20th century it was common to stabilize iron archeological finds by ANNEALING them. So how much can we trust any tests that are now done on a blade found a century ago? Hmmm...

And yes, I'd have to assume that it's *possible* that surface treatments have corroded away. Plus iron will swell as it corrodes, though you can have metal loss as well, *and* there can be bits of scabbard clinging to the blade. All of which will screw up your cross-section drawings, which of course are just drawings mostly done for the vague amusement of archeologists and typologists, not laser scans designed for making perfect replicas.

Good research questions the answers more than it answers the questions.

Matthew
Len Parker wrote:
Nice video here: https://www.youtube.com/watch?v=4iANEih6SFo Metal info at 4:30


Oh, NICE! He's a bit more meticulous than me, and has some better machines as well as more patience. I pretty much agree with everything he says, though I do think his blade could be a *little* thinner and lighter. BUT it's entirely possible that it's spot on, either because low-quality blades were just chunkier (which I'm not convinced of--why not cheat by using *less* metal??), or because the ancients found that lighter blades of crappy metal just folded up too easily. Don't know! And the concept that "most blades were low-quality because they were iron and not strong enough and the smiths didn't care so they made them thick" doesn't hold up for early Greek blades, which were all made for men with money. But heck, it shouldn't even be true for that hispaniensis, since all legionaries were landowners who bought their own gear, at that point. So I think we just don't have enough good information, yet, and may be missing something.

By the way, I believe that's the same blade my own hispaniensis is copied from! There were 2 found at Smihel, very similar. And yes, very asymmetrical! Like him, I basically traced out one edge and then copied it for the other side.

Neat stuff!

Matthew
Re: Iron Age Swords
Michael Beeching wrote:
Friends,

I have a rather broad topic for discussion, which then has the potential to evolve into a project... maybe.

I am curious about "iron age" swords, namely early ones (think "BC"). I am not sure exactly where to look for examples of sources, but I figure someone here might have some suggestions.

... I am contemplating weapons which are considered to be "iron" rather than steel.

For the period before Radomir Pleiner's book and Janet Lang's 1988 article on imperial Roman swords, chapter 5.5 of my PhD thesis is the only one-stop shop I know. There just is not a lot surviving in good enough condition to analyse, and not a lot of that has been cut in two and studied.

Like others have said, its actually harder to make a consistent load of good wrought iron in a primitive smelter than make the mixed products that the Japanese sword-smiths use. So you could find swords made of anything from slaggy iron to quenched and tempered medium-carbon steel at any big market in Eurasia at any time between 200 BCE and 1850 CE.
Thanks for all the great replies and resources so far!

Len, that was a great video to watch. I do currently have another A36 "sword" project on hold as I'm busy with other things at the moment, but what I liked most about that video is the demonstration of how much you can do with a relatively simple set-up. Indeed, a LOT of those processes were familiar already. :D Now, as per the issue of draw-filing, the tooth size on the file was a really interesting matter I had not thought of. In my case, what I found worked for me when using a common bastard file was copious use of the file card to get build-up out of the teeth after or between strokes. Tedious and miserable work, but you avoid the deep gouges that way. I also loved the humor - maybe I need to pry into the chap's channel more?

Tyler, (Dr., I presume?) Foll's website is amazing, and I already sat down completely absorbed in it for some time. Not that I absorbed it completely, if I may make that clarification, but hey! Thank you, I will be going back to that one.

Sean, thank you for a link to your thesis! I for one cannot imagine compiling something of that scope - I started reading just a bit of it and I'm already very impressed, and look forward to crawling through the rest of it. Along with Pleiner's book, which a good friend of mine helped with in terms of nabbing the PDF, it looks like I've got a lot of reading to do!

Matthew, one of the things I actually quite enjoyed about the small bit of watching and researching I've done so far is that early iron age swords indeed seem to have been hit-and-probably-mostly-miss in terms of their performance. It's a great slap in the face to contemporary programming like "Forged in Fire" where they tend to assert that if the weapon is not hardened or tempered, it is not "a weapon," or such are the likes. Rather, the early iron weapons were probably just what you thought they were for the most part - not that great. Indeed, after browsing Foll's site, a great many early weapons were not even heat treated. The steel we're looking to use (probably, again, A36) might get tougher from heat treat, but it won't become much better than what it is when you grind it to shape. Actually, what I want to try is (a.) work hardening and (b.) hammering the edges.

Neat. :)
Speaking of the fellow who did the A36 reproduction of the Gladius Hispaniensis, he also did a video on carburizing:

https://www.youtube.com/watch?v=aFqdWXDZiq4

...The materials for carburizing would conceivably been easy to acquire in the early iron age - you would have a crucible and charcoal. Knowledge of this process and how to handle it are a different matter entirely. However, from the skimming of materials I have gathered so far, it's not wrong to believe that early smiths would have known the difference between high carbon and low carbon steel. It's also very possible that they would have made the connection between charcoal - a source of carbon - and the production of steel. Otherwise, why would examples in Pleiner's research demonstrate dispositions, at least on occasion, of steel specifically concentrated in one area of a blade while not in another? Of course, if we determine the heat treatment of any of those items was questionable, I'm not sure how much better steel would be over iron in that case. That said, more reading is required on my end.
The Celts used blister steel, which tends to have the carbon content very unevenly distributed throughout the metal. This needs considerable refining (folding and hammering) to drive out impurities and redistribute the carbon more evenly, but it wasn't always done.

Some smiths, such as the Japanese, prevented this by intercepting the bloom after the smelt and breaking it apart. From there they sorted the pieces into different grades of high, medium, and low carbon before smithing it.
Well, this is what I'm wondering after reading this article (thank you, Dr. Manning):

https://bookandsword.com/2018/10/26/a-luristan-akinakes/

Given bloomery iron, if one is NOT able to properly reduce or distribute the carbon content from refining, by whatever means constitutes refining, is a potential solution to simply either not quench or just anneal a weapon? And, if the latter is a selected solution for the purposes of a more durable (and thus ductile) weapon at the expense of hardness / sharpness / springiness, does more carbon near the edge - make for a better edge - in an unhardened iron weapon?

In noting the article, I was surprised at how high the carbon content was on average, though it was apparently not at all consistent. If one was ignoring the consistency issue, that carbon content would in fact look very similar to some conveniently garnered information on modern steels, courtesy of Mr. Craig Johnson, on this very site:

http://myArmoury.com/feature_bladehardness.html

...Of course, consistency is generally questionable in an iron age implement. Despite my general ignorance on this subject so far, I am generally confident of that position. Given that inconsistency, the hardness of the akinakes' blade is stated as varying through 111 to 260 on the Vickers scale, and is furthermore apparently random in that disposition as well. All of the potential techniques for hardening the blade previously in this thread are also in Dr. Manning's blog post as well, with the same concerns for why we probably can't find traces of implementing those techniques in this particular archaeological find.

Regarding hardness, note that Craig's feature article focuses on Rockwell C. Those interested in knowing the difference between the current hardness testing standards should investigate them individually, but for simplicity, here is a convenient chart available on Wikipedia which compares values between the various tests - I assume that these values aren't always interchangeable in practice:

https://en.wikipedia.org/wiki/Hardness_comparison

Now, for the purposes of experimental archaeology, consider the commonly available A36 mild steel:

https://www.azom.com/article.aspx?ArticleID=6117
http://www.matweb.com/search/datasheet.aspx?m...amp;ckck=1

A36 has half the average carbon content of the akinakes in question, but it is of course consistent. If you were to produce a simple replica via stock reduction, you would not have any real concerns about weird internal cold shuts or what have you, etc. In terms of hardness, the latter site gives estimates based on other tests in the Brinell and Rockwell B scales. Brinell hardness will range between 119 and 159. Rockwell B hardness will range between 67 and 83. Without any more real knowledge on this subject matter, it is curious that the Brinell hardness can increase over 34% of the lowest rating, and the Rockwell hardness only increases 24% over the lowest rating. I think this is important to hold in retrospect going forward.

Now, comparing the test values of the akinakes to common A36, we can make these rough comparisons:

111 Vickers (set to 110 per the chart) ~ 105 Brinell ~ 62 Rockwell B ~ NO VALUE for Rockwell C

260 Vickers (set to 261 per the chart) ~ 261 Brinell ~ 103 Rockwell B ~ 28 Rockwell C

Note that the iron age sample, or what survives of it, will in its softest [tested] parts, be softer than A36. It would be interesting to know the carbon content of the soft parts of the akinakes, given my prior question, all the while we are generally confident of the composition of the modern mild steel. Likewise, the iron age weapon will be harder than A36 in its more extreme tested points - note the 28 Rockwell C rating. In contrast, the test-derivative-numbers for the maximums of A36, Brinell 159 and Rockwell B 83, will have a Rockwell C rating between a mere 2 and 3.

...Therefore, if I am able to toy around with making a simple reproduction at some point, I will be very curious how much I can work harden an edge with just a hammer. Of course, I will need to at least have a simple set of hardness tester files before I can do that, and that's not in the budget for this month.

As per why the akinakes is such an item of interest here, well, it is an iron age implement, of course. Cyril Stanley Smith estimated it was from the 7th Century BC, but I will defer to Dr. Manning that this age is questionable. I will assume, however, that a similar weapon could be found along side this one:

https://www.trocadero.com/stores/101antiques/items/1398911/Ancient-6th-5th-century-BC-Scythian-large-Iron-Sword-Akinakes

I'm not sure the article above is real, but I frankly don't care. I think that's a neat sword, and if I do tinker around with something, it will be based on the item above. In any sense, that will be a ways off regardless. :)
If you're really keen and have a forge you could try a chunk of bloom from someone like Lee Sauder ... modern mild steel is definitely the cheaper option though!

I don't know of many people who have tried copying one of the big akinakai! I have not done any ironworking.
I did my first test on a piece of A36 with the new Tsubosan hardness tester files I bought earlier this month. The test piece began as a 3" x 3/16" cross-sectioned steel strip. Areas on the strip with the original outer "mill surface" intact (where the strip is extruded or otherwise forced through a die) were quite resistant to the files. The first apparent gouging caused by a file was noted with the 55 HRC tester! In contrast, hardness on the metal where the work-hardened surface had been ground away could be gouged with the 40 HRC tester.

...I figure this is a useful datapoint to bring up before moving forward, whenever that happens. Regardless, I think the test does demonstrate that A36 can in fact be work-hardened to very suitable levels of hardness. It will be interesting to see the results of completely removing the outer scale on a piece of strip from the stock removal process, and then proceeding to hammer-harden the edge for effect.
Michael Beeching wrote:
I did my first test on a piece of A36 with the new Tsubosan hardness tester files I bought earlier this month. The test piece began as a 3" x 3/16" cross-sectioned steel strip. Areas on the strip with the original outer "mill surface" intact (where the strip is extruded or otherwise forced through a die) were quite resistant to the files. The first apparent gouging caused by a file was noted with the 55 HRC tester! In contrast, hardness on the metal where the work-hardened surface had been ground away could be gouged with the 40 HRC tester.

...I figure this is a useful datapoint to bring up before moving forward, whenever that happens. Regardless, I think the test does demonstrate that A36 can in fact be work-hardened to very suitable levels of hardness. It will be interesting to see the results of completely removing the outer scale on a piece of strip from the stock removal process, and then proceeding to hammer-harden the edge for effect.


That is a pretty interesting result.

And I kind of hate to say it, but I think the piece of steel you got should probably have been rejected by the steel mill's quality control department.

I would assume that low grade structural steel like A36 is supplied in hot-rolled condition, or, ideally, in normalised condition, meaning that the steel is allowed to slowly cool down to reduce hardness and gain toughness. Such high surface hardness would probably cause it to fail in Charpy impact testing as required by most standards.
Huh - never really stopped to consider that. I suppose it's possible that the pickling process did not remove as much scale as it could have. But, again, this stuff is run through a die, so some work hardening should be expected. The stuff I have does not look to have been cold-rolled, which does have a tendency to crack when bent (doing a hardness test on that would be very interesting, actually...). But, for some items we have produced in the shop which are known to be hot-rolled, 90-degree bends on 3/16" material can cause some minor fissures. These are not really concerns for our industry in most cases, however.

...I do recall some instances in which some of the welders would grind away some of the surface of the flats in order to clear the outer scale. Once the scale is gone, the steel is soft much like you'd expect A36 to be. I don't think I've ever seen a flat like that shatter. Again, the closest thing I've seen to that would be when attempting to form cold-rolled flats, especially when forming them "against the grain of the steel."

It is an interesting thing to keep in mind, though: if you have to make a sharp implement out of steel flats, count on the outside being harder than the inside. Use a chisel grind!

*EDIT:

I'd also like to note again the testing mediums used: hardness tester files. This is my first time using them - they are certainly NOT a challenge to use and understand, but I am not sure you should consider them as calibrated scientific instruments. That said, I do think they seem to have given some good "ballpark" numbers for the steel sample. I could do a second test later, to attempt to see where the shift in hardness might be. In that regard, I don't think I'd have the instrumentation available to accurately see just how thick the hard surface layer is, but I am confident in saying that it's not very thick at all.
Michael Beeching wrote:
Huh - never really stopped to consider that. I suppose it's possible that the pickling process did not remove as much scale as it could have. But, again, this stuff is run through a die, so some work hardening should be expected. The stuff I have does not look to have been cold-rolled, which does have a tendency to crack when bent (doing a hardness test on that would be very interesting, actually...). But, for some items we have produced in the shop which are known to be hot-rolled, 90-degree bends on 3/16" material can cause some minor fissures. These are not really concerns for our industry in most cases, however.

...I do recall some instances in which some of the welders would grind away some of the surface of the flats in order to clear the outer scale. Once the scale is gone, the steel is soft much like you'd expect A36 to be. I don't think I've ever seen a flat like that shatter. Again, the closest thing I've seen to that would be when attempting to form cold-rolled flats, especially when forming them "against the grain of the steel."

It is an interesting thing to keep in mind, though: if you have to make a sharp implement out of steel flats, count on the outside being harder than the inside. Use a chisel grind!

*EDIT:

I'd also like to note again the testing mediums used: hardness tester files. This is my first time using them - they are certainly NOT a challenge to use and understand, but I am not sure you should consider them as calibrated scientific instruments. That said, I do think they seem to have given some good "ballpark" numbers for the steel sample. I could do a second test later, to attempt to see where the shift in hardness might be. In that regard, I don't think I'd have the instrumentation available to accurately see just how thick the hard surface layer is, but I am confident in saying that it's not very thick at all.


I'm not trying to be argumentative (and I think I might come across like that) but I think it's unlikely that a cheap strip of low grade structural steel is anything else than hot rolled. That process is the easiest and cheapest for steel mills. It's hot rolled, coated with a reddish-brown weldable shop primer coating. The shop primer is weldable but some welders may clean it up before welding, which is a good idea if the weld needs to be as perfect as possible.

Any small fracture / crack after bending is a flaw and should lead to rejection if used for structural purposes.

High hardness, which can occur after welding, is a good indication that the test piece will fail at impact testing.

Back to the topic: a long time ago, I have made a Turkana wrist knife (e.g. https://www.amazigh.it/old-turkana-fighting-wrist-knife-ararait-kenya/) in the same way as (what I understood) they do it in Africa: from the bottom of an old-fashioned enamelled cooking pan. It kinda worked, but even after trying to carburize it (soaking in a charcoal fire for a couple of hours) and heat treat it (heating and then quenching in oil), it was still pretty soft and more or less unusable as a knife (edge rolled very quickly when cutting with it). I didn't try work hardening though. I suppose it could still be used as a weapon but after playing around with it, I failed to understand how to possibly use it as a weapon. Maybe if you look at the picture of the two guys in the link above, it's purpose is for some kind of ritual duelling rather than actual self-defence.
Paul,

You do not come across as argumentative at all! I'm getting the impression you work with structural fabrication - our company generally only makes carts. Either way, our A36 has never caused us problems so far as I'm aware.

And, I agree, the stuff I have must be hot rolled. I only mentioned cold rolled steel as we have used that as well in the past, namely when no forming was planned for its use (but ended up being what was available). It's generally apparent which process has been used on a piece of flat stock, but rather than make an incorrect assumption, I took the route of playing it as safe as possible. In fact, yes, it's hot rolled!

Now, to clarify the testing process, I only evaluated the material on the sides (flats) of the material, and not on the corners. Actually, why not look at the piece in question again?:

...When I scratch along a side of the flat with a 40HRC file where the scale has been removed, a clear scratch is made on the surface. When the file contacts the surface scale, two things can happen: (1.) The file catches the scale from beneath, and lifts it off of the surface - thus digging under, or (2.) comes into direct contact with the surface scale and skates on top of it, only leaving a superficial scratch.

As with the previous test, only the 55HRC file will clearly gouge the surface scale on the material. I do not have a softer file than 40HRC, so the material under the scale layer seems to be a uniform hardness, less than 40 Rockwell C (the material is ground at a taper). And, the scale layer is very, very thin. Once removed, everything settles to a uniform, fairly soft hardness. Unfortunately, it's a bit too soft for my current instrumentation to evaluate!

Page 1 of 1

Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum




All contents © Copyright 2003-2006 myArmoury.com — All rights reserved
Discussion forums powered by phpBB © The phpBB Group
Switch to the Full-featured Version of the forum