It took me all the night but finally made some samples and put them together with the information necessary to explain, more or less, why I choose these samples.
May be it´s clear, most possible not.
I did it mostly for people who knows little or nothing about steel and it contains some generalizations and creepy graphics.
The purpose of this experiment is related to the following question:
Does this picture represent reality?
This question possibly should not be answered by a simple yes or not. But , since the author of these famous illuminations is universally considered as somebody who was familiar with military life of the XIII century (certainly more familiar than we are), the question is taken seriously by most scholars.
Sadly in the majority of cases the answers can be put into one of two groups: those who take the obvious authority of the painter literally, as in this example from a well know HEMA association:
“The clear effect of even single-hand sword blows against steel helmets. In the center a sword splits the top of a helm from behind. On the right two swords hack into the sides of helmets while a spiked mace delivers a crushing blow. Notice also the portions of each sword (their center-of percussion) that does most of the striking is invariably the last third to second-half of blade.”
And those who argue the complete opposite, often including experiments like this one:http://www.youtube.com/watch?v=-h0e0NSwYNg
But may be the case is not that simple, nor as easily probed.
So I want to explore this problem even when I am completely sure an impossible one. However its implications may be useful in understanding how this kind of armour was made and what was considered desirable in their design.
First- Reading art
When we see an action movie today most of the actions portrayed are false.
Depending on the film they range to the little exaggerated to the grossly unrealistic. But even in the latter case some realistic aspects are preserved. A bullet could shown piercing more or less things than an actual bullet of that caliber would but it does not explode, for example, it does not changes its direction in the middle of the air, it does not makes a green flame of fire when leaving the mouth of the gun. The same mechanism is used in art of all times: to adjust the nut more than usual, but not to the point it doesn´t fasten anymore.
So, the question we should do may be is more “how much of this picture is real? To which point?”
Second- If it is possible …It would be that way?
One method to start this analysis is to discard what is blatantly impossible. If we can probe that , not matter the force required, a helmet cannot be cut as portrayed we could conclude that the artist imagined that kind of damage.
That should be the case if a helmet does not get cut but collapses with plastic deformation into the head, for example. In that case we could think that the artist never actually saw neither case: not a helmet being cut, not a helmet being hit by such a powerful blow. (this conclusion is debatable since the artist could want to show an extremely sharp blade which can cut into a helmet instead of just denting it, but we are not here working for absolute answers, just posible ones).
Third- The actors.
Here we have a long series of mensurable variables and not so mensurable ones.
In hitting a person wearing a helmet and cutting that helmet we must have the following:
1- A sword edge which will take the hit and cut without crumbling before.
2- A sword body, which would resist the effort and provide enough kinetic energy
3- A person wielding the sword, exerting enough force to put the sword in the required movement. (There should be taken into consideration the relative speed between agent and subject)
4- A helmet able to be cut
5- A head with a mail coif into that helmet not providing enough support to the helmet to avoid the cut
6- A body below the head not absorbing enough energy to avoid the cut.
In this experiment I take for granted 1 and 2. I´ll discuss the reasons later.
3, 5 and 6 are exclude to focus in testing the possibility that, given enough force, a helmet can effectively show a failure pattern similar to that portrayed in the Maciejowski bible.
Fourth Defining a goal.
So we have to find if the helmet can be cut.
A helmet like this one is a piece of an unknown thickness of an unknown iron alloy, with an unknown macrostructure (lamination, inclusions, lack of uniformity in alloying elements) and unknown microstructure (heat treatment, cold working, composition , etc.
There is, of course, the related archaeological record. Sadly for this time frame (ad 1200-1300) is pretty scarce. We don´t really have a lot of this helmets nor were them studied to such extent to have a clear idea how they were made, with which material, which hardening treatment received, if any. And the few analysis show very varying figures.
But we do have hints:
1- We can assume that the helmet isn´t thicker than 2mm nor thinner than 0,5mm
2- The iron alloys of the time seem to range from nearly no carbon content to rarely more than 0,8%.
3- Iron alloys of the time have more phosphorous and sulphur than today.
4- Anisotropies were widespread.
5- Heat treatment, deliberated or not, could be anything.
6- They surely knew cold working techniques.
Our goal will be to probe if any of the possible combinations of the above data can produce the kind of failure shown in the imagen. Which is similar to this
This helmet was subjected to a traditional test. It is an antique and the pattern of the cut shows a remarkable similarity with the one shown in our picture. Since the authenticity of this test has been contested (even when similar test have been performed in a Yoshihara´s blade in front of a crowd) we´ll try to replicate this pattern.
Fifth: The theory
To understand in which circunstances this pattern of failure occurs we should know how steel behaves.
To our question the stress/strain graphic is what shows the forces involved.
Every deformation implies a stress applied. If we vary the composition of the steel the stress to obtain the same deformation varies to.
At first we´ll show just the behaviour into the elastic limit, id est the range of efforts which flex the steel and still it recovers its initial form when the force stops.
Beyond the E amount of deformation the steel bends and remains bent. We can see that the higher the carbon content, the higher the force needed to reach the same deformation. So the higher the carbon content the more difficult to damage permanently the helmet.
As a curiosity I have included the behaviour of pattern welded steel. The elastic limit of patter welded steel can be everywhere since it depends a lot of the position and quality of the welds but before reaching its elastic limit it needs an even higher force to be deformed. This curious characteristic of pattern welded steel has been noted by knifemakers and by the Japanese while researching to find a suitable military sword in the 30`s.
But what happens once we go beyond the elastic limit?
It depends a lot of the heat treatment
Unhardened steel yields, if the stress persists it starts to deform permanently but at higher and higher forces. Beyond certain point it starts to yield at lower and lower forces and breaks.
That means that the stronger condition of unhardened steel is that where it was deformed to such extent that it doesn´t add any additional strength with more deformation.
But (and it is a big “but”) If we left our work hardened steel just in that point any deformation would find less and less strength in the piece, leading to its fracture. So the key for work hardening is a balance between the place where it gained some strength but still is able to oppose resistance to future deformations.
This point is hard to guess, to say the least.
So we find old tools, arms and armour with clear cracks of overworking.
In the case of heat treated steel the elastic limit reaches a much higher point but if it is left too hard it will break with little or not deformation. The higher the tempering temperature the bigger deformation it will allow.
If the temperature in the tempering is high enough the steel will behave as an untreated piece. (stricto sensu= a subcritic annealing)
Sixth The samples
What does it all mean?
That we have such different behaviour in heat treated and untreated steel that we must to make samples for both cases.
And that the same is the case for carbon content.
But the most evident fact is that we don´t have period materials with different carbon content to start. So we must constantly keep in mind in which sense the material we are going to use is different from the ancient one.
The first is the great care we, as a civilization, take to not let impurities make our steel more fragile.
Phosphorous is major problem, it builds hardness while cold worked. Since many things we build are cold worked this is a major problem. Because a cold worked steel is harder but at the same time more brittle.
We do use steels specifically designed to be hardened by cold working, most commonly to make the skin of our cars. But sadly I didn´t have any in my workshop at the moment.
What I did have was SAE 1010 and SAE 1070 which are steels with very low carbon and high carbon respectively.
So I prepared five samples. I used SAE 1010 and SAE 1070.
P1- SAE 1010 “untreated” (annealed)
P2 – SAE 1070 “untreated” (annealed)
P3- SAE 1010 work hardened
P4. SAE 1070 work hardened
P5 SAE 1070 hardened and tempered. (heat treated)
The steel was 2mm thick in origin but reduced at 1,5 mm by work hardening and then to 1mm by grinding.
In the case of the probes not to be word hardened they were thinned to 1mm just by grinding.
The following picture shows the sample made out of SAE 1010 a with a side work hardened. A cut was made on either side. The unhardened one was bent back and forth ten times without fracture. The other one resisted just two folds and broke, but needed more force.
The same was made with a 1070 sample
Beside the fact that work hardening this steel took a much higher toll in my chronic epicondylitis this time the unhardened side took two folds without breaking but just a 90 degrees bend to fracture on the hardened side.
Two more samples were made with annealed steel and a fifth went to the oven to be heat treated.
Seventh- Punishing things
My first move was fixing the sample this way, because I was very confident in the degree of hardness the worked samples reached.
And the blade to be used was a Oakeshott XIIIb Alexandria pattern following the dimensions given by Clive Thomas in an old Park Lane catalog.
The blade is just a 5 % shorter and where mr Thomas measured 1,5mm I left something like 2 mm just because I wasn´t brave enough.
But I didn´t finished the hilt yet so I improvised a handle with leather and tape. The final blade obviously could hit as strong as a blade with a pommel but it wasn´t the point to reach a realistic force, just a realistic pattern of damage made by a realistic edge. The edge is constructed just like it seems to be in the detailed available pictures of the three or four extant specimens of this kind of blades.
My bad at the fixing solution, the samples collapsed easily and needed to be fixed into other ways
At this point it was evident that the samples didn´t take enough hardening. The 1070 sample showed just a little crumbling which can be seen just below the number “7” of “1070”.
Unfolding both samples something appeared:
The 1010 showed that no cut could yield the steel. Only one seems deep enough to show the beginning of a fracture, just in the middle.
But unfolding the 1070 sample was a different history:
Several points had collapsed. It´s important to stress that no complete cut was achieved, just few crumbled zones below the cut. The cut just stretched the steel to the point it couldn´t take no more. This was exactly what I was expecting but in a more extended form.
At this moment it was evident that I under worked the steel, or the steel wasn´t cold hardenable enough or simply 1mm was too much. Which would imply that the cutting of a helmet was impossible to probe without knowing exactly how much more brittle ancient work hardened steel was.
So I decided to give it another try and reworked half of both samples to give them a new series of hits.
This is what I obtained:
This time they kept their shape more and some places which seemed undamaged in the 1070 sample showed a complete cut, so they were very near to be complete since the start.
If the work hardened samples were soft enough to be completely collapsed before being cut it was evident that their annealed version would collapse easily. So I put a hammer under them.
Then I missed a cut and hit the hammer which got cut along with a good nick in the vice. Luckily no damage on the edge.
Something was left yet. The sample hardened and tempered BD asked.
Quenched and tempered at 210 C. But something didn´t went as in the book. Possibly my SAE 1070 isn´t SAE 1070 after all. Because at 210C it should be more resilient. The f*cking thing exploded in sharp pieces.
Having a feeling that tempered steel was too much to risk in that blade I used a saber blade I made some time ago.
It was late by this time but I made a mistake so I tried to fix it by making another 1070 sample from the beginning
Tempered this time at 260 C which should be enough to present some plastic deformation. And I hit it really hard.
Too much for the poor blade and again the sample was too hard. But this time crumbled in the place hit, so it broke not by the deformation on the whole piece alone, but also in the zone actually hit by the blade. Which make me think that samples tempered at higher temperatures would continue following more and more this pattern of less damage in the whole piece and more along the hit itself.
That´s something left to do: intermediate heat treatments. Incomplete hardening, tempering at higher temperatures, etc.
But what happened exactly?
I think that this happened:
The two heat treated samples are in the left, the other ones in the right. We need samples in the middle but it´s predictable that they will have their fracture points well beyond human arm possibilities ( you can see the hypotetical fracture limits over the “strongest blows” line).
Another interesting conclusion is that 1070 annealed seemed to be stronger that 1010 work hardened. So there is a possibility that work hardened SAE 1010 is completely useless to replicate nothing.
Only SAE 1070 work hardened showed a fracture point between the reach of my blows and a lot of stiffness into its elastic range compared to the other three.
To make a great end of this not so serious experiment I did what everybody did: Smash a SAE 1010 helmet to probe nothing conclusively.
But it seems to me that making helmets with soft modern iron would result in either too poor performers or too heavy ones if we try to compensate with greater thickness. Modern mild steel and even modern high carbon steel will deform in a more historic probable way (id est the way shown in the M. bible or in the kabutowari tests) , even requiring much much higher efforts.