This question came up in another thread, but I did not really get an answer.

The question is, what should be the reasonable standard for testing the durability, cutting power, and thrusting ability of a medieval European-style sword? What materials, what dimensions, etc?

I am not talking about 'tests to destruction' (like cutting metal barrels) or tests meant to develop technique. What I am asking is; a sword of a given type should be able to cut through _____ and thrust through _____ without sustaining any more damage than _____.

I mention 'sword type' since likely the standards would be different for, e.g., a type X cutter and a type XVII thruster.

After years of reading on-line sword reviews, my conclusion is that the standards out there are so variable and subjective that one cannot really call them standards at all. (Unless there are industry standards that most of us regular customers are not aware of?) Maybe it is time to set those standards for the community, or at least try.

-JD

I'm a little confused about what you're looking for. I understand the principle of your question, but are you looking for how far into a given medium a sword of a given type should be able to cut or thrust and what damage is sustained to the blade by the act, or how much damage is reasonable for a blade to take when cutting/thrusting into a medium an x amount of distance? It's a fine, but distinct, difference.
I also see a problem with standardization in your question. For one, what's the medium? For another, what type of cut, how hard is the sword swung. For example, an Oakshotte Type XIIIa will outperform a type XVa when cutting hard targets, but the reverse is true for cutting lighter targets like pool noodles I just don't see how standards can really be developed for something like this.

I suppose that's my round-about and hopefully somewhat coherent way of saying that I think the best we can do is try to match history in materials, proportion, and heat treat and assume that by replicating the physical variables (and tested by destructive tests such as flexing until the blade sets, or cutting metal barrels) the variables of technique can be worked out afterwards.

I don't disagree with what I believe that you are trying to do in principle. A lot of current sword models being offered seem to have published statistics and dimensions that are close enough that one might assume direct comparison in test cutting is fair.

Difficulty arises in how one decides various swords have "similar swings." An actual cut is partly a linear motion, and partly an arc, depending on the technique, type of cut, and how one steps when they do it! The CoP of the Albion Chieftain versus Theign only seem to be a few inches (about 5 inches) apart, but comparative masses and techniques of these are worlds apart (speed, mass, etc. probably different by roughly a factor of 2) in terms of what really happens when people test cut with them.

An empirical based (kinetic energy transferred ) experiment might be the fastest and most realistic way of developing a quantitative measure of the energy available at a sword's sweet spot, as wielded on average by a proficient group of test cutters. Imagine batting a cut resistant, tire tread armored pendulum and measuring how far it is displaced upon impact. Anyhow, if you can approximate how much energy is available at the test cutting "sweet spot", then you can move on to the next problems of evaluating efficiency in cutting versus plain deformation (either of target or rolled sword edges.)

I would think that the "hard target" of interest would be fresh bone. Flesh (meat or wet rolled newspaper, etc.) has not proven prone to deforming swords that I have tried out. Opponents may have tried to bludgeon each other to death by bashing on helms, etc., but I doubt "efficient cutting" through metal or wood was a standard historical expectation. I have found it hard to cut through more than 3/4" of hardwood (solid limbs surrounded by a couple of inches of wet newspaper.) Incidental contact with metal (garage door rails, garbage cans in case of missed targets, steel barrel rims, etc.) has not really been a big problem in terms of rolling my own swords' edges. I guess you have to define what you expect to cut through, any why it is reasonable to expect your period type of sword to have done so.

Depending on what is looked for, we may not even have to test the cutting performance at all.

If we want a sword that could be used in combat, then in my opinion the best way to do it is to reproduce the properties of the historical originals. These properties are far more objective and easier to measure than cutting performance, even though such detailed stats are almost never published.

Then test-cutting can be used just to check that the sword is solidly build and that it endures the impact reasonably well, but to use test-cutting itself as a criteria is dangerous as it could introduce a bias: swords are a compromise between impact performance and handling performance, and often the handling properties are overly important. A sword is rarely be the best tool to cut through a given static medium... It is designed to be able to cut sufficiently well in a rapidly moving, self-defending medium ;)

As to the difficulty of what you want to do, assuming the swords are designed for a sufficiently similar context:

The choice of the cutting medium is not obvious. Ideally several would have to be tested, with various amount and type of armour, various weights, etc. This multiplies the experiment to be done quite rapidly.

The human factor is of considerable importance. A given person will cut more or less well with different swords, because the techniques have to be adapted to the sword. Personal taste is also a factor here... And of course different persons will obtain different results. Obviously you can think of suppressing that by designing some sort of machine. But then the problem is that the machine will not adapt to the sword nearly as well as a competent human in general.

The cutting performance is ordinarily conceived as depth of cut, and that depends on the sharpness of the edge and the thickness of the blade at the very least. So measuring the energy as Jared suggests is good, but not sufficient. And pushing the ball away may be best done with certain techniques that would not cut.

Choosing the point to test along the blade is not obvious either. Depending on the properties of the target, the optimal spot for cutting moves around, and the versatility of a sword is also important. Why having an edge so long if you use only a small portion?

In the end, I think it would not be practical to do on an interesting, varied sample of swords. Too many parameters...

On the other hand, testing a sword's strength is possible. Once you have a very precise type of blade, it is possible to check for defect using basically a powerful impact to see if the weapon holds together. Such a method is evoqued by Henry Wilkinson in this text, unfortunately I do not have any details on the practical implementation of it. Note how this is a check for toughness and not really power, which is provided by the machine. Note also, for completeness, that the explanation of CoP is not physically correct at all :)

Anyway, it all probably sounds very pessimistic and discouraging, but if you believe you have a way to go around these problems by all mean go for it! I know I'm stubborn with my own ideas and I mostly benefited from that :D

Assuming good and provable quality control destructive testing of a representative sample swords should give some idea of what kind of damage a similar sword will take and continue to be useful as a sword.

When does one decide that that the sword has been ruined for use ?

One can be very strict and count aesthetic damage as enough to ruin the sword: This might be the criteria for collectors or people doing cutting exercises where their sword should have no more than very easy to clean up scratches or very minor dings that can be sharpened out.

For the " pure " collector any damage is bad !

For real fighters/soldiers a weapon is still good if it can still do the job and that the cumulative damage hasn't made him worry about it failing: Aesthetic considerations being very low as a priority even a blade looking like a saw is still good if cleaned up so as to minimize stress risers. Today a collector/shooter with a personally owned M16 may view it as in bad shape if it's all marred by scratches and dings but a soldier worrying about surviving will only care about does it shot when I want it to shoot reliably and after all " it isn't mine anyway " If the soldier is smart he will clear the gun and maintain it in good working shape. He or she may still avoid careless damage even if cosmetic damage to the weapon but won't obsess about it I think?

One thing is metallurgy: A specific type of steel properly heat treated to a specific hardness is going to have a hardness and flexibility that is predictable and test done on one should be representative of all treated the same within a narrow margin of error. Steel hitting something harder than it is, is going to notch or have the edge roll. Steel hitting very hard materials that are softer will be dulled and may roll at the edge but will probably not be notched unless too hard and in that case the edge may chip.

Except for sword on sword hard contact most things a sword would hit would be softer: Steel armour, polearm languettes, steel or iron rims on shields, leather and wood and flesh and bone. ( If one misses and hits a stone wall or a large tree trunk edge damage or blade bending or snapping is a high probability ).

Handling is mostly a matter of mass distribution independent of the metallurgical qualities that only become important on impacts.

I'm talking about a modern standard, but one which could be informed by historical parameters. And let's say aesthetic damage is allowable, but not structural damage.

The above responses show the complexity of the subject, which is all the more reason for setting some (perhaps arbitrary) standards. For example in experimental science one often studies complex systems with many degrees of freedom. To make the problem manageable, it is desirable to hold all those parameters constant except the ones you are interested in (the independent variable and the dependent variable). If one does not do this properly, it leads to a 'confound' and the results are not publishable. Things that cannot be controlled (like random noise) one deals with through repetition and statistics.

So, to compare one sword against others one should hold all other factors constant, or to test one swordsman against another one should do the same.

It's not just science geeks who do this, it's everywhere in sports. If you watch a professional hockey game, one takes it for granted that each team has the same rink dimensions, net area, puck size & weight, etc etc., so we can focus on what we are interested in; skill and teamwork. The parameters don't change for every team and every game. Same can be said about any professional sport or Olympic event. It's part of what makes these 'serious' sports.

So perhaps to focus this topic, I can ask the question in a different way. Suppose there was a sword cutting event at the next Olympics (how cool would that be?). The events are as follows:

Cutting in the one-hand, double edged sword, 2.5 pound category.
Cutting in the two-hand, double edged sword, 3.5 pound category.
Thrusting in the one-hand, double edged sword category (where the sword must also meet some minimum cutting standard so people don't use sharpened bars).

The measure in each case is depth of the cut or thrust.

So now the question is, which media (and what dimensions) would you choose for this Olympic event?
- should push the limits of the sword without causing structural damage
- should be readily available to the average person so the community can adopt this standard
- should have dimensions that lead to easy measurement of performance
- should be something easy and SAFE to set up

I have a personal interest in this because I would like to standardize my own tests, but one can see where this would help lead to western martial arts being taken more seriously as a sport.

-J.D.

PS, it would also be cool to have Olympic historical fencing (seriously) but that's another topic.

Oh, but keep in mind that sport equipment tends to become over specialized on it's particular application.
Just a silly example: A pair of soccer shoes are so different from "regular" shoes that they may be quite annoying to use in every day's life.

Some fencing swords, for instance, may fall into sports category... but an historically accurate war sword?
Sadly as it is, and besides everything a sword may represent, it is first of all a weapon. Originally born and meant to do harm to others.
Here lies a great problem with standardization. How can someone reliably measure how well a war sword manages doing harm to others in combat?

So, as you clearly said in a previous post, such standard to measure a sword performance has to be inevitably arbitrary from the beginning.
And it should aim to provide just some ideas about a given sword's capabilities, because aiming to reliably and universally measure how "good", historically accurate or valid a sword is, would be a mistake.


A question for the most knowledgeable ones:
Are there no historical documents stating "what a good sword should be able to do"... or something of the sort... just to consider as a reference of the contemporary vision when they were still in use on the battlefields?
I vaguely remember having read something like that about katanas, but it does not apply to this post. And my memory is really lousy.

Thanks!

I like the idea of this kind of standard, but establishing standards and doing the testing it will be nigh on impossible. Here's how I see it:

1) Create a mechanical testing rig with variable speed/strength of cut/thrust. The variability will be necessary when first dialing it in. The human factor is too unscientific to be used, so a machine is necessary. Test a large number of humans to find out average or typical force and speed. Use that to calibrate the machine.
2) Find a cutting medium. Ideally, I'd want to see a medium with gradations of toughness, again to help dial things in: try an easy target; if the sword passes, go to the next toughest target and try again (and again) until you establish a benchmark.
3) Test antiques. The only way to really judge a replica's performance is to test as many antiques of varying types as possible. You'd need to be able to figure out averages by type/purpose, so you'd need a large sample set of antique swords whose owners will let you strap the sword into a machine. Good luck. :)

I think there is a ton of value to this system, which would give a scientific look at durability. But I don't ever see it happening. I think that without knowing what antiques were capable of, it's nearly impossible to set up a scenario to test replica swords. Since collectors and museums likely won't let their antique swords be used in such a fashion, I doubt this will ever happen.

To drive this point home, there is a maker at a local Ren Faire who likes to show that their swords can break concrete blocks. I'm positive that test is invalid because an antique (designed for real use by people with real practical knowledge of its effects) wouldn't pass that test. So we need to know the limits of antique swords before we can evaluate their replicas.

It's easy to say that modern steels and heat treat make many replicas better than antique swords. However, we don't have enough scientific data to really know if softer, lower carbon steels had different properties in terms of vibrations, etc. Without a thorough study objective of antiques, any rating of modern replicas will continue to be an educated guess at best or wild speculation at worst, both of which are subjective.

A couple of quick comments fwiw. In the 19th c, when folks still occasionally relied on blades for their lives in the melee, military blades were routinely tested in bending and torsion to prove a blades temper before assembly. The sketches of the tests that Wilkinson used are easily found on the internet and in books. These are the kinds of tests that are relatively straightforward although each blade type will have its own tolerances. today, we also have a wide array of non-destructive tests that could be used to examine a blades steel properties, with the caveat that unlike steel used in construction where uniformity is generally a plus (and tests are design to spot abnormalities), the steel in a blade generally has variable properties over its cross section depending on its heat treatment prior to final grinding. These kinds of tests can help to assess "durability" the way I think of it, and I believe that is what K. Cashen is saying......

As for cutting, there are just so many variables (edge thickness, edge geometry, bevels, tapers etc) involved that standards imho would be almost impossible to develop.... not to mention that different types of blade edges are better at cutting different media, I have found there is no "universally good" one size fits all edge when it comes to cutting different media. Swords are tools in that regard. Some work better in certain applications than others and vice versa.

I say almost impossible as over the years I have developed my own personal test of blade edges that enable me to readily assess a blades performance before I swing it at the typical "soft" targets we use, i.e. mats, bamboo, noodles, rolled paper, bottles, cardboard etc. generally, I can tell by running a soft medium like a pool noodle fragment over a stationary blade edge as to whether it is even worth swinging it at a tatami mat or other similar soft target or whether the edge needs more work before doing so. The amount of pressure applied during this kind of test is obviously subjective but generally I can tell very readily whether a blade edge will perform well or whether it needs honing. Obviously a dulled edge does not make for a failed sword, just a sword which needs a little work and it is true that to a certain extent additional energy input can overcome a dull edge at least when cuttng soft media. But over time, even with soft targets, edges will dull if used routinely. Scotchbrite pad scrubbing which some folks seem to routinely do with their "workhorse blades" in WMA will also readily dull an edge.

just my $0.02. TR

p.s. you mentioned thrusting ability. That would seem to be something that probably could be standardized if one really wanted too (required axial pressure to penetrate a certain distance), but again the geometry of the blade would probably be a bigger factor than the actual steel properties of a sword. (i can shove a tine of a pitchfork readily into something - probably even easier than most swords - but that doesn't tell you anything about the quality of the steel)

I think this is a fair question.

I have spent time testing the durability of swords in the sub $300 category. My test is purely on durability but I do offer some personal opinions as well on sword performance. The durability test is objective. I strike the sword 2,000 times on a tire-pell over a three month period. If it doesn't break it passes. I have tested 11 swords. Two failed. One broke in 30 blows the other in about 200. I also broke a third one which passed the 2,000 hit test but broke a year later after about 30,000 blows. It was one of my oldest. Still I have others that have lasted past 30,000 blows. Any comments on performance or handling I offer are subjective since everyone has their own personal taste in that respect.

Here is a link to the website:

http://mysite.verizon.net/tsafa1/swordreview.htm


Why did I pick 2,000 hits instead of 100 or 10,000 ???........It seems reasonable to me personally.

How hard do you have to strike with a sword???...... That too is a personal issue. I imagine a worse case scenario where a mounted knight has just broken his lance and pulls his sword. He strikes while charging at 30 mph adding the horses momentum to his sword swing. He is striking into another moving target which is covered with a surcoat so he can not see the armor underneath.

I think that the wrong questions are being asked, and being asked for the wrong reasons. As collectors, we are very temped to imagine swords that can survive our abuses and still look as pristine as the day we bought them. I have put a lot of swords, most of them my own, through some very abusive test that are as close to a simulation of what the historic orignals would have gone through as I can conceive and replicate and have up come up with a set of better questions (at least I think so).

1. Can a sword do the job it was designed for? e.g. is a XV still a XV if its point is not acute enough to slip into a 6mm mail ring?
2. How much is required to bend the sword? Does it bend too easily?
3. When a sword bends, can it be easily and quickly straightened?
4. Is the edge too brittle or is it soft/strong enough to fold and nick rather than chip?
5. How hard is the edge to repair? To reprofile?
6. How sturdy is the hilt? Do I have to keep peening/tightening the hilt after every use? If the peen comes loose, does the rest of the hilt fall apart or does it stay together?

There is no such thing as a sword that won't take a set...there are only swords that haven't taken sets yet. There is no such thing as a sword whose edge will never be damaged, only a sword whose edge has not been damaged yet. I have had very hard swords that are difficult to bend and whose edges resist rolling, but those same swords are extremely difficult to repair. I have come to prefer swords that are on the soft side..easily repaired and straightened. They perform the same as the harder swords, require sharpening just a tad more often and here's the fun part...bend/nick under the same circumstances as the super hard, super "strong" swords, but are much more easily repaired.

These posts represent exactly the range of opinion that I was referring to at the beginning.

On the one end of the range we have criteria like "2000 hits on a tire pell".

On the other end we have the very complex criteria pointed out in the last posting. I agree these are all important things in choosing one's own sword, but very subjective and hard to communicate to someone else with different ideas and tastes of how a sword should behave. Moreover, qualitative impressions are easily colored by other factors.

I have seen Vassilis take a lot of criticism for his method, but from a scientific standpoint it wins out in the sense that at least he is gathering empirical data and reporting it in a quantitative fashion. (Although we would call it anecdotal, since sword durability may vary a lot from one unit to the next, especially in the range of swords that he has tested). However these durability tests do not tell me other basic things I want to know, like how well the sword cuts.

I started thinking seriously about this after reading the following review:

http://swordforum.com/summer99/dt2130.html

Here Gus Trim launched two swords several times each into a piece of plywood, and reported the depth of the cut for each swing. Not a ton of data, but the results were consistent enough (and presumably his swings were consistent enough) to convince me that one sword was the better cutter for that medium. This is the sort of thing that I would like to see more of, instead of statements like 'the sword cut well through both light and medium targets', which really tells me nothing except that the reviewer liked it.

Now, if everyone started using a few standard criteria, then we could all get on the same page, at least on the basic points. Just like CoG gives a solid clue, but never the whole story, for handing characteristics. The subtleties of a sword would still require a more qualitative review, but we would at least have a few common points of reference.