Mask mesh testing - there must be a better way

[PeterGustafsson]

Hi!


From the "Video of how to do a mask punch test" thread:

Originally Posted by Veeco

I'm still unsure about the whole mask testing thing myself... If the test can damage the mask, when done too often, I'd prefer it not done to my mask.

Veeco gets directly to the heart of nondestructive testing (NDT) in the sentence above.

Before I start, I would like to point out the difference between this and my previous thread on mask testing. In the previous one, I asked for the ramifications of a new mask test procedure being used, without prior express acceptance from SEMI or other relevant bodies. In this thread, I deal with the technical properties of such a test procedure, as desired by the f.net community.

Also, I would like to point out that I do not disapprove of the testing done today - it is a whole lot better than no testing. What I am going for is testing that get the same information, but with fewer (preferably no) bad side effects.

In general, testing of mechanical parts and items are either done nondestructively or destructively. The pro of the former is that the tested object can be used after testing, while the pros of the latter are that it is (generally) cheaper, yields more info, and is easier to interpret.

(Concerning destructive testing: The rules specify destructive testing procedures for masks, in order to homologate them before a mask model can be sold. I personally think that those test procedures look good, but there is some room for relatively minor improvement.)

There are many types of NDT, but they all share one thing - they do not degrade the test object to any appreciable degree, hence their name. Destructive testing, OTOH, will wreck the test object so that it is instantly obvious that it is rubbish, and can not be salvaged.

Mask punch testing (MPT) falls into a category of its own - not intended to be destructive, but can easily be so. As an engineer with some experience in mechanical testing, I can say that if a test engineer would have come up with a test of that sort it would not have been considered good enough, and the search for an acceptable test protocol would have continued.

MPT seems to combine the worst of NDT and DT - the test object is subjected to appreciable stress, no information is found about nontested places, results can be ambiguous if not observed properly, and the test procedure is not standardized between performers.

So, this thread is about how a better mesh resistance test might be designed, and the technical details thereof. How should one design a test which can find out of the threads in a mask mesh will separate too easily? What should such a mesh tester look like, and what physical phenomena should it use while testing? How can one test the mesh without actually subjecting it to loads which will destroy it if the test is done in the same place in too few times? If one guesses that a mask will undergo at most 100 test procedures before the owner replaces it for other reasons (wild guess there!) one would want to have a test which can be done 100 times in the same place of the mesh without noticeably damaging the mesh, while at the same time yielding the desired information. At the same time, the test device should be non-static, easy to use in a standardized way with very little training, and not exorbitantly costly.

The current MPT is cheap, very portable, easy to design. Alas, it can easily be used in the wrong way, and does not fulfill the other desiderata stated above.

Any ideas on how to construct this wonderful thing? I have some ideas, but I do not want to steer the brainstorming in any particular way.

There is also another type of classification of test procedures other than NDT vs. DT - whether they test one place or many. In that regard, one can classify test procedures as either:
1. Point tests. They obtain a test figure for only one very small part of the test object, and a new setting must be done to perform the test in another part of the test object. MPT is an example of this.
2. Scanning tests. At any one time, the test procedure examines a part of the test object, but it is easy to change that part without any resetting. When Sam held up the mask against the light to check for wire dislocations he was doing such a test.
3. Global tests. These test examine the whole test object in such a way so that one number is obtained from one test setting, which is indicative of the whole test object. However, if such a test indicates a failure, it does not say where the failure is. Conductivity testing of a sabre mask bib is an example of this.

Many test protocols are designed so that the test objects are first subjected to some global method (or quick scanning method), which has a threshold set so that one will not get any false negatives, but several false positives. The test object which pass this test are good (to a quite high degree of certainty) while those that fail the global test are subjected to at least one type of point or scanning test. Those tests are designed to pick out the really failed objects, while indicating that the false positives are good. By this combination, a test protocol can have low numbers of both false positives and false negatives, while having a reasonably good throughput. When Sam first held up the mask and then subjected it to MPT, he in effect did just this.

A good starter on the topic of nondestructive testing:
http://en.wikipedia.org/wiki/Nondestructive_testing


Have a nice time!

Peter Gustafsson

[KD5MDK]

Given that testers have no information about the quality or condition of the materials involved, it seems extremely hard for me to picture a strength test which doesn't involve stressing to a designated point the structure.

[SJCFU#2]

Originally Posted by PeterGustafsson

There are many types of NDT, but they all share one thing - they do not degrade the test object to any appreciable degree, hence their name. Destructive testing, OTOH, will wreck the test object so that it is instantly obvious that it is rubbish, and can not be salvaged.

Mask punch testing (MPT) falls into a category of its own - not intended to be destructive, but can easily be so.

Peter,

While I agree that a better test might be preferable to the MPT (especially if it looked at all the mesh rather than just select points), I do not agree with placing it in a category of its own. There are a number of forms of NDT which apply a load that could result in degradation of a defective test article. Examples include hydrostatic testing of pressure vessels and pull testing of lifting appliances (in the latter case failure is based almost entirely on permanent deformation). I feel that the MPT fall into this same category - especially in the case of FIE masks which are supposed to be able to withstand a significantly higher load (25 vs. 12 kg), but can degrade over time.

[D'Art]

Quick thoughts of mine, whilst I'm on a break from decorating-

One simple way would be to use something like a cutlery polisher that many restaurants use. (We hire ours at a cost of about £20 a day) Basically, all it is is a shaker with grit in it. The masks would be stripped of all remaovables, and put in the shaker for say, 10 seconds and removed. A quick and easy way to tell if the mesh is coming apart, as, assuming a small enough grit was used, bits would get trapped where the mesh was failing. The con side would probably be the cost of obtaining a large enough one for masks to fit in.

Anything which passes this could then be passed on to some sort of table-top press machine which would be a mask shaped bowl and has two pressure settings, one for FIE masks, and one for non-FIE masks. A quick squeeze and check for deformities. I realise that this isn't an NDT, but still far quicker and less open to stupidity than the punch test you lot use. Again, though, the cost would likely be prohibitive.

Pass both, and you would be as certain as it's possible to be that your mask is safe.

Back to the painting now....

[Rick Thompson]

One simple way would be to use something like a cutlery polisher ... The masks would be stripped of all removables, and put in the shaker for say, 10 seconds and removed. A quick and easy way to tell if the mesh is coming apart, as, assuming a small enough grit was used, bits would get trapped where the mesh was failing.

There are several problems with this test. First, the sand in this would be too abrasive to the lame' material on a sabre mask, and would probably be WAY too abrasive for the painted mesh on a foil/epee mask.
2nd, the weak points in the mesh aren't always evident because of the cracks between the wires - which is where the sand would stick. Weak mesh solder joints can be caused by stress which does not necessarily break the joint, crack the paint, etc.

Anything which passes this could then be passed on to some sort of table-top press machine which would be a mask shaped bowl and has two pressure settings, one for FIE masks, and one for non-FIE masks. A quick squeeze and check for deformities.

This doesn't necessarily test for anything useful. A mask may be able to bend in several dimensions and still pass all requirements - in fact, fitting a mask does just that.
Also, stressing the entire surface of the mask, while a significant test, doesn't address the real problem. A mask doesn't fail against a blade over a large area - it fails at the point of contact, which is pretty darn small. The punch test addresses this issue more accurately than a whole-surface test.

The punch test, while not inherently a destructive test, *IS* one when improperly applied. One of the major causes of improper is "gouging" the probe - moving it around while attempting to insert it, or pushing it in at an angle (which may result in a false positive - passing the mesh joint when it should fail).

One modification to the punch test might be to put extension "legs" on the tester, or a wide collar. This would sit firmly against the mesh while the probe is being applied between the mesh joints. The problem is that the curvature of the mask is not consistent, so the collar/extensions would need to be easily adustable (or spring-loaded, or something) to ensure a firm seat against the mask when the punch is initially set against it.

Another testing problem with the visor masks, as mentioned in another recent thread, is that a blade can penetrate the intersection of the visor plate and the mesh. There is no current test for this at the armorer's table (beyond the visual inspection).

[Fencergrl]

Originally Posted by KD5MDK

Given that testers have no information about the quality or condition of the materials involved, it seems extremely hard for me to picture a strength test which doesn't involve stressing to a designated point the structure.

Keep in mind that metal fatigue, denigration of metal materials etc... get tested around the world all the time. Airplanes, pipelines, boats, turbines, etc... they all require scheduled testing. It's about keeping a large number of people safe.

So as Peter mentioned, there's Non-Destructive Testing available. I used to work for a supplier who sold all types of NDT equipment and supplies, so I know a little about the different methods available and their costs.

Where I see a problem with this, is finding an inexpensive means of doing this. First off... many NDT techs will not get out bed in the morning for less than $100 hour. This may seem like a lot of money, but you've got someone who crawls into dangerous, uncomfortable places for a living with very expensive equipment.

So lets say, they'll do it for a bit less because you're not asking something dangerous or uncomfortable. The equipment and materials he uses to do the testing is very expensive. He needs to use very precise equipment for measuring the thick and thin areas of metal to find the weak areas.

This is not something an armourer can use. I suppose the FIE could train a select number of armourers to travel around the world. They could invest $1000s of dollars in test equipment.

The truth is, is cheaper to use a destructive means of testing and let the fencers replace their damaged masks.

[veeco]

Not for the fencers it isn't...

[InFerrumVeritas]

Originally Posted by veeco

Not for the fencers it isn't...

Well, I assume that training these "super armorers" and providing their equipment would be expensive on the people running the tournaments, especially if they have to be flown around the world. These costs would probably be translated to tournament costs and membership dues, so yes, it would be.

[Fencergrl]

Originally Posted by veeco

Not for the fencers it isn't...

Yes and no. Yes it's expensive replacing masks that are damaged by testing (especially when it is done incorrectly).

On the other hand, where would the money come from if the FIE or USFA had to send armourers for specialized training and buy expensive NDT equipment and supplies? Your wallet my friend... suddenly you will either have a requirement that all masks be tested every 3-6 months and the fencer would be charged for this testing.

The truth is, our existing method is probably the most efficient and economical means of testing despite it's inherent flaws.

[PeterGustafsson]

Hi!

Originally Posted by Fencergrl

The truth is, our existing method is probably the most efficient and economical means of testing despite it's inherent flaws.

This sounds depressing...

Anyway, despite your credentials, I still want to hope for something better.

If we can not design a better test for the present mask mesh type - why not try to imagine another design of the mesh itself, which would be wear-obvious? If it is designed in such a way so that shortly before it becomes dangerous, it changes it appearance in some obvious way (preferably in a step-wise fashion) so that it is easy to test that new mask mesh design?


Have a nice time!

Peter Gustafsson

[Gav]

I'd like ask a another question.

I'd like to see a breakdown of injuries and deaths caused specifically by mask failure.

Before embarking on a round of chin stroking (and possible mask destruction) I would really like to know whether this is a solution looking for a problem.

[Rick Thompson]

The truth is, our existing method is probably the most efficient and economical means of testing despite it's inherent flaws.

Well, the existing APPROACH may be the most effecient & economical. This doesn't mean that it can't stand improvements. The mask punch device is currently $70-$100, and can last for years. If the cost of the mask punch device were upped by $50-$70, it would still be a bargain.

Gav asked if this weren't a solution in search of a problem, and asked for a statistic on mask-failure related injuries. Like many athletic activities, only a portion of the injuries might ever be officially reported - injuries sustained at a local club would probably never make it into the statistics that are recorded at USFA tournaments. And the USFA will know of those injuries only IF the tournament directors report it. And the near-misses, where a mask might fail, but there's no injury would probably never be reported anyway. All that being said, ONE injury is too many.

More accurate, efficient and easier testing methods would go to IMPROVE the safety factor, and it's hard to gainsay that. How many times have you been to a local tournament where one of these conditions exist?
A) the person testing the masks has just been recruited by the tournament organizer, and given a 2 minute lesson on how to inspect masks.
or B) there is NO mask inspection.

Improving the process, making it simpler and more accurate to test, would be to everyone's benefit. And this would much easier than to get the SEMI commission to approve a new and easier-to-test design of mesh.

So - we go back to Peter's initial question. Here are some specific questions, which might provoke some enterprising engineering types:

1) Given that the mask punch is a valid means of testing the strength of the mask mesh joints, is there a way to change the design of the mask punch device to reduce operator error?

2) Is there a means to identify manufacturer's or brands, or runs of production, that are suspect (e.g. some manufacturers who do not meet the FIE standards on all their brands of mask).

3) Can there be produced a cohesive training guide for inspection, so that there is a standard by which to train armorers/inspectors? (given that we're not going to fly in experts at every tournament to do this inspection).

4) The mask punch is currently used from the outside going in. Is there a means to perform the inspection of the mesh joints going from the inside-out? Would this yield different results or make the process easier (since the visual inspection usually involves looking from the inside of the mask.

5) Are there statistically areas of a mask that are more likely to fail, and should therefore be tested first or more thoroughly? (e.g. some manufacturer's masks, known to dent near the top front of the mask, are more prone to fatigue there. )

6) Are there rules or standards for the visor masks? How do you test the plexiglass (or whatever) visor? How do you test for the spacing between the visor & mesh?

Should this thread be moved to the armorers section?

[PeterGustafsson]

Hi!

Originally Posted by SJCFU#2

Peter,

While I agree that a better test might be preferable to the MPT (especially if it looked at all the mesh rather than just select points), I do not agree with placing it in a category of its own. There are a number of forms of NDT which apply a load that could result in degradation of a defective test article. Examples include hydrostatic testing of pressure vessels and pull testing of lifting appliances (in the latter case failure is based almost entirely on permanent deformation). I feel that the MPT fall into this same category - especially in the case of FIE masks which are supposed to be able to withstand a significantly higher load (25 vs. 12 kg), but can degrade over time.

I expressed myself incompletely. MPT can result in the test object being noticeably degraded, and in a way not obvious to the naked eye. The tests that you list should behave in such a way that loading sufficient to cause degradation (which I take to mean plastic deformation) will do so in such a way that the deformation is either visible to the naked eye, or at least relatively easy to measure.

Hope that clears up matters a bit.

(Note to self: proofread at least thrice when posting big threadstarts! )


Have a nice time!

Peter Gustafsson

[great bowyer]

It has been quite a while since I have been on fencing.net, but after seeing all of the posts about mask failures and testing I decide to post a few comments.

I do agree with all of you that mask safety is most important. If there is any question as to the safety of the mask, I feel that that mask should not be used. With this in mind though, I think that the current mask testing is not able to tell us the exact condition of a mask. I remember several years ago there was a post about the mask that failed and was on display at the nationals. If I remeber right, after the mask failed it was tested and it would have still passed if it were not for the hole.

It has been a while since I have worked out any physics problems, but if I am looking at the problem correctly then the way the current mask tester is designed, the spring is calibrated so that 12 Kg are needed to hold the probe in the depressed position. There needs to be a force that will then move and actually accelerate the movement of the tester to the depressed position. It seems to me that the faster someone pushes down on the tester the more force is then going to the probe and into the mesh causing a possible failure.

As for a suggestion of a possible fix, I am not sure. The best idea that I can think of at the moment is maybe a light sensetive liquid spray that is sprayed onto the mesh and held under a special light that will illuminate any fractures in the mesh. This does not sound very cost effective though. I think something like a extensive test like I mentioned above done every 6 months or so combined with a peridical punch test could be effective.

On the other hand though, it seems to me that fencing is a very safe sport the way it is. If you look at the stats, then it is far more dangerous getting to the gym to fence than the actual act of fencing.

[veeco]

Originally Posted by Gav

I'd like ask a another question.

I'd like to see a breakdown of injuries and deaths caused specifically by mask failure.

Before embarking on a round of chin stroking (and possible mask destruction) I would really like to know whether this is a solution looking for a problem.

+1

I've never had my mask tested, at any event in Europe. I've never heard of or seen any catastrophic mask failure in Europe. Only the United States feels compelled to test masks, and that's the only place I've heard of catastrophic mask failures.

So either the punch tests administered in the US are in fact weakening the masks, or Americans don't know how to properly care for their equipment, or Americans have to stop buying cheap crap from China...

You'd have to think, with the large number of people who fence everyday in Europe, compared to the relatively low number of fencers in the US, that the incidence of catastrophic mask failures would be much higher in Europe...

[Rick Thompson]

So either the punch tests administered in the US are in fact weakening the masks, or Americans don't know how to properly care for their equipment, or Americans have to stop buying cheap crap from China...

I think that's something of an oversimplification. I trust that was intended to be humorous....

You'd have to think, with the large number of people who fence everyday in Europe, compared to the relatively low number of fencers in the US, that the incidence of catastrophic mask failures would be much higher in Europe...

Or the reporting of incidents in Europe is lower (for any number of reasons), or the European coaches/fencers toss out their masks more readily when they see that it's getting old, or they get less expensive masks and are less concerned about doling out for a new mask....

There are too many variables & unknowns to arrive at a definite conclusion.

[Rick Thompson]

Great bowyer wrote:

It has been a while since I have worked out any physics problems, but if I am looking at the problem correctly then the way the current mask tester is designed, the spring is calibrated so that 12 Kg are needed to hold the probe in the depressed position. There needs to be a force that will then move and actually accelerate the movement of the tester to the depressed position. It seems to me that the faster someone pushes down on the tester the more force is then going to the probe and into the mesh causing a possible failure.

Take a look at the mask punch test and how it works. It's NOT an impact test. It's a conical probe intended to stress the mesh joints with lateral pressure, using a 12kg spring to apply a controlled amount of force. Since there is a maximum distance between the wires (as specified by the FIE), if the conical probe is able to spread the wires to the point that it can fully penetrate, one or more of the solder joints on the wires have failed.

If you've seen someone actually using the punch test to PUNCH a mask - yeesh... give me shivers....

An impact test would be subject to MUCH more variability, and to abuse. Not to mention the issue of damaging the wires and/or chipping off paint.