Why “Bounce Testing” should be avoided

(Just moving my response to the safety standards here so it's not clogging up the thread)

federal and industry regulations specify that slings that support a shock load should be immediately taken out of service.   A proper proof test is a smooth pull to 2x working load.   A “bounce” test would immediately render the sling unusable due to its being too unpredictable.   Apparently OSHA, ASTM, ASME, and rigging companies base their best practices on junk science and bullshit just like I do.

Industrial rigging and safety standards are TOTALLY different than climbing standards.  It doesn't make sense to try to apply random rigging standards to climbing.  Here are some examples of other industrial rules (edited formatting for clarity):

• Slings must be immediately retired if they have a knot in them (ASME B30.9). 
  • Retire every rap anchor instantly because it's tied with a knot. (https://www.certifiedslings.com/load-limits/inspection-of-slings, https://www.osha.gov/laws-regs/regulations/standardnumber/1910/1910.184)
• Carabiners must be rated to 16 kN when cross-loaded (Climbing carabiners is 7 kN).  
  • Time to throw away all your carabiners. (satra.com/ppe/ANSIZ359.12.php)
• Harnesses (which must include chest harness) must be immediately retired after one fall.  
  • My proj just got really expensive! (https://blog.msasafety.com/fall-protection-harness-expiration-date/)
• Slings must be 100% proof tested to 2X the rated load and have a MBS of 3X the rated load.  
  • All climbing slings would instantly fail this test, even brand new. (1910.184(i)(8)(ii))

Just like we don't follow the OSHA/ASME rules for knots, carabiners, and harnesses, it doesn't make any sense to blindly follow their rules for shock load either.  Attempting to apply OSHA/ASME/ANSI workplace safety and fall protection rules to climbing gear is a ridiculous proposition, they are intended for a totally different set of circumstances.  For someone who has been continually ranting about how textbook and lab data isn't applicable, it's hypocritical to be suggesting that crane regulations are applicable to climbing.

The reason they have to immediately retire slings after a shock load (and knots, and all of the other things) is because it's IMPERATIVE that the sling is able to hold the load on the label with a safety margin.  That's because riggers use slings to lift huge weights overhead, and if you drop a 20,000 pound steel pressure vessel because a 20,000 pound  (WLL rated) sling broke at 19,000 pounds, you might kill a bunch of people.

Climbing gear is completely different.  Slings are rated by their breaking strength, not their working load limit.  We all know and accept that in real use, they will hold less load than the tag, sometimes WAY less.  That's totally fine, because we don't use them to lift metal with a crane.

(As an aside, I find it odd that someone with "25 years as a mechanical engineer often with a focus on safety systems" needed help to find basic industrial safety standards that every factory installation in the western world needs to comply with...)

This whole thing is a red herring.  You STILL haven't provided any data or sources for any of your original claims.  Now you're pulling out industrial crane requirements that somebody else helped you find.

Lol the personal vendetta continues. I’m not sure who is trolling who here.

My two cents: if there was a rap anchor that I thought was sketchy enough that it might fail bounce testing - by which I mean shock loading to a couple kN - I would just leave a sling. If I didn’t have any option other than rapping off sketchy tat, I wouldn’t be shock loading it at all.

Kyle Tarrywrote:

haven't significantly altered anything I've previously posted (except my original post in this thread)….,

Ok Kyle  

It appears that you are deliberately lying.   Both about what you have said, and most importantly altering and misrepresenting what I have supposedly said —like about the RMRU analysis.  It shows deliberate editing.   THAT is total bullshit and borders on sick/sinister.  

Then This last post above,  total La La land of irrelevancy.  A lot of space about nothing relevant or in contention and missing the point that shock loads create uncertainty in used materials.  Apparently to you it’s IMPERATIVE that cranes carry safe loads but not so imperative that climbers live?  and you’re now saying shock loads are fine for tat climbing anchors.  You were for it, then against, now for it again.   But you never said that.  You never actually say anything.  Just tell us how YOU would test a sketch alpine ice anchor and or tat bail anchor in the rain — no cop out with “I’d back it up”.  Yeah, of course… just stop the gibberish of posting every useless thing you search up after one of my posts and stay in context. 

I mention those ASME stds only because they are based on the same principles you’ve been fighting me on this whole time.  That materials become unpredictable through their useful life and shock loads are a bad safety test for such materials - and leave them more unpredictable. 

Now you’re proposing I don’t understand that industry and climbing stds are different and I’m advocating for OSHA regs at every rap anchor.  Ok Kyle

Now the reasoning for the stds must be flawed too.  
and Jason EL must be a liar and that never happened, because Kyle knows if someone just used it, it must be good.  That never happens in the real world.  Junk science.  Ok Kyle  

You’ve never said “here’s what I do and why…”  to advance the discussion.  Curious.  

Others have engaged usefully and made good points.   I sincerely appreciate good faith disagreements.    

Maybe it all boils down to what I said in the accident thread, that most climbers agree 99% but go nuts over the minutiae of that last 1% 

 like in this case, it’s likely what constitutes a “bounce” vs another static test.   Maybe that is the root of most contention.  To me, and maybe most? others, a “bounce” is a ballpark 2-3x shock load to the anchor.  Maybe even some intentionally try to “shock fail” the anchor thinking that’s a better test to weed out a bad one.  That is potentially a dangerous practice to tat and ice and that is why this thread was created and still stands.  Climbers dying sucks even though it’s rare, it’s not rare enough.  let’s work together constructively to align on a best practice and dispell myths.  Good controlled testing is safer,  bad uncontrolled testing is dangerous.

My position has evolved slightly to better find and fit a best practice for a team (at first I was myopically in a soloist context and forgot a team essentially gets 2 free tests) while never abandoning my original correct premise — that old stuff is not at all like new stuff, and shock loads are worse than static loads for a proof test (still not clear if you understand that or agree, and even weirder if you agree, why/what are you arguing so much about?).  

Dont needlessly add more bullets to the roulette chamber.  

I know you’ll likely break this out into a dozen quotes to attack, but I won’t respond.  Consider it “Ok Kyle”  

Unless someone adds interesting new info or raises a sincere question, I think we can finally put a fork in it.   Absolutely minimize shock loads in a test of sketchy anchors (tat  or ice).  If it’s not sketchy, do whatever floats your boat.  

Edit to Kyle.  as to data or sources, I don’t need to.  It’s in your own textbooks you claim you have.  It’s the basis for the standards.   Read them and step away from the keyboard.   Yoda gets it

Mark Pilatewrote:edit To Tom: Totally agree on proof testing. That is basically what we are getting at. Best proof test of a questionable anchor. But your argument also tends to take results and theory from the normal zone (green) which is very repeatable and predictable, and mis-apply them to non-uniformly deteriorated and altered material with a high degree of internal variables. Proof tests need to be carefully constructed and implemented. 

The test is plenty repeatable and predictable.   You can see some actual load measurements in this video I posted before.      https://www.youtube.com/watch?v=gq3_DfyHg1A   14:14 to 14:26  nylon sling generates 2.3 kN with roughly 160 lb climber and quite small bounces on nylon cord.      14:45 generates 2.3 - 2.7 kN  using spectra sling, still quite small bounces.  2.7 kN is 607 lbs, approx 3.7 times body weight.    2.7 kN is  times his body weight, double what can be reached when rappelling.

You keep speculating that the webbing now may have some crazy sudden loss.  If that were true, a bounce test to 2.5 kN will provide more safety factor than a milder test by the heavy climber on rap.  Again, while backup in place.

Taking into into account that 

1. Smooth static is less variable in material response than higher impact shock loading

2. known load is less variable than unknown load  

So, given the topic is best proof test for sketchy anchor would you still recommend 

A: unknown, higher impulse, field bounce test from sling  

B:  known weight, careful load test with rope

Which would you suppose is statistically safer in “filtering” results ?

My theory and concern is not that doing a pretest test horrifically weakens something every time .  Pretest is better than no pretest.  I’m trying to get to best/safest pretest, and known is better than unknown, and smooth is better than sharp to avoid the rare, but unwanted unfiltered catastrophe of the proof test.  

A bounce test is actually not a very severe "shock test."  It's only about 3.5 times body weight,  not a funkness test (yanking with a hammer).   Much of the test energy of a short fall is absorbed by your harness squeezing your body,  plus any nylon in the anchor. 

I take it your "smooth" test may be as low as 2 times body weight, which is not enough to overtest all rap situations.  Just moving around on rap can generate 1.5 - 2x bodyweight due to your motions.
Table of results at 12:00. https://www.youtube.com/watch?v=nr3YBDnOI8Q  Shows how the bounce test is a reasonable amount of overtest compared to the rappel.

......Always back up, but if absolutely forced to rely on sketch for some reason.. 

1.  Test smoothly and carefully with both partners backed up, 

2.  Heaviest partner raps backed up

3.  Last partner removes (big why?) and raps.  and use smoothly.   

 seems like a contradiction to say you're relying on sketch, yet you have a backup.   A backup usually means you're not relying on sketch unless you later remove it.  

Once you have placed a backup sling, you're right, why remove it?   - assuming you don't desperately need it later.    Maybe you have enough slingage to backup only 1/3 of the raps.

But maybe you backed up an existing long sling with some other gear.   Once you conduct a 3.5x bodyweight test on the existing anchor, you can be confident that you can remove your backup stuff that you may need later.

Thank you Tom. This entire thread was created  by some childish need to argue.  somehow he took offence to the fact that for about the last 30 years if I am in a situation where I am backing up a rappel anchor and wish to remove the back up because I might need it later or because its expensive I have  bounce tested the anchor to prove that it is worthy, then removed my back up and rappelled.  The decade previous to that I just removed the back up and prayed a lot while I tried to rappel as gently and smoothly as possible.  somehow mark felt the need to prove his superiority by ridiculing this method which INMOP has helped keep me alive  and sane for the last 3 decades... 

Nick Goldsmithwrote:

Thank you Tom. This entire thread was created  by some childish need to argue.  somehow he took offence to the fact that for about the last 30 years if I am in a situation where I am backing up a rappel anchor and wish to remove the back up because I might need it later or because its expensive I have  bounce tested the anchor to prove that it is worthy, then removed my back up and rappelled.  The decade previous to that I just removed the back up and prayed a lot while I tried to rappel as gently and smoothly as possible.  somehow mark felt the need to prove his superiority by ridiculing this method which INMOP has helped keep me alive  and sane for the last 3 decades... 

That’s all anecdotal evidence, it has no place in science.

Nick Goldsmithwrote:

Thank you Tom. This entire thread was created  by some childish need to argue.  somehow he took offence to the fact …,

Sorry you feel that way Nick.  This thread was created due to a recent fatal accident and was not directed at you personally in any way.  No one knows all the details or sequences that led to the final failure there, but climbers bailing in the rain off old anchors is relatively common and many may not fully be aware of all the risks of “testing” it with higher shock loads.   And the effects are cumulative so that it becomes a ticking time bomb. Jason’s example is just one where it just held a load minutes before.  It happens.  Luckily rare, but we don’t have yearly accident compilation books for nothing.  

As for the need to argue, it is amazing, but not just on me.   Now we’re parsing the last 1% into .01% chunks about mild bounce or hard bounce.  I have no argument with Tom and his responses have been thoughtful.  

Have no idea why a narrowly focused technical debate has become emotional and personal for some.    Everything I have posted is true.  You can decide how to proceed.  For you, apparently expensive gear is the motivation.  You do you.

You will not find any experienced engineer from gear manufacturers, industry, or government that will disagree that,,,,,

1. Old deteriorated webbing and waterfall/alpine ice “on the edge” ESPECIALLY wet does not respond the same as new/good stuff and is highly variable in what your “test” outcome expectations should be.  

2. shock loads are more unpredictable that smooth static loads and should be avoided in such tests. It’s in the weeds and out of scope to argue about mild shock loads.  As I said, I do not have any real argument  with Tom and would say we appear to be 99.8% aligned…easy to reach consensus. 

I have continually modified my own thinking as the discussion progressed and admitted errors in logic.  But the technical takeaways stand and are bolstered by acceptanced engineering practice. 

As stated in OP.   Goal was best practice.  If you want to take offense to it, I don’t know why. 

We should try to find a way to combine the heavy breathing and willful ignorance in this thread with the fallacious arguments and dipshittery in the dog rant thread so as to create a wormhole thread that will somehow eradicate bounce-testers and irresponsible dog owners.  Let’s get on it, “engineers”.

Mark Pilatewrote:

As stated in OP.   Goal was best practice.  If you want to take offense to it, I don’t know why. 

And the "best practice" YOU came up with was bounce testing or "testing to 2x the static load" whatever you want to call it. Jeeze man, give it a rest

Ian Lauerwrote:

Jeeze man, give it a rest

The title should have been:

Why “Rappelling On Super Sketchy Tat” should be avoided.

And the thread should have gone like this:

Because it can break and YGD.

</end thread>

I am not advocating the use of tat for load bearing but weird situations do occur. I can think of cases where I might bounce test non placed gear such as a bush or small tree. It's not a very scientific load application but if you can apply a force of at least twice the expected load without damage, I think there might be confidence that the anchor will survive. I still don't understand why this sort of test can't apply to webbing slings.

OSHA section 1910.184(i) relates to synthetic web slings. OSHA allows repaired slings to be returned to service if they are proof loaded.

1910.184(i)(8)(ii) Each repaired sling shall be proof tested by the manufacturer or equivalent entity to twice the rated capacity prior to its return to service. The employer shall retain a certificate of the proof test and make it available for examination. 

One problem with climbing gear is that there is no rated capacity or application of factors of safety in any sense that is familiar to engineers.

At any rate, I have not concluded that the OP is wrong on any point but I would like to see some relevant research or data. Call me a skeptic but I value data over opinions.

Dan Merrickwrote:

I am not advocating the use of tat for load bearing but weird situations do occur. I can think of cases where I might bounce test non placed gear such as a bush or small tree. It's not a very scientific load application but if you can apply a force of at least twice the expected load without damage, I think there might be confidence that the anchor will survive. I still don't understand why this sort of test can't apply to webbing slings.

I think we are well aligned.  Again, for purposes of this discussion, I am defining a “bounce test” as what I understood to be the most common method of a short shock drop from static anchor clip in tethers like slings or PAS or similar that generates a 3-4 kN high impulse load.  

Doing a test is good.  Always test before using.  But not all tests are equal and some could add unneeded dangers.  Some tests do damage and create the issue they intend to screen for  

(Link below on general Proof test considerations.  Section 6 has dangers of bad or Poorly considered/mis-applied tests.  Many of these principles are how ASTM/OSHA developed their tests and guidance for nylon slings) 

https://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fntrs.nasa.gov%2Fapi%2Fcitations%2F19950010857%2Fdownloads%2F19950010857.pdf&data=05%7C01%7C%7C558361eb6bbc486205f708daafdd20b7%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C638015661543916695%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=2pPiVPEz9iv8fqFgi%2BMZPWoX4J6oa%2BuiZMLPOMllJX8%3D&reserved=0

Goal is coming up with the best test to maximize safety for those scenarios at the edge of the envelope like old tat and bad ice anchors.

Trees, bushes, anchor pieces (pro), aid are outside the scope of argument, due to being more robust against shock  — though a good test is still appropriate 

In general,  a 2x intended max load smooth static proof test  is better than a 3-4x shock load test to degraded, non-uniform materials  is all I’m saying.  Especially wet materials if bailing in the rain,etc.  the dynamic response capability (if needed in an emergency out of the blue like rockfall, a slip, etc) may be used up in the bounce, leaving no margin and weaker if an operational shock mishap occurs. 

OSHA section 1910.184(i) relates to synthetic web slings. OSHA allows repaired slings to be returned to service if they are proof loaded.

Yes, repaired if visible damage but not shock loaded slings.  And the proof test is slow static application   No shock testing allowed and not 3 or 4 x the load.  2x.  In industry,  “x” is 1/5 of design rating which would be more equivalent to climbing ratings.  

1910.184(i)(8)(ii) Each repaired sling shall be proof tested by the manufacturer or equivalent entity to twice the rated capacity prior to its return to service. The employer shall retain a certificate of the proof test and make it available for examination. 

Exactly.  This smooth, slow pull to 2x intended load is the basis for the “optimal” test for a climbing team  

One problem with climbing gear is that there is no rated capacity or application of factors of safety in any sense that is familiar to engineers.

Exactly.  It’s why variable, uncontrolled tests on highly variable materials is not recommended  

To compare apples to apples, in industry -where safety and legality matter (therefore sound reasoning) and they have loads more experience (pun intended) in such matters, a safe proof test is 2/5 of design breaking strength.  Not shock loading it just under breaking strength.  If that gave a better statistical margin of safety, they’d do it.  Let alone on slings they would have taken out of service a long time ago.

Proof testing involves a large degree of probability and statistics.  

At any rate, I have not concluded that the OP is wrong on any point but I would like to see some relevant research or data. Call me a skeptic but I value data over opinions.

Agreed and good policy.   But you are not going to get a return on a google search for testing old webbing and ice.   You’ll need to synthesize data and studies from proof testing, nylon materials studies from industry and military, statistics and probability analysis, failure studies, etc.    but ASTM had done it for you.  Just apply the underlying foundational principles based on sound engineering, science, and math…not the verbatim requirements   

but…any experienced engineers, proof test developers, or better yet, actuarials anywhere, would never say a that highly variable 2-4x shock load test is equivalent or better than a 2x smooth static load test to highly variable deteriorated materials.  

There are reasons behind the ASTM and OSHA stds.  Even slings in service (still well in the useful green zone) are too unpredictable after a shock load.  Imagine how unpredictable items are that would be considered well beyond useful safe-life (in the red zone).
and there are reasons behind the slow 2x proof. Not even a 3x slow static pull is recommended. Because good proof tests are designed to not introduce more damage than they were intended to detect.
 

Even if the difference is slight (and no one here knows) but I’ll do the test that has been determined to statistically/ probabilistically have better outcomes.  If you have a chance to remove a bullet from the  big Russian roulette chamber of life, why leave it in?

To Ian:  the only thing I have modified from my original premise is to go from a 1+ static load test of the largest guy, to a 2x static load test of both team members.   For a soloist, I’d definitely test using the rope to minimize shock. 

Thank you for the link to the NASA paper. I can't say I read all 110 pages in their entirety but I read enough to understand what their point is. They basically say that proof loading is a poor choice if you have other, preferably non-destructive, test methods available.

Mark Pilatewrote:

Trees, bushes, anchor pieces (pro), aid are outside the scope of argument, due to being more robust against shock  — though a good test is still appropriate 

Wood has a significant relationship between load duration and strength . Wood strength resisting impact is roughly twice what it is for permanent loads. I have demonstrated this by testing enough pieces of wood to failure with impact loads to get a good statistical representation of the strength. Then take several pieces and apply about 75% of the impact strength and wait. In a week or two, the loaded samples will start to fail. Unfortunately, this effect is cumulative so the the time for multiple load applications is added to get the total. Just because a wood anchor supports you when you begin a rappel doesn't mean it won't fail before you get down or when the next person descends. A 2x bounce test would give me some confidence that a wood anchor will last long enough.

Wood behaves like the composite that it is and I believe I know a few things about wood structural behavior. I have to admit that I don't know for certain the failure mechanisms of polymers or of woven components like webbing. The NASA paper seems to primarily concern itself with crack propagation in metals. I don't think polymers and their failure mechanism are specifically mentioned. This doesn't mean that the paper isn't applicable but that there are always more questions. If nylon has a significant load duration strength relationship, I am not aware of it. I think it does creep a lot but I don't know if the strength decreases under long term loading. I don't know about the effect of cyclic loading on polymers.

A couple quotes from the paper:

"The value of the proof test is greatly enhanced as a flaw screening method when combined with non-destructive examinations of the component." Yep, look at that webbing before trusting it. 

"Proof or overstress testing a component consists of applying a load greater than it would experience during service. This practice is well established as a means of detecting, possibly destructively, gross manufacturing and material defects before the product is delivered." I think this applies to things like brand new webbing. Tie the webbing around an anchor and then test it to make sure the knot, which might have a gross defect, is right. There was an accident where a brand new piece of webbing failed because the person using it did not know that bulk webbing can have cuts joined with a piece of tape. Not long ago I bought a spool of webbing that was two pieces butt joined with a wrap of masking tape. Depending on the colors, the splice can be surprisingly hard to see.

We all know that rappelling is one of the most dangerous parts of climbing and it usually terrifies me. I really don't care how ignorant people think I am but I'm bounce testing any system that I plan to rappel off. I have some degrees, licenses and certifications that suggest I am not stupid but self assessment of competence is not reliable.

The entire  argument is absurd.  Any anchor that can't  easily withstand a bounce is not remotely strong enough to rappel safely from. 

Nick Goldsmithwrote:

The entire  argument is absurd.  Any anchor that can't  easily withstand a bounce is not remotely strong enough to rappel safely from. 

Are you listening to The Dude’s story?

Dan Merrickwrote:

Wood has a significant relationship between load duration and strength . Wood strength resisting impact is roughly twice what it is for permanent loads……. Just because a wood anchor supports you when you begin a rappel doesn't meanwon't fail before you get down or when the next person descends. A 2x bounce test would give me some confidence that a wood anchor will last long enough.

Okay, wood was completely outside the scope of original discussion here, and it is a given than an over stress proof test is required,  but you raise some puzzling points that seem to favor my position.

You are basically agreeing and proving my “point 6” that some others contested and Patrick wanted confirmation of….That some materials can sustain a high impulse load, but then fail at  roughly 1/2 that load under longer static conditions.  I agree with you and it was that knowledge that led me away from using a higher impulse load test (bounce) vs a longer smoother 2x static stress load 

It would seem from the info you posted and have experience in, that one WOULDN’T want to do a quick impulse test that may mislead me into trusting something that could then potentially fail later for the longer time duration of the rappels

Review section 6 of the paper on dangers of badly constructed proof tests — basically either undetected issues for intended use, or doing more unknown damage just from the test are the main concerns I’m trying to discuss  

As for nylon web, after the many degradation variables that I listed in the OP, accumulate over the years, it fails in a random mix of brittle and ductile mechanisms, as well as unseen “micro tears” (yes, here they are real) due to embedded grit, etc. yadda yadda.  That response to a shock load that could be at or near its unknown remaining ultimate strength, is completely unpredictable.  

So, given that we seem to agree on all the lead in principles, and that others in industry who know the principles, outright reject using high impulse loads on non-uniform, highly variable materials,  I’m not sure why you are sticking with “bounce”  vs static load proof test.   Maybe it’s as simple as our understanding of “bounce” is just different and we do in fact agree   

Nick Goldsmithwrote:

The entire  argument is absurd.  Any anchor that can't  easily withstand a bounce is not remotely strong enough to rappel safely from. 

No disagreement from a dead horse perspective, but I’m drawn to the logical fallacies presented here like a moth to flame.  

Almost all anchors you will encounter will be fine and have WAY more margin than you need, so you can do whatever test floats your boat.  

But the scope of the discussion here is emergency, unplanned bailing where for whatever reason, you have limited resources and will need to rely on sketchy materials you found in place.   You are saying you are doing a “bounce test”.  The danger is the ones that pass your shock test and leave you in a more dangerous position.  You have no idea what the actual failure strength is ….just like in the Tahquitz tragedy.  Point is to just limit higher “shock” loads than needed to avoid the condition that you now think it’s safe and it fails later during use.  Rare, but can happen.  

Your primary objective is to ensure it’s reasonably safe enough to use in the current context.  You want to employ the test that gives you the highest PROBABILITY of ensuring that it won’t fail under use  

Maybe not you, but at least some climbers are bouncing on static tethers and imparting higher shock loads to highly variable and degraded materials and expecting that this “helped them” sort out a bad anchor.  
maybe it did, odds are good. But maybe it actually misled you and left you even more exposed with less margin to sustain the planned load or an unplanned shock during actual use —like ledge you’re standing on suddenly giving way -as you know.

Based on scientists and engineers from NASA , sling mfrs, plastics scientists, regulatory and stds organizations, statisticians, etc. all would say that you are way out on a limb and in uncharted territory applying logic, proof testing practices, or material response expectations used for new material and applying them to end of life material with unknown variable stack ups. It’s way unpredictable. Your goal is to make it slightly a bit more predictable by testing it properly. 

I am suggesting to apply the same logic and science that has been vetted by industry and ASTM for nylon slings.  Compared to industry stds, even my proposal is way out of bounds.   As the engineer I spoke to on the phone said, “I know you guys are suicidal (climbers), but don’t be crazy”

In my opinion, referencing tests developed for testing rocket motor housings, and trying to apply them to climbing rappel anchors, doesn't really make a lot of sense.  How the heck is space ship testing relevant to some old nylon webbing?

This thread is such a mess.  There's effectively zero data here that indicates that any of your "proposals" will actually work, or will result in people being safe.  Alternatively, we could simply encourage people not to rappel on non-redundant, old, crappy tat, full stop.  This thread could give people who don't know any better the idea that it's ok to rap off sketchy tat, as long as they follow some official-sounding "test protocol."  That's not good.

Kyle don’t go out of context again.  This definitely ISNT rocket science.   Nothing about climbing is “safe”, but some things are safer than others.  If this threads a mess, it’s due in large part to you sidetracking with straw men and out of context info.    Again you post irrelevant data about numbers not related to the debate and skirt the salient points about bad proof tests.  I apply that logic and use the factors that apply to nylon slings vetted by the rigging industry.   I apply info to the context in which it is valid.  I suggest you do the same  
If you have a more appropriate source, post up

The MAIN gist of thread and OP is not to do it at all.   The last bit, is — if you do (and some do) at least do it right with sound principles.   If anyone has led noobs astray it is you.

1. Limit shock loading!   

2.  Use the proof test principles and logic developed for nylon slings by ASTM.  

Confine your debate and criticism to refuting 1or 2 above and state your reasoning, or just Tap out and go debate NASA, and ASTM and hound them for their data and “junk science”

Edit to Kyle:  haha, Ok you got me there.  I thought I said “section 6” is most relevant.  Will fix in original reference.  

The First 2/3 of the OP thread is devoted to “Why not to rely on sketchy anchors”. But Some do, some have been killed over the years. It continues to happen. Some, including you, have proposed that a bounce test may save lives if relying on sketch.  That is less safe. 
 I’m merely pointing out a cautionary consideration in adding extra shock loads in testing such a system.   

 “My” proposed test isn’t any more “convoluted”than a “bounce”.  Just use smoothly applied static loads vs a bounce of high impulse.   Limit shock loads.  People can do whatever they want.  Take or leave the advice.   Again, people have to use their god given common sense and apply the principles to their situation. 

Edit to Nick below :   TA Da!!  I always knew we were essentially 99% in agreement all along.   Now it’s 100% from your last post. Won’t split hairs.  Can close thread 

Everyone be safe out there. Back up everything with known good materials.