Do you need to retire dropped carabiners?

Back to the Bush Guide

– Bjorn Sturmberg

This is a controversial topic that comes up every so often and which, due to the often singular importance of carabiners, is worth some further research.

If you want the one word executive summary, it would read; “No” (as in no, you don’t need to immediately discard dropped carabiners). If however you want to know when you should and shouldn’t, and why, then read on…

Firstly, lets look at some of the specific testing that has been done on the topic. Basically they all boil down to dropping some ‘biners from a great height, setting up an industrial tension rig and testing said ‘biners to failure. A summary of a few such reports follows.

According to Chris Harmston, the quality assurance manager at Black Diamond: “I have test-broken hundreds of used, abused, and dropped ‘biners (even some that fell 3000 ft. (1000 m) from the top of the Salathe Wall on El Capitan). Never have I noticed any problem with these unless there is obvious visual damage to the ‘biner. While somewhat reassuring, this does not give you carte blanche to use carabiners that have been dropped a significant distance. Immediately retire any carabiner that is crooked, has deep indentations, or has a gate that doesn’t operate smoothly.” (If you dont know what the Salathe Wall is check out Mayan Smith-Gobat climbing it!)

Impact point on Wild Country ‘biner dropped 255m onto hard surface (photo Richard Delaney)

Closer to home we have a new test rig set up in Katoomba that is churning out some very interesting results (and is always looking for some damaged ‘biners to test).

Richard has started a facebook group that includes lots more discussion and videos of ‘biners being loaded until failure. He has for instance tested 6 locking gate carabiners (3 x aluminium, 3 x steel) that were dropped 255m onto a steel plate (a fairly extreme scenario).

Post drop observations:

  • 3 steel ‘biners all obviously broken and unusable
  • 3 aluminium biners some observable impact marks but appear normal

Post drop destructive pull tests:

  • 2 recovered steel biners: gate open strength as expected (normal)
  • 3 aluminium biners: 1 landed on crotch and impact resulted in complete gate separation. Destructive pull-test performed (gate open). Failed at 19.20kN (rated gate-open strength 20kN). Another survived with gate intact and failed at 23.79kN (rated to 24kN). A video of the later is here.

Gate blown off ‘biner that was dropped 255m onto a steel plate sitting on concrete (photo Richard Delaney)

From these tests we can conclude:

  • Steel carabiners, for many reasons, appear to suffer obvious physical damage from large drops
  • Aluminium carabiners may not have any obvious physical sign of damage from large drops
  • 3 aluminium carabiners suffered no loss of strength from large drops
  • This was a very small sample size. However even with a sample size of 1000 one cannot be assured that number 1001 may not land in a particular orientation and suffer some undetectable damage resulting in a significant loss of strength.

So what would be a good legal recommendation for a website to make? (legal disclaimer; no one in any way associated with this post is qualified or generally recommendable for legal advice) Follow the manufacturers’ instructions with regard to use, inspection, and maintenance for all your equipment. If they say destroy after a significant fall then send it to “Rope Test Labs, PO box 193, Katoomba, NSW 2780” and Richard will gladly destroy it for you. This is a way for them to increase their sample size and thus validity of the results (so we can follow their recommendations with greater trust).

Presumed impact point on aluminium ‘biner dropped 255m onto hard surface (photo Richard Delaney)

Same ‘biner post pull test. Peak force was 23.97kN, just short of rated strength or 24kN (photo Richard Delaney)

And lastly 2 more experiments;

  1. In a test conducted by REI, thirty carabiner bodies (half ovals, half Ds) were each dropped six times onto a concrete floor from a height of 33 ft. (10 m). Following the drops, their open-gate strength was measured and compared to thirty control samples from the same production batch that had not been dropped. The statistical result was no loss of strength.
  2. At the International Technical Rescue Symposium, 2000, Garin Wallace and Kevin Slotterbeck of SMC presented data on the strength of carabiners that had been dropped 27’ or 54’ onto concrete or asphalt. One by one, they dropped 115 new, SMC locking D aluminum carabiners. Then they broke the carabiners, measuring the breaking strengths. What do Garin and Kevin say about using climbing hardware that has been dropped? Retire your hardware if you drop it (they do work for a manufacturer, after all). What do their numbers say? The carabiners that were dropped were no weaker than the un-dropped carabiners. In fact, the average strength of the dropped carabiners was slightly stronger than the un-dropped carabiners, but the difference was not statistically significant (during the presentation a certain engineer from an unnamed airplane manufacturer in the Seattle area was heard muttering something about forge hardening).

All in all the data from these diverse range of experiments using a mix of different ‘biners seems to be confirming the same thing, that unless there is clear visual damage it is unlikely the ‘biner has been weakened. That said, when it comes to essential safety gear that is used in life and death situations, it is probably best to err on the side of caution.

Finally, if you want to inform yourself about some more rope work related concerns / myths you can check out ‘On Rope’ by Smith and Padgett, onrope1 (not related to the book) and geir hundal. Some of these are climbing specific but most are life and death issues for every canyoner.

Back to the Bush Guide


8 thoughts on “Do you need to retire dropped carabiners?

  1. Back in the early 90’s Blackheath Venturer scouts had a policy of placing any dropped carabiners into a purpose made wooden box. These were periodically tested at QANTAS on an electron microsope (or some such device) for internal fractures. The steel ‘crabs’ always came back undamaged but according to John Wylie, the Blackheath Scout group leader, the alloy crabs often didn’t pass the test.

  2. Should you not take a Factor of Safety into consideration here? A carabiner rated to 24kN should fail OVER 24kN, probably by at least 25%. Companies need to include a factor of safety into their design.

    Failing at 23.97kN actually illustrates a compromise of strength…

    • 24kN is the breaking strength. Giving a lower number than the true number encourages overloading because “yeah dude, they’re like waaaaay stronger than they say.”

    • Sport climbing gear ratings DO NOT include an extra factor or safety (or has a safety factor of 1, however you’d prefer to look at it). Working gear, made for other purposes than sport climbing, usually does.

  3. I’m not a climber nor an engineer, but a number of sections jumped out at me.

    “This was a very small sample size. However even with a sample size of 1000 one cannot be assured…”

    This statement is misleading, and unfairly diminishes the benefits of larger samples. HIgher sample sizes increase confidence. You are far less likely to be surprised at the 1001st result than you are at the 4th result. This was a very small sample size, and the results are of limited value. Don’t evade that, and don’t draw strong conclusions yet. I applaud your efforts to continue testing.

    Again, I am not a climber, by my understanding is that part of the “throw away after one drop” advice is that if you don’t have that policy, you soon lose track of how many times it has been dropped – if the weakening is cumulative, it is the risk of using a mechanism that has been dropped 20 times, rather than once. The test after 6 drops is the closest to the real risk, as it was explained to me.

    “The statistical result was no loss of strength”

    “Statistical result” doesn’t mean anything here. If you mean that the *mean* strength remained normal or within an acceptable range, that is a misleading result. If a single one of them failed below the spec, it is negative outcome. (If the controls also failed below spec, you have a bigger problem!)

    • I am a climber and an engineer, and agree with pretty much all of the points above.
      However I would also be interested in how damaged vs. undamaged ‘biners perform in a fatigue test, since many failures in everyday life occur waaay below the ‘ultimate’ strength, due instead to fatigue. It may be possible to load-up an undamaged carabiner to near its capacity thousands or even millions of times, but add a big nick (or worse, a near invisible crack) resulting from a fall, and it is more likely to reach its limit after dozens or maybe hundreds of times. Bearing in mind how often you load/unload a typical carabiner, it wouldn’t take long to rack up more load-cycles than a damaged ‘biner can handle.
      So I echo the above comments – a larger sample size will always give you a more reliable result, but also – the greater the variety of tests, the more potential failure-modes you will capture too.

      P.S. If a carabiner fails significantly above its rated strength, then it hasn’t been optimised & you are effectively carrying unnecessary weight. Both figures quoted above (23.79/24 & 19.2/20) are within 5% of their ‘rated’ value (which is not necessarily a ‘minimum’) which does imply that they are optimised. Although, yes, I would be more comfortable if they were consistently just *above* rated strength!

  4. I’m not sure if this is true or not but I was told the that the stress fractures became a problem under heat. Heat being genterated from friction with the rope while rappelling. Maybe the next tests you should experiment with the carabiners heated. Or do a few tests with droped fig 8’s and belay devices.

Join the conversation. Share your thoughts.

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s