Fretting fatigue of Karabiners

[Simon Wilson]

Just watching them as someone comes up a pitch rigged in that way gives me nightmares.

Why, what do they look like? I've never noticed anything scary while watching the krabs as somebody prussics.

You can watch them flex with every prussik!

Next time you fly watch the aircraft wings flapping about. They are made of similar alloys and also have a big FoS (factor of safety).

 

[TheBitterEnd]

A I said I think the FoS is such that a dramatic failure is extremely unlikely but just interested to know if this is something that anyone has investigated.

Just to go back to the study linked to above, it's interesting but I'm not sure it's especially relevant to this discussion.

Well it is for me, I was made aware that fretting lowered the fatigue limit in alloys. The situation we have in caving is an alloy krab fretting on a steel bolt and just out of mild interested I wondered if this had been investigated. The study Bob linked to, whilst not directly about fretting, is pretty much exactly the conditions we have in that they had an alloy krab connected to a steel pin in their test machine.

Just to be clear, I was never suggesting that krabs would fail under normal use but it is interesting to know if the life may be reduced by fretting (or anything else for that matter) to the point where it might be advisable to retire a 15 year old club krab. However it seems not  since said club krab if used by parties of 8 every weekend on a typical dales pitch will have undergone less than 600000 2.5kN load cycles

 

[Bob Mehew]

MJenkinson quote:

See low anchor warning in CNCC rigging guides for example.

I wonder if this is this the example that Bob's referring to just above?

yes  see https://cncc.org.uk/fixed-aids/technical-group-anchor-warning.php

 

[andrewmcleod]

...since any single bolt failure should be tolerated...

Ah.  Some pitches have single anchor rebelays.  And of course links on yourself are 'single' not 'double'.  There has been at least one accident in the past where a guy managed to open his crab gate when crabbed directly into an anchor and fell.  (I can't find a link to show how this can occur.)

But (if the anchor placement is good) a single anchor rebelay should be more than 5m below the previous anchor and sufficiently high off the floor such that the anchor failing (or you unclipping the rope from the krab or the krab from the anchor) results in you having a low fall factor fall. Big caveat in this is of course 'should' but there should always be redundancy somewhere.

Your choice of attachment krab is different to your choice of anchor krabs. It is just funny that cavers use screwgates for every connection in their rigging but are then happy to clip in with a pair of snapgates (and on a traverse line if being lazy are often only on one snapgate) whereas climbers will use snapgates/quickdraws throughout the rigging, including when building a belay, but will only ever have a personal attachment that uses screwgates (because this, unlike any rigging, is not redundant).

I use twist locks on my cowstails for this reason (and am happy on just one, meaning on a traverse I don't have to either be on a single snapgate or mess around using the hand ascender krab as a third cowstail etc).

 

[Antwan]

Try http://web.mit.edu/sp255/www/reference_vault/Fatigue_Presentation.pdf

Thanks Bob, that's exactly the thing    I thought someone would have done something like that.

Just to go back to the study linked to above, it's interesting but I'm not sure it's especially relevant to this discussion. In their lowest-force gate-closed tests, they cycle the repeated loads up to 8 kN. Let's say a caver + kit + tacklesack weighs 120 kg, which puts a force of ca. 1.2 kN on a rope. Assuming poor prussiking technique and rigging from a single anchor, that force could be amplified to about 2.5 kN1 whilst climbing a rope, which is only about a third of the minimum force they tested in that study.

From the 8 kN tests, the carabiners tested survived an average of about 11,000 cycles, and the graph appears to be levelling off as the force is reduced. Testing at lower forces is described as, "prohibitively time consuming", suggesting many more cycles would be required for the sorts of forces SRT would impart.

1 page 58 of http://www.speleo.no/redning/Life%20on%20a%20line%20part1.pdf

just to add a bit of force to your assumption, you could quite easily get 4Kn if it is a badly rigged Y-Hang. For that reason I think krabs are fine as personal gear when you know the use/history but not as club tackle when they are likely to get abused (unintentionally) by novices or just people that shouldnt be rigging things.

The OP does raise an interesting point, something to think about regards fixed rigging.

 

[Mike Hopley]

Personally as a climber I sometimes wonder why all pitch rigging is always done with screwgates when snapgates should be acceptable, since any single bolt failure should be tolerated...

I can see the benefit of the krab as something to clip into and the general reassuring security of a screwgate though...

I would personally prefer to rig off snapgates, as they are light and convenient. However, this would only be suitable for an expert group -- or at least, an experienced group who have been thoroughly informed about the risks.

Snapgates are generally not suitable for clipping into with cowstails, due to the risk of unclipping (as the Swiss accident demonstrated). Experts will know this and clip into the knot loop instead (which is the safest place anyway). Exceptions: if you are hanging from the anchor, e.g. at a rebelay, it's okay to clip into a snapgate.

Rigging off snapgates also carries some risk that the rope may unclip itself, as it moves around. This will never happen under load, but could potentially happen when you are leaving a pitch head or moving along a traverse line. Experts will be aware of this and will be vigilant checking the rigging, particularly by looking behind them.

So snapgates introduce extra hazards. These hazards can be managed, but many cavers will not be aware of them. Also I think rigging off snapgates would terrify most people. Therefore snapgates are not suitable for general use -- at least, not yet!

Some European cavers are not only rigging off snapgates, but accessory krabs like this, which are weaker but cheap! This is not as insane as it sounds, but I don't really understand it when for slightly more money you can get these instead, which are full-strength and lighter.

 

[MarkS]

Try http://web.mit.edu/sp255/www/reference_vault/Fatigue_Presentation.pdf

Thanks Bob, that's exactly the thing    I thought someone would have done something like that.

Just to go back to the study linked to above, it's interesting but I'm not sure it's especially relevant to this discussion. In their lowest-force gate-closed tests, they cycle the repeated loads up to 8 kN. Let's say a caver + kit + tacklesack weighs 120 kg, which puts a force of ca. 1.2 kN on a rope. Assuming poor prussiking technique and rigging from a single anchor, that force could be amplified to about 2.5 kN1 whilst climbing a rope, which is only about a third of the minimum force they tested in that study.

From the 8 kN tests, the carabiners tested survived an average of about 11,000 cycles, and the graph appears to be levelling off as the force is reduced. Testing at lower forces is described as, "prohibitively time consuming", suggesting many more cycles would be required for the sorts of forces SRT would impart.

1 page 58 of http://www.speleo.no/redning/Life%20on%20a%20line%20part1.pdf

just to add a bit of force to your assumption, you could quite easily get 4Kn if it is a badly rigged Y-Hang. For that reason I think krabs are fine as personal gear when you know the use/history but not as club tackle when they are likely to get abused (unintentionally) by novices or just people that shouldnt be rigging things.

The OP does raise an interesting point, something to think about regards fixed rigging.

Unless I'm missing something, a Y-hang would need an angle of over 140 degrees between the two arms to generate 4 kN per anchor in the situation I described above. I think it would be quite difficult to rig one with that degree of force amplification.

 

[SamT]

As a climber who's taken whippers on to 3mm worth of brass soldered onto two strands of 1mm wire which,  having been well placed, have held perfect well.. I always read these threads with mild bemusement and a pinch of salt.

Perhaps if some folk feel "terrified" by caving rigging they ought to take up knitting instead.

Mind you.. those big needles look a bit sharp.. deadly even.. #tremble

Sent from my SM-G930F using Tapatalk

 

[Barny]

As a climber who's taken whippers on to 3mm worth of brass soldered onto two strands of 1mm wire which,  having been well placed, have held perfect well.. I always read these threads with mild bemusement and a pinch of salt.

Perhaps if some folk feel "terrified" by caving rigging they ought to take up knitting instead.

Mind you.. those big needles look a bit sharp.. deadly even.. #tremble

Sent from my SM-G930F using Tapatalk

Finally a voice of reason. Well said Sam.

 

[Antwan]

Try http://web.mit.edu/sp255/www/reference_vault/Fatigue_Presentation.pdf

Thanks Bob, that's exactly the thing    I thought someone would have done something like that.

Just to go back to the study linked to above, it's interesting but I'm not sure it's especially relevant to this discussion. In their lowest-force gate-closed tests, they cycle the repeated loads up to 8 kN. Let's say a caver + kit + tacklesack weighs 120 kg, which puts a force of ca. 1.2 kN on a rope. Assuming poor prussiking technique and rigging from a single anchor, that force could be amplified to about 2.5 kN1 whilst climbing a rope, which is only about a third of the minimum force they tested in that study.

From the 8 kN tests, the carabiners tested survived an average of about 11,000 cycles, and the graph appears to be levelling off as the force is reduced. Testing at lower forces is described as, "prohibitively time consuming", suggesting many more cycles would be required for the sorts of forces SRT would impart.

1 page 58 of http://www.speleo.no/redning/Life%20on%20a%20line%20part1.pdf

just to add a bit of force to your assumption, you could quite easily get 4Kn if it is a badly rigged Y-Hang. For that reason I think krabs are fine as personal gear when you know the use/history but not as club tackle when they are likely to get abused (unintentionally) by novices or just people that shouldnt be rigging things.

The OP does raise an interesting point, something to think about regards fixed rigging.

Unless I'm missing something, a Y-hang would need an angle of over 140 degrees between the two arms to generate 4 kN per anchor in the situation I described above. I think it would be quite difficult to rig one with that degree of force amplification.

Like I said, in a club situation you will get realativly novice users who think they know what they are doing and do just that, having witnessed an individual try and rig a perfecr 'T' using a figure 8 knot, if it can be rigged wrong it will be is my philosophy these days

 

[cavemanmike]

Have a little faith anwan.
As long as you educate those who rig wrong.
I know a lot of rigging is different to everyone, but wrong rigging is wrong rigging.
Just saying 

 

[Simon Wilson]

Like I said, in a club situation you will get realativly novice users who think they know what they are doing and do just that, having witnessed an individual try and rig a perfecr 'T' using a figure 8 knot, if it can be rigged wrong it will be is my philosophy these days

It's nowhere near as dangerous as you might think. It's impossible to rig a perfect T because you just can't tie the rope tight enough for a start and then the knots will tighten as load is applied so even if you tried to tie it as a T it would end up much nearer to the recommended maximum of 120 degrees when loaded. There is a big factor of safety as well. The components in the system all have a strength of around a couple of tonnes-force, give or a take a tonne or so. If nylon rope was used the rope would stretch around 30 to 40% before it broke. So, if it was possible to load the rope to breaking point the Y angle would be below 90 degrees. That's not allowing for the double rope in one or both loops which would change things.

 

[Tommy]

Remembered this old discussion and thought I'd add some insight from another sphere.

There have been a number of aluminium carabiner failures in paragliding due to fatigue.

• A much higher rate of cyclic loading occurs in paragliding than SRT.
  While two carabiners are used, they are single point of failure as wings don't fly very well when only partially attached...not as much of an issue under good SRT safe rigging practise.
  I'd imagine most caving krabs fail due to corrosion or plastic gate "flop out" or even abrasion long before fatigue is a concern.

I've attached a file that's doing the rounds with some info.

Video with sobering graphic content of what happens when carabiners fail. https://www.youtube.com/watch?v=mEIKg0DsNKY
(The pilot should have been using tandem rated carabiners, and had a reserve parachute, but sadly the knowledge and resources we take for granted don't always make it to other countries).

Stay safe. Inspect your gear. Inspect your rigging.  :hug:

 

[Tseralo]

There have been a number of aluminium carabiner failures in paragliding due to fatigue.

Slack/highlining have found the same issues with ally krabs as the loading is incredibly cyclic and also much higher than caving high due to the force angles and the line being tensioned.

 

[Loki]

Slightly different but I came across some maillons worn almost all the way through attached to plate hangers and pull through chains. Unfortunately we?d already committed and had no choice but to use them. I can only think the winter floods must vibrate the chains so much the wear between the different components got accelerated. Check fixed gear!!

 

[Bob Mehew]

Slightly different but I came across some maillons worn almost all the way through attached to plate hangers and pull through chains. Unfortunately we?d already committed and had no choice but to use them. I can only think the winter floods must vibrate the chains so much the wear between the different components got accelerated. Check fixed gear!!

And perhaps not unsurprising the same is true for rope.  I had one sample strung up in a passage with its end thought to be well above likely high water levels.  It was not and the rope was rubbed to bits after one flood.