Relaxing Cowstail Knots

[cap n chris]

I?ve tried to envisage a plausible scenario for putting even a FF 1 onto a rope, let alone a FF 2, and find it difficult.

A dutiful SRT novice on a club trip might easily experience a significant FF if they slipped off the foothold after clipping their cowstails into the rigged traverse here:

 

[global_s]

Forgive this novice for asking, but I don't get that rigging. Why not have a tight traverse line to a bowline on the bight on the last two bolts?

 

[Glenn]

'cos the rigging for the Y hang is pants!

 

[Benfool]

Glad somebody said it.......

 

[ianball11]

Maybe the overhang beneath the y-hang bolts confused the issue and they've lowered the y-hang knot to be below that rub point so needing a big loop, though it's just two fig8s on the y-hang?  not a double bolt share.  Is there a rebelay further down so it's only a short section of pitch?

 

[r_walklate@hotmail.com]

Is there an argument that you would not use your cows tails on that traverse? Surely using a stop would stop any FF?

 

[cap n chris]

it's only a short section of pitch?

Not short, really, as it's about 15m to an anchor which can be rigged either as a rebelay or deviation according to whim/group/lucky seaweed.

 

[Bob Mehew]

I started on this topic following work by the French on Cows Tails, see http://british-caving.org.uk/rope/lanyard_tests_v6.pdf.  On the back of it we wrote an article for Speleology in April 2008 whcih I will convert to a suitable form and put up on the BCA web site.

 

[cap n chris]

Is there an argument that you would not use your cows tails on that traverse? Surely using a stop would stop any FF?

Improved rigging would be my preferred solution.

 

[Penguin]

I have been wondering about the fall factors in exactly that sort of fall recently, where the attachment point slides down a rope. 

If one fell from a cowstail length above the solid anchor the cowstail was clipped into, that would be a fall factor 2. 

If one fell a from cowstail length (say 1 m) above a vertical steel cable 2 m long eg on via ferrata to which the cowstail was attached, this would be a fall factor 4. 

But if one fell with a cowstail clipped into a semi-static or dynamic rope as in the rigging photograph, surely this would reduce the fall factor?  Consider a fall with a cowstail (say 1 m) from the top of the loop which descends to the Y-hang (which is about 1 m) - so effectively climbing 1 m above the bottom of the loop as the karabiner will just slide down.  There are now 2 m of rope (1 m semi-static, 1 m dynamic) involved in the 2 m fall, so a fall factor of 1.  Is this correct?  Surely, if the caver fell from an anchor at waist height with tied to a 2 m semi-static rope this would also be a fall factor of 1.  So what happens when we combine ropes? 

Could you fall down a 10 m pitch into a rebelay loop tied to your 1 m cowstail a scenario which could give a fall factor of ~1 (0.91)?


And how does fall factor equate to force?  Or does it not?  Is there a magick formula combining weight and fall factor to give a force?

 

[ianball11]

http://www.myoan.net/climbart/climbforcecal.html

 

[Cave_Troll]

That calculator has got to make a lot of assumptions.
elasticity of ropes when manufactured vary due to manufacturer / rope dia
elasticity of ropes may vary with age, use dirt etc.
shock absorption by knots.
elasticity of person.

 

[caving_fox]

I have certainly encountered rigging similar to the above, and would always treat the last butterfly knot as a rebelay. Eg put stop onto the short descent to Y hang - unless there's enough decent footing around to justify only safetying with cows tails. I suppose technically then I could slip, but as above you'd have to get it really really wrong to fall onto that slack rather than slither/bump. Hence nowhere near the same peak load.

I fig8 my cowstails and restrain the knot with rubber clip at the krab. Although I've come across the advise to loosen them I have yet to hear convincing enough argument that it is worth the extra risk/hassle from loose knots.

 

[ianball11]

That calculator has got to make a lot of assumptions.
elasticity of ropes when manufactured vary due to manufacturer / rope dia
elasticity of ropes may vary with age, use dirt etc.
shock absorption by knots.
elasticity of person.

I like how static and dynamic are in relation by 2:1

 

[Bob Mehew]

On the back of it we wrote an article for Speleology in April 2008 whcih I will convert to a suitable form and put up on the BCA web site.

see http://british-caving.org.uk/equipment/Cows%20Tails%20article%20v2.pdf.

Five observations on what has been written so far. 

When I used the term "relaxed" I never thought people would think I was implying that knots should be so loose as to be at risk of becoming undone in a fall situation.    It has been an interesting learning point.    When you dress a tied knot, you may subject it to a force of up to 0.5kN / 50kg.  When you lower yourself onto a knot you then subject it to a force of around 1kN / 100kg.  And in doing so, you will almost certainly tighten the knot up a little more.  Lyon's report http://www.hse.gov.uk/research/crr_pdf/2001/crr01364.pdf suggested poor rope work might place a load of 2kN / 200kg on the rope and thus the knot.  The translated French Cows Tails article suggests that the maximum force a knot is likely to see under most deliberate caving action is around 3kN / 300kg.  Our research work suggests a fall of even 0.5m is likely to impose a peak force of 6kN / 600 kg on the first drop rising to over 10kN / 1000kg (that is over 1 ton) for subsequent drops.  We also find that the rope is most likely to break within the knot at the first bend.  So by tightening a knot beyond its dressed state you are likely to increase the peak force seen in a fall situation and thus increase the likelihood of the rope breaking.  By relaxing the knot I mean taking it back from 2 or more kN to 0.5kN; not leaving it at risk of undoing.   

The concept of Fall Factor is almost useless for rope lengths below a few metres since the set up is not a simple single length of rope but complicated by the presence of knots, rope loops and movable metal work such as crabs, as Penguin perceptively comments.  I have yet to push force measurements below 0.5m sample length but when you think about it, a crab and two knots plus their loops will take up much of that length.  (Just to make clear, that 0.5m under a 1kN load, so the length prior to placing the load on the rope will be somewhat shorter.)

Drops survived is the reported parameter from dynamic testing of rope.  The value of 4 mentioned by Fulk related to a brand new 9mm static rope.  The Long Term Rope Test http://british-caving.org.uk/equipment/LTRT.pdf reported the average drops survived of even lightly used caving rope at between 2 and 3 drops.  The standard (BS EN 1891:1998) requires a rope to survive 5 drops.

The chances that one might seriously fall on one's cows tails or whilst abseiling or prusiking are low but the consequences could be high or even fatal, see http://www.hse.gov.uk/research/hsl_pdf/2003/hsl03-09.pdf.  I am talking about a simple precaution to reduce those consequences.  Our Speleology article recommended the complete untying of the cows tails.  Following some representations after the publication of that article, I backed off from suggesting a complete untying to just relaxing the knots so one can efficiently hand wash them without loosing lengths.  I have found that I can get the length of a rope sample with knots when under load to within a few centimetres by simply measuring the point at which I make the second bend in the rope, having made and dressed the first knot. 

The calculator at http://www.myoan.net/climbart/climbforcecal.html gave 11.25kN for a 100kg mass on a 0.8m length static rope 0m from the last anchor (ie FF1.0) in comparison to our measurements of between 6 and 9kN.  Can someone look at the coding behind the web site? 

 

[TheBitterEnd]

It's pretty simple, weight is weight, lrope is length of rope, lanch is distance from anchor, the first three IF statements do units conversion and the final one does static/dynamic. It boils down to

form.shock.value=weight/80*((lanch+lrope)/lrope)*4.5

for static rope

function compute(form) {
var weight=parseFloat(form.weight.value)
var lrope=parseFloat(form.lrope.value)
var lanch=parseFloat(form.lanch.value)
form.fallfact.value=""
form.shock.value=""
if (form.tanch[0].checked==true) {
lanch *= 0.3048
}
if (form.unitrope[0].checked==true) {
lrope *= 0.3048
}
if (form.tweight[0].checked==true) {
weight *= 0.45359237
}
if (lanch>lrope) {
alert("The rope must be at least as long as the distance from last anchor.")
}
form.fallfact.value=(lanch+lrope)/lrope
if (form.trope[1].checked==true) {
form.shock.value=weight/80*((lanch+lrope)/lrope)*4.5
}
else {
form.shock.value=weight/80*((lanch+lrope)/lrope)*9
}
}

 

[Penguin]

But what exactly do fall factors represent?  Are they a useful way of calculating the capability of a system to absorb forces, or the breaking point of a system?  Or not?  Can we say, for example, that a particular system can withstand a fall of factor 1?  Or that another system can withstand a fall of factor 2? 

Jumping into a shaft, tied to 100m of rope anchored at waist height will generate much more force than a fall onto a 1m cowstail anchored at the same point, but both are fall factor 1.  Can we say that a system can absorb a factor 1 fall when it is possible to generate such a range of force in a factor 1 fall?

 

[Fulk]

In theory, the force on your body as you come to a halt if you jump into a pit on 100 m of rope is the same as it would be if you jumped into the same pit on a couple of metres of the (same) rope (ignoring 'end effects' due to the knots / krabs you'd need to attach the rope to your harness), because although you would convert 50 times as much potential energy into kinetic energy, you'd have 50 times as much rope to absorb it . . . but you won't catch me volunteering to put it to the test.

 

[Bob Mehew]

Thanks Bitter End; I do not recognise the 80 or 4.5 / 9 constants in the formula but there again I have seen little trying to provide numerical descriptions of rope performance.

Fall Factors provide a means of comparing the capability of different lengths of the same rope, provided the overall length does not drop below a few metres.  Fulk's example misses out the feature that whilst jumping on say a new 11mm dynamic rope is likely to hold you, jumping on say a 3mm accessory cord will not.  I would also qualify the force values in Fulk's example by saying they are likely to be roughly similar - there are some surprising factors at work which could modify the forces.  We have observed that the time of the peak force at one end of the rope is at a different time to the peak force at the other end but have yet to get a handle on its implications.  (And I am ignoring air resistance.) 

 

[Penguin]

So could we consider fall factors as an indication of the ability of a body and system to withstand and absorb the potential fall forces, provided both body and system can withstand the forces generated in the potential fall in the first place?  Ie not making cowstails from 3 mm cord! 

So for example we might say that a typical SRT system - anchors, hangers, karabiners, rope, cowstails, harness and caver - should withstand the forces generated in a factor 1 fall but has the potential to fail at fall >FF1?  Whereas a via ferrata system  - cable, cowstails, harness, climber - with dynamic absorbtion built in to the cowstails should withstand fall factors >2?