Cambridge Underground 1984 pp 28-32

An Investigation into Climbing Rope Failures or
Why I Became an Arm-Chair Mountaineer

'I've done some work on mountaineering ropes. Do you want me to give you an article for the journal?' I said .....

I must have been feeling charitable at the time, since having just written page after page of the official project report (the 'work' referred to being my final year engineering project), the very last thing I felt like doing was going over it all again. Anyway, I've done my best not just to regurgitate it, but provide something of interest to cavers, since some of the breed actually use this rope as their SRT cowstails and safety cords. Whilst the opportunity presented itself, I also tested one or two caving lifeline ropes as a check on their safety.

The tests I performed fell into two categories; drop tests, and tensile tests. The drop test rig (see figure 1) is meant to simulate the conditions a rope experiences when it arrests a mountaineer's fall. Because the rope is run through krabs, any fall means the rope will be dragged over and bent round a krab edge - a radius of about 4 or 5 mm. Thus the drop test is not a straight drop, but is carried out over an edge.

The UIAA have set a standard test because of the infinite possibilities of krab placement in a real rock climb. This is shown in figure 2. As may be seen, the laboratory test is much more severe than the UIAA test but, dare I say it, due to a somewhat lethargic attitude by the laboratory technicians and limited space, I was lumped with what I'd got.

Tests were carried out to investigate the cumulative damage effects of a number of falls of a given fall factor. Here, I must explain what the term fall factor means:

Fall factor, n = (total length fallen) / (length of rope supporting fall)

A little thought will show that this can only lie in the range 0 to 2.

Numerous recent articles in the Caving Press (with which you are all, no doubt, familiar) stress the fact that it is not the length of drop, but the fall factor which is important in determining the severity of the fall. (ref., Ramsden, Paul: BCRA Cave Science Vol. 9, no. 4, pp 290-299, Dec. 1982).

Falls of magnitude n=2 can certainly happen in leader falls in mountaineering, and it is not difficult to imagine the same situation for a caver using his cowstails on (say) a traverse line at a point where he must climb above it as far as his cowstail allows.

Anyway, I needed some graphs of, for example, retained rope strength against number of falls, and since you can't (well, I can't) draw graphs on just two points, I had to find a value of n which would give six or seven points (ie., you could drop the rope six or seven times before it broke).

I tested both 9mm and 11mm diameter ropes as these tend to be the popular sizes used in "half-rope" and single rope climbing techniques respectively (well, actually this happens to be true, but I had nothing else to test in any case). Single ropes were tested since even in the "half-rope" (two separate 9mm ropes) method of climbing, it is possible for just one of these ropes to sustain the entire fall if the system has been badly rigged.

n=0.8 and n=1.4 gave acceptable numbers of results for 9mm and 11mm ropes respectively.

The ropes used were of 'kernmantle' construction (fig. 3) consisting of a bundle of cords, held in place and protected by a woven sheath. The cords were made up of three 'strands' twisted together, the strands themselves consisting of many fine nylon fibres. Equal numbers of right-hand and left-hand 'twist' cords were contained in each rope.

The plotted trends of retained rope strength, and extension to failure against number of falls showed a decrease (for 11mm rope) from 43kN at zero falls to c. 12kN after five falls (this last specimen being severely damaged, the sheath and some internal cords severed), and 35% at zero falls to 23% at five falls. The 9mm was equally dramatic, equivalent numbers being l5kN to 5kN and 35% to 22% after five falls.

So! What was causing these failures?

When a climber falls on his rope, the peak force in the rope (assuming rigid belays and climbers, and no damping) can be shown to be given by

Fmax = W [ 1+sqrt(2n/s + 1) ]


Credit for this nasty really ought to go to Steve Roberts because he wrote it down first, but I think I can say with modesty that it would have come out in the wash anyway.

Typically, for 11mm rope, s = 0.03, n = 2, and W = 800N, then Fmax = 12.6W = 10kN, ie. about 25% of the ultimate tensile strength (UTS) of a new 11mm rope.

Now during a fall, the rope is stretched, but does not fully recover elastically. This leads to a lower value of s. Looking at the equation above, it thus increases Fmax on subsequent falls. This longitudinal stiffening of the rope, with subsequent increase in peak load, is not in itself sufficient to cause failure. Also, the climber's body, harness, and rope friction all add to damping in the real case, reducing peak forces.

The implications for SRT ropes are much more serious however. By their nature they are low stretch (otherwise we'd all be prussiking on climbing ropes) and so peak forces are much higher, damping is also less. The lesson is obvious: SRT ropes should not be used in situations where they may have to sustain falls (eg., lifelining) and in normal SRT use this is not the case.

However, I digress.

It was found that severe abrasion occurred at the "edge" in each of these drop tests, due to the rope being dragged over it in tension. This abrasion caused a gradual decrease in UTS, and once the sheath was ruptured, all the central cords were immediately vulnerable to abrasion (fig. 4), leading to a rapid fall in strength with increasing numbers of drops. (In fact, in nearly all the tests, once the sheath had been torn through, the subsequent drop led to failure, despite the sheath only representing some 20% at most of the rope's UTS).

To conclude, in new rope, the decrease in UTS due to abrasion and increase in Fmax due to increase in stiffness eventaully cause failure by tensile overload. Failure always occurred at the edge where stress and abrasion were concentrated. We may be thankful that in caving use, our cowstails, although bent round the edge of karabiners in the end knots, are not usually subject to severe abrasion.

NB: One area of rope weakening that I was unable to investigate sufficiently was that due to environmental degradation and fatigue. As an indication, a 10 year old 11mm rope (not kernmantle), which had sustained two severe falls, numerous small falls, and considerable later use for hauling equipment up rock slopes was found to fail at c. 9-10kN, ie. one quarter of its new counterparts' strength, or alternatively would only sustain a fall of n = 1.4. (These results were independent of the amount of surface abrasion).

Now to the caving lifeline.

I tensile tested one sample of each of four specimens, namely:

Both hawser-laid ropes were "pretty old" but I had no details of their age, other than that they were older than the Braidline. It had been decided to retire the hawser-laid on the grounds that it was now probably unsafe.

The tensile test results were as follows:

for Knots
(for n=2)
New Braidline0.0920.47 kN12.282 kN6.08 kN
Old Braidline0.0911.125 kN6.675 kN6.08 kN
14mm Hawser0.09-9.976 kN6.08 kN
12mm Hawser0.06-7.231 kN7.24 kN

(Assuming that knotting the ropes leads to a reduction in strength to 60% of the initial value).

The value of 60% is debatable, but figures using it show that even the grottiest bit of hawser-laid is actaully stronger (but not safer) than our well used Braidline. Perhaps some more tests are called for (it could not be ascertained what reduction in strength the grips caused in the Braidline).

These figures for Fmax are a bit close to the UTS for comfort. However, there are reasons to believe that a caving rope deteriorates pretty rapidly on its first few trips, but after that, its mechanical properties tend to level out. It is fair to assume that this has happened for all the used ropes. It is also fair to assume that in all caving circumstances, the worst possible fall encountered whilst lifelining would be n = 1.

So, it is safe to use all the ropes tested with the possible exception of the 12mm hawser-laid, but the Braidline should be regularly tested again to check for any further deterioration in strength.

Bill Hawkswell, April 1984

> Index to Cambridge Underground
> Table of Contents for Cambridge Underground 1984
> Back to CUCC top page