Specialty

Cords for Glacier Travel

What UIAA 110 Fixed, and What It Left

In 2025 the UIAA approved a standard built for exactly these products.³ UIAA 110 defines a static rope as one “intended to support the weight of a human body or bodies in mountaineering, caving, canyoning, and glacier traverse when rappelling, ascending, positioning, and rescue rigging scenarios,” and sets no diameter limit. They explain the reasoning in a footnote: new materials may produce smaller diameters, and such ropes can meet the standard if they meet the performance requirements, because “diameter is an important consideration in choosing which devices and techniques to use in the field; it is also an important consideration for the wear life and ease of gripping.”² The scope statement names the uses these cords were already being sold for, and the standard tests against them. That starts to fix the certification problem. It does not fix the guidance problem, for reasons set out below.

A second edition of UIAA 110 is dated May 2026 and took effect in September 2026. It added the requirement that measured elongation be lower than the manufacturer’s claim, a check of the stated mass, and a requirement that strength meet the manufacturer’s own claim as well as the minimums. The figures above are from the second edition.

Why the Numbers Disagree

The spread is large. Petzl recorded nothing above 2 kN with a mobile belayer and people falling into real crevasses. ENSA recorded close to 4 kN on a 5.5 mm Dyneema cord and called a fall like that unholdable. Smith, on a bridge abutment, recorded 3.1 to 5.0 kN, and ENSA’s 2012 tests on dynamic ropes came in at roughly 1.6 to 2.2 kN, read from their curves.

There are several things that differed in these studies that could play into the differing results: fixed against unfixed anchors catching the fall, the crevasse lip characteristic or its simulation, the location and type of release for the fall, the amount of rope in service, the rope type and diameter, human against rigid drop mass, and whether there were knots in the system. I will take these one at a time.

How much the anchor could give

ENSA’s 2012 report notes that on a real fall “l’assureur avance et résiste sur plusieurs mètres”, the belayer moves forward and resists over several metres. They simulated that first with a dragged haul bag, and in 2013 with a pulley set to slip at a constant 75 daN for 3.5 m before locking onto deadmen. In 2013 the same 9.1 mm Joker was then dropped at the same 12 m twice, once onto fixed deadmen in their spacing series and once onto that mechanism in their knot series: about 172 daN against 138 daN, 20% lower with an anchor that could move, both without brake knots in the rope.⁷ Their 8.1 mm Ice Line runs the same way across the two years, about 178 daN on the 2012 deadman against 120 daN on the 2013 mechanism, though the 2012 drop was on 15 m of rope against 12 m, and a second 2012 drop on the same rope read about 158. Neither pair is set side by side in the reports, and the 138 is printed and the other three are read off their published curves. Their own repeat runs scatter too: the Joker at 15 m reads 182 daN in one series and 172 in another, so 20% is above their noise but not far above it.

Petzl point the same way. They publish no spacing, masses or fall geometry, so their numbers cannot be set directly against ENSA’s, but with a belayer free to move and catch the fall their figures land in the same low ballpark, typically 1.1 to 1.4 kN, with nothing in their testing above 2 kN. So a figure measured on a fixed anchor overstates what a team would feel by roughly 20%. That is worth knowing when two studies sit that far apart, but it does not explain the larger gaps between them.

What the rope ran over

Only Smith’s edge was artificial, a three-inch steel pipe that in his own words “did not provide any energy absorption that would come from a snow corniced crevasse edge”.⁵

ENSA measured forces on each side of the edge in the same fall. Smith, with one load cell, measured each side on separate drops, and his readings show no consistent loss over the pipe. Over ENSA’s snow lip the faller saw 238 daN against 172 at the surface with a rigid mass, and in the fall with a person their curves show about 173 against 118. So a snow lip takes something like 30% out of the force before it reaches the belayer.⁷

Comparing belayer-side figures across the studies lands in the same place: Smith’s dynamic ropes came in at roughly twice ENSA’s over a real lip. If the edge is what separates them, the two studies should agree on the faller’s side of it and diverge on the belayer’s. Scaling each force by the weight of what was falling shows exactly that: Smith’s 8.8 mm reached 3.55 times its mass and ENSA’s Joker 3.03 times theirs on the faller’s side, within about 20% of each other, where on the belayer’s side they are a factor of two apart. Neither study presents it this way, and it leans on the rope behaving like a simple spring, which no rope exactly does, so take it as a rough approximation. So Smith’s figures read as an upper bound rather than as what a team would feel, and the edge is doing a large share of the work in every result that used a real one.

How the fall started

Smith released 1 m above the edge and 1 m beyond it, ENSA 1 m from the lip at a height they do not report, and Petzl do not say how their falls began at all, only that people fell into real crevasses over lips that varied in hardness and shape. Comparing the release means holding the anchor constant, which leaves ENSA’s 2013 drops onto the slipping mechanism against Petzl’s falls with a belayer free to move. Both ran over a real lip onto a restraint that could give, and they land in the same ballpark, so the release does not look like it costs much, though with Petzl’s spacing and masses unpublished that is a trend rather than a like-for-like comparison.

None of them accounts for the snow bridge itself. ENSA and Smith dropped a mass out from the edge to stand in for falling through one, so whatever energy goes into breaking a bridge is missing from their tests, and Petzl, who had real crevasses, do not describe how their falls started. The evidence here is thin, both because the reports say little about it and because what can be compared shows little difference.

None of these is a straight drop. Smith’s mass hung 1 m above the edge and 1 m beyond it, yet the mass fell 5 to 11.5 m on 25 m of rope, because the rope has to swing from horizontal to vertical before it loads.⁵ Those distances are hard to square with his own forces, because a descent that long implies about twice the elongation a dynamic rope shows at 3.5 kN. I cannot account for it from the report, and it suggests something in the system was giving that is not described. ENSA describe the same shape in their crevasse, the fall pendular, the mass striking the edge and then swinging about its equilibrium. The fall factor, the usual way to compare one fall with another, does not describe any of this cleanly.

How much rope was out

Smith ran his whole set again on 6 m instead of 25 m and the forces rose sharply, from 3.1 to 5.0 kN up to 5.0 to 7.3 kN, and he concluded that wherever a low-friction fall is possible every effort should go into putting more rope into service.⁵

ENSA dropped the Joker on 15, 12, 9 and 6 m, onto fixed deadmen as immovable as Smith’s bridge abutment, and report the reverse: the force on the belayer falls as the roping distance falls, 182 daN at 15 m down to 155 at 6 m, read from their curves.⁷ Their explanation is that the release height was fixed, so the extra rope only added stretch, and 15 m lengthened the fall by more than 90 cm against 6 m. Both varied the rope the same way, by moving the anchor rather than by adding slack, and both held the release unchanged.

Smith measured a much larger effect than ENSA did. His forces change by half again to double; theirs by 17% across four lengths, in steps of 4 to 7%. The one repeat visible in their work is the Joker at 15 m, which reads 182 daN in one series and 172 in another, so each of their steps is about the size of the gap between two runs of the same setup. They publish no repeat values, no error bars and no test order, and anything drifting through the session, snow warming or the lip cutting back or the deadmen settling, would show up as a trend if they worked through the lengths in order.

The same scaling makes the two comparable. A fall caught with no slack and no drop still peaks at twice that weight, whatever the rope, because the mass descends as the rope stretches, and that is the floor. ENSA sit just above it, at 1.98 times their 80 kg mass on 6 m and 2.32 on 15 m. Smith sits between 2.7 and 5.1 times his, and 7.4 on 6 m, which takes a real drop before the rope loads. More rope spreads a fall that exists, and theirs barely did. The spring model behind that is fair for ENSA’s short drops and poor for Smith’s long ones, where his forces rise less than it predicts and the distance his mass fell implies more rope stretch than those forces can account for.

All of this is about force, and spacing is not set by force in the first place. It has to be long enough that two people are not on the same bridge at once, which scales with the crevasses you are crossing, and that constraint sits above anything the drop tests say.

What the rope was

Each study compared rope types inside its own setup, which makes these the most reliable comparisons available, and they point in different directions. On peak force the dynamic rope was lower every time: ENSA’s dynamics at 1.6 to 2.0 kN against roughly 4 kN for the 5.5 mm Dyneema,⁷ and Smith’s small low-stretch rope about 1 kN above the small dynamic, which he says held regardless of the test mass, though at 118 kg his own figures show about 0.35 kN. On fall distance it reverses, with his low-stretch 8 mm cutting the drop roughly in half, 5.75 m against 11.25 m.⁵

On whether the fall gets held at all, the only study that recorded outcomes favoured the cord. Petzl held 47 of 58 falls on the RAD LINE, 81%, against 36 of 61 on the PASO, 59%. They saw the same two belayer reactions on both, and some falls were not held on either.⁶ The one clear negative result is narrower than “static”: ENSA’s 5.5 mm Dyneema static cord, of which they wrote that “it is clearly impossible to hold such a fall,” which is not the same object as a sheathed 6 mm cord built for the job. That is one study, by the cord’s manufacturer, with no detail on who fell or how, so the stats look more favorable towards static cords in this test series but do not conclusively settle the matter. A dynamic rope lowers the peak force and a static one shortens the fall, but neither follows through to whether the fall gets held: Petzl found the cord’s steadier loading easier to work against than the dynamic rope’s bouncing when the belayer was pulled into motion, though harder when the belayer stood firm against the first shock. What would settle it is the load at which a belayer stops being able to hold, and no study established that. Petzl recorded whether falls were held without publishing a threshold.

Diameter inside a type gives no rule at all. Smith’s two low-stretch ropes went the wrong way together, the 10.0 mm averaging both more force and more drop than the 8.0 mm, 4.8 kN over 8.25 m against 4.5 kN over 5.75 m, where his two dynamics ran the usual trade, the 11.0 mm giving a shorter drop at a higher force than the 8.8 mm at 100 kg; at 118 kg it gave the shorter drop at the lower force. The cords in the table above run from under 2% to 6% static elongation at much the same diameter, which is why: stretch is the property that matters and diameter is a poor stand-in for it. That is where the UIAA landed when they took the diameter limit out of 110.

What was falling, and where it was measured

The falling mass was 80 kg at ENSA, 100 and 118 kg at Smith, and real people of different weights and skill at Petzl. ENSA dropped both and found a rigid mass produces 32% more force than a person of the same weight, so the rigid-mass figures read high. Most figures were read at the belayer, or at the anchor standing in for one. The exceptions are one of Smith’s series, read on the falling mass, and ENSA’s rigid-versus-human pair, where the person was measured on the climber and the rigid curve looks like a surface reading. That distinction is worth more than it sounds. ENSA put sensors at both ends on the same drop, a 9.1 mm Joker at 15 m, and measured 238 daN on the falling mass against 172 daN at the surface, 38% more at the bottom than at the top. With a person falling instead of a mass they put the reduction at 45%, which they attribute to the rope running through the lip and to the stretch in the rope itself, though their own curves for that fall show about 32%.⁷ What this gives us is that the two ends of the rope see different forces in the same fall. Nearly every figure here is belayer-side, and nearly every belayer-side figure comes from a rigid mass, so they read about 30% above what a team would feel from a real person. The faller sits at the other end, where roughly 30% of the force is lost before it reaches the belayer.

How the force arrived

While most of these studies focus on peak forces, how fast the force arrives matters too, and Petzl addresses this: “the human body feels the difference in absorption between a static or dynamic rope, even if the maximum impact force is the same.”⁶ On their curves the RAD LINE rises sharply enough to unbalance a belayer where the dynamic PASO 7.7 rises gradually, and both stay under 2 kN. That cuts both ways though, because the stretch that softens the onset is also what makes the rope bounce, and through the arrest Petzl found the RAD LINE loaded steadily, a “regular feeling of tension, which facilitates belayer action,” while the dynamic rope “alternates suddenly and unpredictably between low and high tension” and destabilizes the belayer. They conclude that the RAD LINE can be used for this, and that its lack of stretch “can even be favorable in cases where the belayer is surprised by the fall.”

In addition to how fast the force arrives, how long it lasts also matters, because arresting a fall means absorbing all of the falling climber’s momentum, and a brief spike and a sustained pull of the same magnitude are not the same problem. The spike might be what takes you off your feet, but the sustained load is what would take you to or over the lip.

ENSA concluded of their Dyneema cord that “it is clearly impossible to hold such a fall,”⁷ while Petzl concluded that the RAD LINE can be used for exactly that. ENSA inferred it from a force trace, with a rigid mass on a fixed anchor and nobody trying to hold anything, where Petzl put real belayers on the rope and recorded what happened, finding that a belayer who was set in motion and then held after a few steps was the most common outcome. They were also not testing the same cord: ENSA used a 5.5 mm Dyneema static cord whose construction they do not describe, and Petzl the 6 mm RAD LINE, which is a blend of materials. Whether the cord or the method accounts for the opposite conclusions cannot be settled on the evidence available, because the Dyneema was only ever dropped onto a fixed deadman, not onto the slipping anchor they used in 2013, and not with a person in place of the rigid mass.

Petzl’s own range is the sharpest case. The RAD Line and the PUR Line are both 6 mm and both under 2% static elongation, so neither absorbs energy — yet Petzl endorses crevasse fall arrest on the first and says the second “must not be used for belaying.” The reason given, that the cord does not absorb energy, is true of both. What differs is the construction: the PUR Line is HMPE alone, where the RAD Line pairs HMPE with polypropylene and nylon in a sheath built for grip and braking. Petzl say as much in the PUR Line’s own notice: “be vigilant with the type of knot used; the PUR LINE’s sheath is extremely slippery.”⁸ Whether that is their reasoning for the different guidance, they do not say.

If one of these studies has to carry the most weight, I would lean towards Petzl’s, because real people fell into real crevasses, real belayers tried to hold them, and they ran it more than a hundred times across varied lips. Nothing else comes close on any of those counts. The catch is that they publish conclusions, typical curves and a held rate and little else — no weights, no spacing, no forces beyond the typical curves — and that they sell the cord, so unlike ENSA and Smith they have something riding on the answer.

Whichever way you weigh them, the lesson is the one this article keeps running into: “hyperstatic cord” is not a category that behaves the same way throughout, or that has been tested in the same way. Most of these cords could theoretically meet UIAA 110, and a result from one product still would not carry to another for this use.

Brake Knots

Brake knots are one of the ways a team can take energy out of a fall. ENSA is the only measured data I could find on them: on a 9.1 mm single, knots cut peak force 20% and the tension left in the rope at the end of the fall by 87.5%; on an 8.1 mm double they did nothing to peak force but still halved the residual tension. That is twelve drops in total, three per configuration, and on the strength of them ENSA wrote that the results “prouvent l’efficacité des nœuds de freinage”, prove the effectiveness of brake knots. Their 2012 series, in colder snow, had reached no clear result.⁷

What the knots are doing, when this works, is taking energy out of the fall by ploughing into the lip and the snow, which is the job the manufacturers say something in the system has to do when the rope will not. A second or third person being dragged a few steps does the same work.

Three things are not known. In 2012, in cold recent snow over an ice crust 30 to 40 cm down that the ropes mostly slid on, ENSA found the knots made no clear difference on the thinner ropes, though they cut the force 40% on the 10.2 mm, so the effect depends on conditions you cannot count on. Neither paper names the brake knot they tied, though ENSA’s film of the 2012 series shows the knot they used in it. And every rope either series tested ran from 7.7 to 10.2 mm, so how brake knots behave on a 5 to 6 mm cord has not been measured in any of these studies. Beyond what ENSA measured, the usual advice on how many knots to tie, how far apart, and which knot to use rests on convention and field experience rather than testing.

What the Testing Cannot Settle

Every one of these studies reports a peak force, a number that can tell you whether the team is likely to get pulled off their feet, but not whether they recover and arrest the fall. The things that decide that are mostly ones the published work does not settle: how fast the load arrives and how long it lasts, which only Petzl describes and none of them put numbers to, the terrain and the snow the belayer is standing on, how many people are on the rope, which nobody varied, and whether there are knots in it, which nobody has measured on anything thinner than 7.7 mm.

Then there is which cords were tested at all. Of everything in the table above, the RAD LINE is the only product named in any published study, and that study is Petzl’s own. So the best evidence here describes one cord, and it bounds the problem rather than answers it.

Practical Implications