Gear Physics: How Ice Screws Keep Ice Climbers from Falling to Their Deaths
Image: My Good Images/Shutterstock


This story is over 5 years old.


Gear Physics: How Ice Screws Keep Ice Climbers from Falling to Their Deaths

The simple piece of gear that made ice climbing a legit sport.

While so many adventure sports in the GoPro era seem to exist for the sake of spectacle, ice climbing remains a practice of deliberation. It's eerie, slow, and beautiful—a quiet sport of ascending glistening chandeliers.

Those chandeliers consist of frozen waterfalls and other vertical or otherwise steep ice formations, such as those resulting from intermittent groundwater seeps or runoff from snowmelt. It's very gear dependent. To ascend, the climber is equipped with ice axes and crampons—foot-mounted arrays of sharp metal points, basically—which are affixed to rigid, sturdy boots. Ascending is a matter of digging axes and crampons into ice and pushing/pulling oneself upward. Climbers are protected against falls by a roped belay system like that used in traditional rock climbing.


In all of its roped forms, climbing depends upon anchors. In a belay system, this means that one climber is roped to another via an anchor, such that, should one fall, the rope will become taut against the anchor while pulling (usually upwards) against the other climber. It's probably easier just to see it:

When the climber in the above scenario falls, the result is a downward force on the anchor resulting from the masses of both the climber and the belayer combined. This is slightly counterintuitive, but we can imagine the scenario as though each partner were attached independently to the anchor, with their own rope. When one climber falls, the anchor experiences the force of two climbers falling. The upshot is that an anchor needs to be able to withstand a whole lot of force (several thousand Newtons of force).

In regular rock climbing, this anchor can take a few different forms, ranging from the cams and nuts of traditional climbing to the drilled-in bolts of sport climbing to top-rope anchors built from nylon webbing lassoed around trees and-or sturdy rocks. In ice climbing, however, we're pretty much just left with ice. And, with ice, we don't have a whole lot of options when it comes to anchors. Enter the ice screw.

While a traditional rock climbing rack can be a dazzling, diverse array of colorful hardware, there's a uniformity to ice climbing protection ("pro," the gear that is inserted into the face and is supposed to support a fall). Screws come in different sizes, but at the end of the day they're still just screws: As a climber ascends an ice formation, hacking ad hoc holds with axes and crampons as they go, they stop at regular intervals to twist a screw in and and attach a rope to it.


Alpine ice

Ice climbing as a sport of ascending waterfalls as though they were granite faces started as a subproblem of alpine climbing and mountaineering. Getting up a rugged mountain peak via glaciers and couloirs could well mean dealing with passages of steep, uninterrupted ice. In the earliest days of mountaineering, this meant literally hacking stairsteps into it.

Climbing's relationship to ice changed dramatically in 1908, thanks to climbing legend Oscar Eckenstein and his newly-developed 10-point crampons. While heavy at three pounds apiece, their aggressive spikes allowed for ascending ice without the need for cutting steps beforehand. In effect, they were the steps.

As a belated obituary for Eckenstein, published in the Alpine Journal in 1960 (he died in 1921), put it: "These crampons were most effective and so recently as 1924 were described as having 'the merit of being the only claw at the present time in which both the metal is rightly wrought and the points are shaped and placed under the foot with any scientific regard for their use.' Eckenstein himself records that, thanks to the use of his crampons he did not cut more than twenty steps in all over a period of twenty-five years, apart from one unfortunate day when he inadvertently took someone else's crampons instead of his own."

Eckenstein was at the same making radical changes to ice axe design. To pair with his new crampons, he developed an axe small enough such that it could be wielded with one hand. As the abovementioned obit remarks, however, the axe was at the time mocked. It wouldn't be until the 1960s that short ice axes became popularized in a real sport-changing sense, thanks to the innovations of Yvon Chouinard, the owner-founder of Patagonia clothing. His 1978 book Climbing Ice is considered to be the founding document of modern ice climbing.


And what was modern ice climbing? Ice had become a sport in itself. The one-handed ice axe meant that an ice climber could have a hand free with which to affix protection, which meant ascending riskier, steeper ice. In the early days, ice-based protection meant spikes resembling the pitons often used in rock climbing (left, below). Obviously, these would have to be hammered in.

Beyond the snarg

Ice screws started appearing in the 1950s and 60s. The earliest version is probably the MARWA screw, manufactured Mariner Wastl in 1958, according to an ice screw history published in the Alpinist. The overthin MARWAs were known to be unreliable, but the ice-screw idea was thus introduced to the climbing world. Still, it didn't really take off for another couple of decades, however, until gear hero Greg Lowe unveiled the "snarg" screw-piton hybrid.

While still requiring the user to pound it in with a hammer, the snarg added the crucial feature of hollowness. This allowed for a much larger diameter because displaced ice could occupy the snarg's open core. The larger the diameter, the more screw surface area could be exposed to the ice. Snargs stuck around through the '80s and '90s, but have now been mostly replaced by screw-screws made by Black Diamond and other companies.

Here's Canadian mountain guide Mike Barter, who would have to be played by Ben Mendelsohn in a movie:

There's surprisingly little research out there about ice screws and ice screw physics, at least compared to the rest of climbing. A key study was published in 1997 by Black Diamond quality assurance manager and materials engineer Chris Harmston. The results were surprising and a bit unnerving.


Black Diamond had become concerned about the guidance being offered by self-described ice climbing experts as to screw placement and usage. A key piece of common advice was that screws should be placed into ice at a generally upward angle, like earlier pitons. Harmston and colleagues spent years and many hundreds of ice screws attempting to answer basic questions about placement, screw length, and even what constitutes "good" climbing ice in the first place.

The researchers created special cells consisting of thick metal buckets that could be used to generate the sort of ice found in real-world (real-ish) conditions. Ice screws were inserted as normal and then drawn out at a rate of four inches per minute as data was collected about how exactly the ice failed given varying screw depths and angles. As far as length goes, Harmston was able to conclude that climbers should go as deep as possible given the circumstances, but that the length to safety relationship diminishes with good quality ice. Longer is better mainly in suboptimal conditions.

More surprising were Harmston's findings about screw angle. As it turns out, screws hold much better when they're angled in the direction of the falling force, as in the second scenario above. Perhaps needless to say, this is not intuitive. Much of the force is then resisted by the threads rather than the girth of the screw itself. This isn't a small difference either: an ice screw placed at a downward angle is as much as twice as strong as a screw placed at an upward "negative" angle. As with screw length, this difference becomes negligible as ice quality decreases. It becomes downright dangerous in scenarios in which "melt out" is possible. (Melt out is just what it sounds like: The ice around the screw begins to melt, thus effectively turning the screw into a hollow smooth-sided piton.)


A new ice-screw study, which comes courtesy of an engineering firm in Ohio, is making the rounds this week. It trades the frozen buckets of Black Diamond's research for a computer modeling technique known as finite-element analysis (FEA). Here, very complex interactions among materials can be captured using ordinarily very burly partial differential equations.

The results support Black Diamond's own research and provide some explanations/visuals. In good ice, again, length doesn't make a whole lot of difference screw-wise. As you can see below, the load in a fall is concentrated in the very uppermost parts of the screw. As for the screw angle, the load difference is huge. (The red areas correspond to ice that's reached its crush threshold; unlike the previous graphic, the screw heads on the right column of this graphic are on the left. The top image is an upward-facing screw, while the bottom is the optimal downward-facing orientation.)

A 2003 study by students at MIT notes something pretty key, which is hinted at in the conclusion of the Ohio firm's FEA study. While the screw angle matters, the rate of the force put on the screw matters much more. That is, if the impact on the screw from a fall can be spread out over time, the force will be much, much lower.

Given the stretchability of modern climbing ropes, this means that the higher the climber is above the ground and, thus, the more rope that's active in the system, the less abrupt of a shock there will be on the screw. This is also the utility of pieces of gear known as screamers, which add in additional stretch to the system such that even in a low fall—where there's less rope available to stretch—the shock can be stretched out, reducing the load per unit of time.

This sort of quantification is a funny thing, however. It doesn't account much for the varying conditions that are guaranteed to be encountered in the wild, including ice that looks totally bomber but is actually riddled with air pockets and-or running water. Differential equations are nice but they don't substitute for experience. Be safe.