UKC

A few weeks ago I had to abandon my much loved blue Dragon after we failed to remove it from a placement. The annoying thing was that this was not at some difficult to access or dangerous spot halfway up a large mountain, but rather at the top of Stanage, at a place where we could sit safely unroped! To add insult to injury we could even get our fingers on and move all the cams of the unit individually! Having never lost a cam to getting it stuck in over 20 years of using them, and having recovered other people's 'jammed' Friends for them, this was upsetting.

What we realised had happened was that the stem of the unit had come unstuck from alloy head of the unit (it looks like it is held in place with epoxy). This wouldn't be dangerous in fall because there is an oversized 'plug' attached to the end of the stem that stops it from pulling through the head of part of the cam BUT when trying to remove the cam you push the stem (and hence the whole cam) in slightly while retracting the individual cams with the trigger. When we were trying to do this the stem was pushing through the head of the Dragon meaning that we couldn't bring the cams in via the trigger. After spending the best part of an hour trying everything we could think of, we both needed to get home so had to leave it. Some one subsequently got it out, so well done them although I imagine if they place it again they will probably face the same issue.

That design of Dragons were recalled shortly after they were released (indeed the cam I lost was a replacement for the original due to that recall) but the recall was not connected to this issue. I sent DMM photos asking whether I could expect the glue to fail on my yellow dragon which is the same design and age. They said the attachment of stem and axle boss is "through interference fit and bonded with glue, meaning it is unlikely for this to happen"*, but of course it did happen with the blue cam!

I've looked carefully at my yellow cam and given it a wiggle and can detect no movement in the stem. Not really sure what to suggest to other Dragon users besides keep an eye on your cams and if you see the glue has failed maybe don't place that cam as there seems to be a good chance it will be very difficult to get out again. I would be very interested to hear if anyone else has had this happen to their Dragons though. Mine were from 2010: https://www.ukclimbing.com/gear/review.php?id=3213 and have been regularly used since then so I fully accept they have seen plenty of wear and tear.

*I should add to be fair that DMM felt from the photos the cam was stuck due to being over-cammed. I would accept that it was in a tight placement at the bottom of its expansion range, but it was not over-cammed in that all the four cams could be manipulated manually. I feel if the stem hadn't been loose it would have come out without much drama.

In reply to TobyA:

Is this a similar thing to what happened with WC friends a while back? In that case a "notice to inspect" was issued and those that failed were recalled.

In reply to gethin_allen:
I remember there being an issue for Wild Country but don't remember exactly what it was. Not sure if they are made in the same way as the Dragons though?

Here's a photo of the stem protruding from the cam head https://flic.kr/p/VgDUzz

In reply to TobyA:

Unless the stem is meant to do that i think you have a just cause in asking DMM for a new one. Surely even with age and usage the stem is not menat to protrude through the head. Like you say it was lost due to the malfunction.

In reply to TobyA:

Yep - exactly the same problem.

In reply to ModerateMatt:

Not sure why someone would dislike that comment. It's a manufacturing fault which should be covered by warenty. I guess they might require more proof than a photo?

In reply to beardy mike:

DMM sold me a new one at a reduced cost, although with VAT it was still the best part of 50 quid. What they did do which I thought was impressive is ask Ben Bransby who does work for them and is quite local to Stanage to go and try and remove it (I felt a bit bad at abandoning gear which could be visible to even a walker exploring the cliff top - because leaving the cost aside, it felt like littering). But by the time Ben got there someone else had managed to retrieve it.

I'm sure DMM get emails reasonably regularly from people who have got some bit of gear stuck and then expect the company to replace it for free, so I see why they took the position they did. Nevertheless I still believe that it was the stem coming loose that led to us not being able to remove it not over-camming.

I can see why you would say it was my 7 years of use that led to the glue failing not a manufacturing fault. I snapped the springs on my favourite no.1 Flexi Friend after about 13 years of regular use, if you use stuff lots I accept it won't last forever.

In reply to beardy mike:

How long do you as the designer expect the glue to last Mike? Are there industrial glues available that you can expect to last almost as long as the metal they are bonding?

In reply to TobyA:

It depends on the type of glue, how clean the surfaces were when it was applied and how rough the surfaces are, and the amount that went in there. So many variables. That said there are old Raleigh M-Trax and Dynatech bikes which were bonded together 20 plus years ago which are still fine...

In reply to TobyA:
I note the DMM response said:
They said the attachment of stem and axle boss is "through interference fit and bonded with glue, meaning it is unlikely for this to happen"
An interference fit is where the stem is slightly larger than the hole it is fitted to so I would have thought that it should be strong enough to retain the stem - the glue would be forced out of the joint itself and just be present at the ends. Fits are of course subject to variability so a poor one might sometimes occur that might lead to this failure.

In reply to keith-ratcliffe:

That depends on howthe have designed the joint. The termination is a piece of Stainless swaged to the stem wire using a talurit swage tool which forms a hexagonal shape. So glue may be squeezed in between the hex shape and the axle housing.

In reply to beardy mike:
Could an interference fit where the 'male' part isn't truly circular but has slight indentations all the way around be a possibility, so that the glue might have more of a chance of bonding the two parts together (if that is something which would work)?

In reply to TobyA:


That is impressive.

In reply to keith-ratcliffe:

When you look at the head of those dragons from below you can see a tiny amount of glue that has come out and dried on the outside, so your description makes a lot of sense. Thanks.

In reply to beardy mike:

Would a true interference fit potentially result in in tension stress fracture and possibly galvanic corrosion issues with an aluminium part in tension around a stainless steel part?
anyhow, it shouldn't be difficult to have some form of key on the two glued surfaces to keep them together.

In reply to gethin_allen:

I suppose it might be bulkier, to have them key together somehow? It's a good idea though.

In reply to Timmd:

Well it's not truely circular - look at tobys photo and you will see what I mean - there is a hex shape on the wire termination where it's been crushed. It's meant to be like that. For Epoxy to work you have to have adhesion and volume - it's not like anaerobic glues like Cyanoacrylate which need virtually no space at all to form a bond. So in this case I'd say it's a lot of use, mixed with maybe not quite as much glue and a little bit of a film of contaminent as it was glued. Really pretty easy for it to happen...

In reply to Timmd:


By "key" I was thinking along the lines of how you scratch up an old layer of paint to key the new layer to it.
You could do this by slightly modifying the crimping tool to add a knurled finish but I'm not sure how you'd do something similar to the inside of the hole in the cast and milled alloy part.

In reply to gethin_allen:

Yes, a true interference fit rather than a transitional (which this basically is) would result in anodising being compromised unless you preheated the aluminium axle housing before sliding it on - that's not going to happen in this case as it's simply too expensive a process to do. Galvanic corrosion would not really affect the glue though, and it's pretty unlikely to occur in the first place - think you're looking for something that's not really there. The mass of Aluminium is still relatively low compared to the steel, and whilst it's not an ideal combination, it's the ratio which counts and the environment it's in.

As for forming a key - how do you form a key properly on a 9-10mm inernal surface which is anodised without causing a problem unless you do it prior to anodising? And then how do you form a matching surface on the stainless? It's not that easy in practice, not to do in a consitent, repeatable way and I'm not sure they will be having that many problems that it warrents going into it in that much detail. Sure, it's not ideal but Toby's cam looks like it's had a fruitful life. Don't forget, 3 sigma requires only a miniscule percentage to fail - lets say they have 3 a year back compared to at a guess 14000-20000 units sold per year - that's well within what you'd expect...

In reply to TobyA:

Obviously clueless with engineering as will soon become apparent, but could the size and weight of end termination be reduced by hot forging the axle connection onto it or separating the strands of the wire stem like a connection for a suspension bridge or in post tensioned conc to give a direct transfer of force over a wider area?

Also why the plastic cover, how do you inspect (do you ever actually need to?) for fatigue at each end of the stem? (was thinking same on Metolius cams which have a funky spring, presumably as some sort of stress relief?)?

(am sure proper engineering bods find this sort of thing hilarious, so apologies in advance)

In reply to Mattilda:

The terminations are in the smaller sizes hotforged and then swaged/crimped into place on the end of the wire. You can't hot forge the termination onto the end of the wire - that just doesn't make any sense as you'd have to heat the wire to high temps for no good reason. It would still only be a mechanical joint whether hot forged or crimped. Crimping is effectively just the same a cold forging - you're just squashing metal! Could you make a cup and cone termination? Yes, but that is a low volume process. Crimping is fast, cheap and effective.

PLastic cover, not really sure. It's what people do these days, but there is no real requirement other than to stiffen the stem a little. I think BD were the first to completely enclose the stem as their design uses a single long loop of cable which means you have to contain the wire somehow. Previously Wild Country had an open stiffner which was specifically so that you could inspect the wires. That said, the middle section of the stem is not where failures occur - that generally happens at it's ends where repeated flexure of the stem cable can lead to failure of wire strands - largely due to stress created by crimping the terminations on. This is where Metolius have an advantage as their brazed terminations are very very strong - if a braze is done correctly it results in a joint almost as strong as a solid lump of the parent metal. But it's a tricky process to control and requires expertise - metolius have 3 guys doing it by hand day in day out!

In reply to beardy mike:

Thanks! A little knowledge can be a dangerous thing, i was also thinking about the old Forged friends (think they have a design simplicity and elegance if that makes any sense) and was wondering for super super light (but sacrificing flexibility and value for money) if there was anything like Delrin you could make them out of, (or bond a woven loop of carbon into a cast flexible stem?) some ideas probably best left that way.....

(thank you, it's all interesting stuff)

Also a friend has some first generation twin stem Camalots he used until quite recently, have climbed with them - it focuses the mind.

In reply to Mattilda:

At one time there was an australian company who made a plastic cam. I have no idea whether the strength was up to much, but they didnt last long as no one trusted them! But in theory you could do something with composites.

In reply to beardy mike:
This Australian company was Altitude Equipment. They marketed the Poly-Hex in 1998, and the Poly-Nut in 2000.
I have never heard that they produced platic cams. Please beardy mike, could you let me know a little more about it?
To my knowledge, the only company that produced “plastic” spring-loaded camming devices is Coyote Mountain Works (USA) with the Samson, in 1986. This cam was made of composite, (long fiber glass, carbon, nylon).

In reply to nutstory:

My mistake - you're right!

In reply to beardy mike:
Just out of interest can you clarify the difference between swaging & crimping in this context - a Google search produces some interchangeable results for the two terms.

In reply to beardy mike:
Just for fun here are (to my knowledge) all the "plastic" nuts ever made
- MAMMUT (Swiss) Bi-Keil (1968)
- PECK CLIMBING EQUIPMENT (England) Ny-Chock (circa 1968)
- FORREST MOUNTAINEERING (USA) Foxhead #3 (1970)
- COYOTE MOUNTAIN WORKS (USA) Coyote Nut (1986)
- ALTITUDE EQUIPMENT (Australia) Poly-Hex (1998)
- ALTITUDE EQUIPMENT (Australia) Poly-Nut (2000)
- GEAR4ROCKS (Ukraine) Stopperplast Nut (2007)

In reply to nutstory:

There was a patent applied for in the US for a composite cam lobe skinned with metal. I don't know if it was ever pursued. And a reversed Camelot with the axes of rotation opposite to the way a Camelot works. And a bill belcourt one for improved grip on microcams. And a Zedel petzl cam... all sorts!

In reply to keith-ratcliffe:

Well crimping tends to refer to a pressing type process, whereby you out a ferrule between two for tools, bring them together and end up with a joint. Swaging means the same but can also refer to rotary swaging whereby the joint is formed by a rotating tool which incrementally moves in to crush the joint shut. Same same in terms of end result.