*** Carabiner Life *** Fri Feb 10 07:36:52 1995 Message : #21938195 From: Rob Simmon Address : simmon@daac.gsfc.nasa.gov Group : Usenet.rec.climbing Subject : Re: Failure of Aluminum (was Retiring Old Carabiners) Org. : NASA Goddard Space Flight Center -- Greenbelt, Maryland USA Clarifications & additions to the discussions about breaking biners: stuff usually fails in two ways, ductile and brittle. ductile failure occurs when there's enough force inside the metal to move dislocations in the crystal structure through the lattice of atoms, resulting in deformation, and eventually breaking. (note that it doesn't have to break to fail, it just has to be deformed enough to be useless.) nylon, aluminum, and steel usually do this. Godwin Liu: So, as I understand it, you mean to say that aluminum, when it fails, fails by means of ductile fracture (as opposed to brittle fracture). The process you describe, is, I believe what I was trying to describe: *usually* It is possible for aluminum to fracture in a brittle mode, as in fatigue. Godwin Liu: Let me elaborate on my earlier, admittedly simplified, explanation: Several stages of a metal's response to stress: 1. A metal will first stretch elastically (i.e. stress leads to stretching, and when stress is removed, metal returns to original shape.) 2. The metal will then move into a plastic stretch (i.e. stress leads to deformation.. when stress is removed, metal has been permanently deformed) 3. As the metal stretches plastically, its strength has been changed, and will lead to breakage. (ultimate tensile strength). What I meant to say was that if you stay within region #1, you have not "damaged" the biner (see important notes on fatigue below, however). the 'ductile' fracture you describe sounds to me like what happens when you go through regions #2, and #3. Yup. another rough definition of brittle vs. ductile fracture is that brittle occurs in the elastic range, and ductile occurs in the plastic range. The note on fatigue is VERY important, so we can't say "no worse for the wear." on to fatigue ... Godwin Liu: Rob, can you give us an indication on the size of crack that matters? Are we talking about microfractures? or large ones which can be identified by visual inspection? Nope, I can't give you specific crack lengths. Wish I could, though. There is a material property, KIC (kay *subscript* one cee) that relates crack length to applied stress that will cause growth, but it's tricky. It's based on energy arguments - energy of new surface created vs. energy released by reduction of stress. (actually that might be GIC.) Microfractures and internal fractures, so visual inspection isn't infallible. Godwin again: I find it hard to believe that regular aluminum bar stock develops significant fractures under conditions of top roping, even as the anchor biners. well, it does. Aluminum, even under infinitesimal stresses, will eventually fail by fatigue. but we're talking A LOT of cycles. me: after crack initiation, very small stresses can cause elongation, so only a FEW CYCLES, AT LOW STRESSES can cause failure. materials differ widely in their response to fatigue - steel, for instance, can be fatigued forever at a low enough stress level. (really, as close to forever as we can measure.) Godwin: there appears to be something of a double negative here. But I am assuming that you mean that steel can be cycled at low stress WITH these cracks, without fracture. I should have been more clear. If a material is put under severe stresses, cracks will develop that can then be propagated by low stress. Steel and Aluminum both will fail under these conditions. SO - a dropped biner should be retired. However, if the stress is ALWAYS low, (below the 'fatigue threshold') Steel will not suffer a fatigue failure. Aluminum does not have a fatigue threshold, so repeated stressing, no matter how low, will lead to fracture. SO - an "old" biner should be retired. certainly more than three months, certainly less than 25 years. (that narrows it down, doesn't it?) Godwin: Conclusions on Biner Damage due to Fatigue: ------------------------------------------- Even elastic strain can cause the initiation of small cracks, which ^^^^^^^^^^^^^^^^^^^weaken the material. In the case of aluminum, further stresses after this crack intiation, cycled, can cause the aluminum to break. Thus, theoretically, a biner CAN break under regular conditions of climbing, if it is regularly put under conditions of load (because of fatigue). SO, since it is difficult to evaluate these conditions, it may be prudent to change your biners after repeated conditions of loading. YUP. me: there are other things involved, like corrosive environements and the like, but I won't go into it. (except to say that 'biners that have been near the ocean should be wiped with fresh water and dried, and retired sooner than other biners) Godwin: :I am still unclear on this matter. I thought that aluminum was quite :resistant to corrosion. In that, any newly exposed aluminum quickly :oxidized to a very thin layer of aluminum oxide, which does not have :nearly the same weakening effects as, say, rust, iron oxide. Aluminum is quite resistant to corrosion - in air, without the presence of chlorine. However, if some sea salt manages to get into a surface flaw or under the pin that holds the gate on - LOOKOUT! Corrosion may appear as a rough white bumpy area on the aluminum. This will severely weaken the 'biner, and get worse with time. Godwin: Final Conclusions ------------------ As my previous ones, but with: >Rob Simmon: retire a biner after you've used it for "a while" (who knows how long a while. maybe someone should write black diamond.) But let's also be sensible. According to other posts in this thread, no one appears to have ever heard of a modern biner design actually breaking. Which gives us some confidence. Let's just not overdo it with the confidence. I agree completely. I do want people to know, however, that the gear is not infallible forever. It's REAL close to begin with, but time takes its toll ... this is probably on the FAQ, but what are the numbers for BD, MSR, Lowe, Metolious, etc. etc.? especially the R & D departments. BD at least has a materials engineer on staff, who wrote in Climbing magazine about cracking of SLCDs ... Anyone on staff who reads this stuff who would like to reply? -rob simmon simmon@daac.gsfc.nasa.gov Fri Feb 10 08:08:16 1995 Message : #21939298 From: Rob Simmon Address : simmon@daac.gsfc.nasa.gov Group : Usenet.rec.climbing Subject : Re: Failure of Aluminum (was Retiring Old Carabiners) Org. : NASA Goddard Space Flight Center -- Greenbelt, Maryland USA bubb@tools.ecn.purdue.edu (Anthony R Bubb) wrote: True, look at it this way. Material deteriorate with cyclic loads. Steel stops deteriorating after about 1,000,000 cycles. Aluminum just keeps on going. I am not saying that you will ever load the material 1,000,000 times though. That's probably why they use AL... no reason to believe fatigue failure is a whole lot lower than steel, other than the the str is lower to start with. AL and Steel do have about the same str to weight ratios though. actually, the fatigue strength of Steel and Aluminum may be VERY different. In fact, the fatigue properties of two different types of the same metal could be very different, as well. It really has little to do with the yield strength or elastic modulous, and more to do with microstructure. (No, I don't have any numbers, my books are boxed up sitting under my old desk at CMU.) The fatigue failure criterion used in design by mechanical enginerds for withstanding cyclic loads (frequency effects not-withstanding) is generally the Goodman Equation. If I recall correctly (and I don't have a text handy to check) it states that a material will fail when: {(a/Syt + o/Sut) = 1.0/n} (I might be mixing up the placement of Syt and Sut...) Where Syt = Yield str, Sut = ultimate Str, a= 1/2 peak-to peak amplitude of load cycle, o = load average, and n = # of cycles. I don't seem to recall the time-space between loads being relevant to this criterion. rough guesstimate. It would take a LONG time to load the biner by a lead-fall 100 times let alone a million. One can use VonMieses Criterion and the Goodman Equation to predict how many times a biner would withstand a particular load. you only need ONE lead fall to start a crack, after that a top-roped fall could make it grow, eventually leading to failure. again, I encourage anyone who's really interested to read up on it, or at least send me a message in about a month when I get my textbooks back, and I'll be more specific. -rob simmon simmon@daac.gsfc.nasa.gov