| OUCC Proceedings 11 (1983)Corrosion for Cavers II: Corrosion of Alloy Karabiners | OUCC Proceedings 11 Contents | 
Andy Riley
During a caving trip down Swinsto Hole, Kingsdale, N Yorkshire, a heavily pitted karabiner was found at the bottom of the 20 foot pitch lying under 0. 5 m of water in the gravel bed (Figures la and b).
The whole of the body of the karabiner was covered in pits several millimetres wide and deep. At the bottoms of the pits was a white, insoluble deposit. Only the screw-up gate was unaffected by corrosion. The unattacked parts of the surface did not show any more abrasion damage than would be expected to be caused by normal caving use.
A section 
was made across the backbone of the karabiner which was hot-mounted in perspex 
and polished to a 1 micron finish. Under optical examination, the entire section 
was found to be full of cracks. This unexpected discovery led us to make further 
investigations into the nature and origin of the cracks.
Under low 
power observation, all the cracks were seen to run in circles, concentric with 
the outer circumference of the section. The cracks were extremely numerous, 
branched and finely divided (Figure 2). They seemed to be evenly distributed 
throughout the whole section and were not located primarily at the surface. The 
bases of some of the pits were examined. There was a definite orientation 
dependence of the cracks to the base of the pits. In many of the pits, the 
cracks were seen to emerge at the bottom (Figure 3), the pit growth direction 
lying parallel to the crack direction. In other pits, the crack direction was 
perpendicular to the direction of pit growth (Figure 4) and had not yet reached 
the surface. All the pits had a deposit at their base. The aluminium metal at 
the base of the pits showed signs of extensive corrosion attack and was very 
spongy and porous in nature.

The 
composition of the karabiner was unknown, but was determined by electron 
microprobe analysis to be an Al-Zn-Mg-Cu alloy (see Table 1). An analysis of the 
deposit in the base of the pits was attempted using energy dispersive analysis 
(see Table 2). Wavelength dispersive analysis would have been superior but was 
unavailable at the time of writing. The deposit was predominantly composed of 
aluminium oxide.
Discussion of observations
The 
chemical analysis suggests that the karabiner is a high-strength Al-Zn-Mg alloy. 
Shreir (1976) suggests that these alloys have a high risk of suffering 
stress-corrosion cracking which can be accelerated by incorrect heat treatment. 
It seems likely that the surface pitting was probably initiated by the emergence 
or interaction of the internal cracks with the surface. Once pitting has been 
initiated, the pits grow unhindered by external effects or microstructure. This 
is because pits generate an acid environment at their base which prevents 
reformation of the passive film on the exposed aluminium surface (Robinson, 
1960) and therefore corrosion proceeds rapidly. Impurities in the water can 
assist in the initiation and propagation of pits - a combination of carbonates, 
chlorides and copper ions can be very damaging (Davies, 1959). In hard water, as 
little as 0.02 ppm of these ions can initiate pits (Porter and Hodder, 1953, 
Rowe and Walker, 1961).
The 
orientation of the cracks in the section strongly suggests that the 
microstructure is exerting a major influence on their growth direction. The very 
fine branched nature of the cracks suggests that they are intergranular. 
Unfortunately, I was not able to etch up the grain boundaries in order to 
demonstrate this. Exfoliation corrosion is a well-known phenomenon in high 
strength aluminium alloys. Robinson has examined the effect of elongated grain 
structure and heat treatment on the formation of surface blisters (Robinson, 
1982). It seems likely that an elongated grain structure is formed in the alloy 
karabiner during manufacture as it is extruded and that this initiates surface 
and filiform attack. Grain boundary attack then causes the production of 
corrosion products, creating large stresses at the grain boundary which force up 
grains at the surface to create blisters. If these blisters reach a certain 
size, a pit will form and pitting corrosion will dominate.
Conclusion
Grain 
boundary attack has probably occurred because of precipitation and segregation 
of alloying elements at grain boundaries during heat treatment. Exfoliation 
corrosion produced blisters on the surface which in turn caused deep and severe 
pitting. The intergranular attack and pitting in this karabiner has become 
apparent due to its immersion for an unknown time in cave water which might be 
expected to contain the necessary impurities for this kind of corrosive attack. 
It would be interesting to know how long it would take for such attack to occur 
and whether such slight attack which may occur during normal caving use has any 
effect on the strength of the karabiner.
References
Davies, D.E., 1959. Pitting of aluminium in synthetic 
waters. J. appl. Chem.
9, 651-660.
Porter, F. C. and Hadden, S.E., 1953. Corrosion of 
aluminium alloys in supply waters. J. appl. Chem. 3, 
385-409.
Robinson, F.P.A., 1960. Pitting corrosion - cause, 
effect, detection and prevention. Corros. 
Techno1. 7, 237-239, 266.
Robinson, M.J., 1982. Mathematical modelling of 
exfoliation corrosion in high strength aluminium alloys.
Corros. Sci. 22, 775-790.
Rowe, L. C. and Walker, M.S., 1961. Effect of mineral 
impurities in water on the corrosion of aluminium and steel.
Corrosion 17, 353t-356t.
Shreir, L.L., 1976.
Corrosion. Butterworth, London, 2 vols., 2nd edn.
Table  
1.  
Electron microprobe  analysis 
of karabiner  alloy 
composition 
Standardless EDS analysis 
(ZAF  corrections 
via magic V)
| 
		Element & line | 
		Weight % | 
		Atomic % | 
		Precision 3 sigma | 
		K-ratio | 
		Iter | 
| Al Ka | 92.54 | 6.77 | 1. 18 | 0.9279 |  | 
| Cu Ka | 1.11 | 0.49 | 0.34 | 0.0108 |  | 
| Zn Ka | 6.35 | 2.74 | 0.91 | 0.0613 | 6 | 
+ ca. 1% Mg
Table 2. Electron microprobe analysis of 
pit deposit composition 
Standardless EDS analysis (ZAF corrections via magic 
V)
| 
		Element & line | 
		K-ratio | 
		Weight % | 
		Precision 3 sigma | 
		Oxide formula | 
		Oxide % | 
| Al Ka
		
		 | 0.6654 
		 | 37.40 
		 | 1.10 
		 | A12O3
		
		 | 70.66 
		 | 
| Si Ka | 0.0511  | 6.20 | 0.68 | SiO2 | 13.27  | 
| S Ka | 0.0075  | 0.58 | 0.17 | SO3 | 1.44  | 
| Ca Ka | 0.0168 | 0.77 | 0.18 | CaO | 1.08 | 
| Fe Ka | 0.0358 | 1.42 | 0.36 | FeO | 1.83 | 
| Cu Ka | 0. 1418 | 5.95 | 0.99 | CuO | 7.45 | 
| Zn Ka | 0.0817 | 3.42 | 0.82 | ZnO | 4.26 | 
| O* |  | 44.25 |  |  |  | 
* - determined by stoichiometry