The sulphate cliffs of Cabo da Roca demolish stainless steel. This fixed hanger is an exception.
Many thanks to Luis Fernandes Silva for the sample.
To date, the main thrust of this blog has been to attribute the extreme corrosivity of sulphate crags to the action of sulphate reducing bacteria (SRB).
It has been observed that whilst most anchors installed at sulphate-bearing, sea cliffs are degraded within a matter of a few years, there are some examples of stainless steel that show no significant corrosion even after decades of exposure. From both an academic and a practical point of view, such instances are worthy of investigation.
Home-made fixed hangers like the ones illustrated below can be found across the crags of Cabo da Roca. Their actual provenance has been lost in time, but it is believed they were installed between twenty and thirty years ago.
My current hypothesis is that austenitic stainless steels need to contain at least 10% to 20% strain-induced martensite before damage by SRB becomes a reality. Thus, my first action always is to check whether the part is attracted to a magnet. If so, I know that some conversion to martensite has been induced by the manufacturing process.
I use a magnetic balance technique that quantifies the attraction in units of % martensite. Looking at the hanger illustrated below we see that it is barely magnetic with putative martensite composition well below the 10%. I’m not surprised that this material is resistant to SRB attack.
This raises the further question as to why the manufacturing process induces large levels of martensite in some austenitic steels and not others. I discuss this complex topic in this post. A dirt quick summary is as follows –
The rule of 10 percents: An austenitic stainless steel containing less than 10% nickel is likely, on cold-forming, to give rise to a product containing more than 10% martensite, and thus will be vulnerable to hydrogen embrittlement.
So, what do we have here? Analysis reveals –
| Element | Weight Percentage |
| Nickel | 10.1 +/- 0.1 |
| Chromium | 17.5 +/- 0.1 |
| Molybdenum | 2 approx. |
Once again, no surprises. Cold-working of this 316 compliant alloy would not be expected to yield sufficient strain-induced martensite to render it vulnerable to SRB attack. As a side note, perhaps significant, the molybdenum content of this alloy contributes to austenite stability over and above that endowed by the >10% nickel content.
Conclusion: This one obeys the logical sequence –
- 316 means >10% Ni
- >10% Ni means <10% strain-induced martensite
- <10% martensite means low H-diffusion rate
- low H-diffusion rate means SRB cannot penetrate the metal
So, the challenge is still out there. Show me an example of true 316 that has failed through SRB attack.
12 replies on “Another SRB Resistant SS Anchor”
Many thanks for this interesting post,David.
I’m worry about the weld on the “true” 316 SS glue ins.
What’s your opinión?
Could be a martensite induced point? Although the true 316 on de rest of the tensor.
Hi Jon, you raise a good point. There is no doubt that the weld is ferritic, not austenitic. The formation of delta-ferrite is inevitable as the molten metal of the weld pool solidifies. I don’t believe that there is any way around the problem except by annealing the part post-weld.
I have measured two commercial 316 glue-in bolts that employ welding. Whilst the body of both bolts show less than 2% martensite despite substantial strain during cold forming, at the point of the weld they measure more like 20%.
Is this a problem? I’m certainly suspicious that it could be, but I have no evidence from the field to support this, and I’m hesitant to jump to conclusions. There is no doubt that delta-ferrite will support a greater rate of diffusion of atomic hydrogen, but I believe there is more to the issue.
One of my longer-term wishes is to investigate the role of hydrogen binding energies in supporting a buildup of atomic hydrogen in the part. I suspect that the specific microstructure of the strained material has a role to play here, and thus the relatively unrestrained crystallization of delta-ferrite from the melt pool may take us somewhere different.
So yeah, nah, I don’t know. Let’s see if time gives us some clues. There are huge numbers of Raumer Super Stars out there on sea cliffs. If they are going to be attacked, it will be at the weld.
Thanks David.
I’ll keep a Sharp guard over the raumer 316 SS on the cliffs of Menorca, to find any clue.
I’ll send you the broken A2 SS anchor after Christmas, the posts offices are very busy now.
Thanks
Thanks David! What’s your opinion about the PLX bolts that were sold by Fixe? We have a few of those installed on sea cliffs in the canary islands. Do you think they are safe as 316 or dangerous as 304?
Firstly, it’s important to realise that Fixe used two different duplex stainless steels in their PLX product. One was the cheaper 2304 alloy which contains 0.3% molybdenum, and the other was the 2205 alloy which contains 2.5% -3.5% molybdenum. This information is not from Fixe but is what XRF analysis tells us. The difference may or may not be relevant to what you see happening but needs to be borne in mind.
My objection to the PLX product was that, at the time we really didn’t know why we were seeing the acute failure of SS anchors on certain sea cliffs, and swapping to yet another alloy could only increase the confusion. And confusion is exactly what followed.
There was a poorly substantiated belief that we were dealing with stress corrosion cracking (SSC), and duplex alloys were being marketed as a cure for that ill. As things turned out, the villain has proved to be sulphate reducing bacteria causing failure by sulphide stress cracking (SSC) and, if anything, duplex alloys with their extensive ferritic interlayers, are likely quite the wrong choice for this problem.
My guess is that PLX will be similar, or worse than 304 for the sea cliffs of the Canaries. Both the volcanism and actual wall-wash sampling tells us that sulphate levels can be high on these crags. I know from a UIAA test site in Railay where we have PLX installed next to 304 and 316 that PLX is performing somewhat worse than 304, with both being worse than 316.
Out of curiosity, how is the 316 holding up there? Do you know?
By there, do you mean the Canaries?
I have got samples of failed hardware from there and have got wall-wash samples. I can confirm high levels of sulphate and the presence of sulphide on the surface of fractured bolts. However, I am yet to do detailed metallurgical analysis, or confirmation of actual chemical composition. I think it will be the same situation as we have for other sulphate locations with apparent resistance by 316, but I don’t really have the evidence of that.
Sorry, I should have been clearer. I meant the 316 samples in the UIAA testing site in Railay, how are they holding up?
I haven’t had a chance to recheck the UIAA test site in the last few years. At that point, 304 and PLX were both looking pretty bad. On the other hand, titanium and super-austenitics like 1.4529 were totally unblemished, while 316 was pretty good, but definitely not without some signs of attack.
Obviously, titanium is a case apart. The super-austenitics make sense because of their high nickel content (>20%), so it is going to be interesting see what 316 looks like when I next check. We know that the material specifications for all parts were OK because Martin checked them by XRF before installation. I’ll make sure I bring a magnet so I can add a bit more information. All I can see for now is that the attack on 316 looks superficial, and there is no sign of the “black ring of death” exhibited by the 304. But time could change all that. Let’s see what happens 🙂
That is very interesting. Please make a post whenever you visit it again, it would be super interesting to see how the different types of materials fared over there.
I certainly intend to follow-up. After all the work Martin Roberts put in establishing these tests sites, it is the least we can do. The last few years have been problematical. First there was COVID, then I smashed my brains in a climbing accident and now I simply have old age to beat back with my walking stick. What could possibly go wrong? 🙂