Stainless steel – can we believe what it says on the label?

We often find 304 pretending to be 316, but now I uncover a case of 303 posing as 304.

Yeah, yeah, I know, it was twenty years ago, and we do things better now. But, if we don’t understand how such things happened, how can we be sure they won’t happen again? I suspect this one is far more systemic than the odd bar of rogue steel making its way onto the production floor.

Introduction:

I am about to point to a well-known climbing anchor product, but will refrain from naming the manufacturer, because I believe the problem is more general, and likely applies across many manufacturers. Quite simply put, it comes down to this – there is not the money in the market to support the quality system that the manufacture of safety-critical components demand. Incoming bar stock is not scrutinized, or even specified, as well as it should be.

Here is an anchor of the type we will be considering.

And this is what’s left of such bolts after ten to twenty years on a sulphate cliff at Cabo da Roca.

It was common practice to refer to these bolts as being 304, and I certainly did so because they analysed at 8% Ni, 18% Cr and 0% Mo.

However, the reality is that there are other 18-8 steels such as 303 which fit this description. I didn’t have the means to measure S and thus settle the matter. Now that I look back, certainly there were suspicious inclusions in some samples.

Recently we found this:

Adjacent to the Northern Queensland city of Townsville there is a granite outcrop that has been home to rock climbers for many years. Recently a climber, checking the top anchors on one of the climbs, was alarmed to have both bolts of a double bolt belay snap on being test tightened.

Both bolts were of the longer, double-wedge, version of those I illustrated above, and one is shown below. This bolt would have been installed over 20 years ago.

Although the fracture was readily identified as an SRB attack, I was suspicious of such a diagnosis. It seemed very unlikely given the geology and geomorphology of the location. Furthermore, both sulphate spot-testing and wall-wash analysis failed to find a sulphate source.

Now this is where things get interesting.

In the past, I’ve had reason to wonder whether a stress-crack in a sulphur-enriched alloy like 303 might, at least partially, mimic an SRB attack on 304, but have been put off by the cost of the analysis for sulphur. But now, for the very first time, thanks to contributions to our Crag Chemistry Independent Bolt Failure Testing Fund, I was in a position to send a sample away for Optical Emission Spectroscopy (OES).

And it turns out that yes, those “304” bolts are in fact not 304 but 303, showing the elevated sulphur levels of that alloy. Full report is here.

Sample	Chromium (wt%)	Nickel (wt%)	Sulphur (wt%)
Stuart-1	17.7	8.5	0.19

For those readers who don’t carry stainless steel composition tables in their head, the value of 0.19% for sulphur is high compared with the upper limit for 304 of 0.03%.

Why is 303 used?

AISI 303 is a variant of AISI 304. This is an alloy where the sulphur content has been intentionally supplemented to create a population of manganese sulphide (MnS) inclusions throughout the fabric of the metal. For those production processes that involve the cutting rather than the forming of metal, these inclusions serve as chip-breakers which, by their very presence, improve the efficiency of metal removal.

In the past I have noted that, for the bolts in question, all have cut threads, rather than the more modern rolled type. The difference is immediately obvious when viewing a polished section, as shown below.

Only now does it occur to me that, for as long as manufacturers are thread forming using cutting rather than rolling processes, we can expect to find expansion bolts made of 303 rather than 304.

Sorry – no free lunch:

It’s great we can employ a myriad of MnS inclusions to reduce production effort, but unfortunately, there is a significant downside. These same inclusions, when they break the surface, compromise the passive chromium oxide film that renders stainless steel stainless. So, easier to machine for sure, but no longer truly stainless.

Every novice in the craft of engineering materials learns from his elders that there is no free lunch when it comes to austenitic stainless steels like 303 – their use has to be restricted to non-corrosive environments.

I will expand on this subject in another post. For now, I just want to cover two bullet points –

• There is much written on the pitting corrosion of these steels. It is severe. It is a showstopper.
• There is less written on stress cracking of these steels, but it is nevertheless very real. I can point to two quality studies where 303 climbing anchors have failed through SCC-like mechanisms. One of these is coastal and the other well inland .

The evidence is clear: 303 bolts should be restricted to indoor environments only.

The Euro-muddle of A2/A4:

I am as guilty of propagating muddle and confusion as the next person when it comes to slack use of these Euro categories for fasteners under EN ISO 3506. In this blog you’ll see examples of where I have written A2(304) and A4(316). This arises because it is common to say A2 when you mean 304 and vice versa.

I knew they weren’t strictly interchangeable descriptors of stainless steel, but it’s an easy 90% communication win, so I took it.

But here’s a trap! Firstly, the A2 category applies just to the finished bolt product, not the bar stock used to produce that product. Secondly, the category is very loose with respect to alloy composition, and both 303 and 304 bolt products can quite correctly be categorized as A2.

So, picture this. The manufacturer specifies 303 bar stock from his supplier because this is what his production machinery demands. He then puts the finished product in a box labeled A2. This is a perfectly accurate and legitimate product description to present to the market. The climber sees the product on the shelf and thinks, “Ah good – 304” and there you have a seamless route for getting a dangerous grade of steel onto the crag.

CorrectioN:

Thanks to reader Renato for pointing out the error. Whilst ISO 3506 places wide enough limits on Ni and Cr to allow both 303 and 304 into the A2 category, the sulphur content would force 303 down into the A1 category.
I guess I was trying too hard to be nice and find a plausible reason why 303 might accidently be marketed as 304, but that reason has just evaporated.

Conclusion:

Am I making a mighty fuss over just one bolt? I’ll let the reader judge. However, I feel obliged to sound the alarm with such evidence as I have. More will follow.

How did we come to this pass? I suspect it is because we love climbing, and we hate paying for the necessary quality, with equal vehemence. That’s surely madness? … but only to those who have never experienced the thrall of the climb.

---

Please Read

This post introduces a field of investigation that needs much more work. The keystone will be the OES analysis of such bolts as we encounter. For the avoidance of doubt the OES analysis has to be carried out by an accredited laboratory. This costs money.

To underpin this post we have paid AUD $385 for the one OES analysis. How we go forward from here is totally dependent on being able to pay for further analyses.