One alloy, two standards, three numbers

Published 19 August 2026. Corrected 20 August 2026 — see the note at the foot of this page.

The numbers

Read the table before reading the argument, because the middle row is the one that changes the story. The headline pairing — 450 against 500 — puts an ASTM plate minimum next to an EN cold rolled minimum. Two variables move at once. Hold the product form still and the standards are 450 against 460: two per cent, not eleven.

The other nine points are the product form. EN 10088-2 sets a different floor for cold rolled coil and sheet than for Quarto plate, because cold rolled product is stronger and the standard says so; ASTM A240 covers plate, sheet and strip under one figure and does not make that distinction at all. So the two documents disagree less about the alloy than they disagree about how finely to slice the product.

Both EN rows are on the same page of the same datasheet, in the same table, one line apart. Outokumpu’s Forta range datasheet gives EN 10088-2 minima as its primary column, names the product form per row, and lists ASTM A240 imperial equivalents alongside — which is correct, complete, and exactly the detail that gets stripped out when a figure is copied somewhere else.

Source: Outokumpu Forta range datasheet, May 2015, rows C, H and P for Forta DX 2205.

Why two numbers exist

Not because one body is more careful than the other. Standards bodies set minima by their own process — the population of material considered, the product forms covered, the test practice assumed, the margin policy applied. Two committees looking at the same alloy can land on different floors without either being wrong, and both floors are honoured by mills selling to them.

There is no reconciliation to be found. The alloy does not have one true minimum yield strength. It has a minimum per specification and per product form, which is a subtly different object and the whole point of this note. Two coordinates, not one — and a quoted figure that names neither is not a specified minimum, it is a number that resembles one.

What actually goes wrong

Nobody misreads the datasheet. What happens is slower.

An engineer works a project governed by ASTM. The figure 450 becomes the number they know for 2205 — not written down anywhere, just known, the way you know the density of steel. Two years later they are on a job governed by EN, working in cold rolled sheet, reach for the number they know, and size a section against a floor 50 MPa below the one that governs.

The result is conservative, so nothing fails and nobody notices. Run it the other way — EN figure carried onto an ASTM job — and it is not conservative. It is a section sized against a floor the material is not contractually required to clear.

Neither version produces a dramatic failure. Both produce a design whose margin is not the margin the engineer believes it is, and that is the category of error that survives review, because a review checks the arithmetic, not the provenance of the constant.

Three things this generalises to

A property value is not a property of the material. It is a property of the material as described by a document, for a stated product form. Condition, temperature, test method and the specifying standard are all part of the value, not context around it. Strip them and what is left is a number that cannot be checked. This note began life attributing the 500 MPa figure to plate; correcting that single word moved the apparent disagreement between two standards from eleven per cent to two.

This is the same failure as specified minimum versus typical value. There, one number is a floor and the other is an average. Here, both are floors from different documents. In both cases two different objects are wearing the same units, and the units are the only thing carried forward.

Which is why a source is not decoration. Under the E-ladder, a specified minimum reaches E5 only when the record carries the standard designation, its edition, the clause or table, the product form and the condition. That requirement is not bureaucratic thoroughness. It is precisely the information that makes 450 and 500 distinguishable rather than contradictory.

Our own version of this problem, since we are recommending discipline

On 2205 vs 316L we quote both figures, and both sit at E3, not E5 — because the ASTM standard is paywalled and we did not open it. Under the demotion rule, a claimed level a reader cannot verify from the record alone drops one level, so that is where they sit, and the pages say so.

Which is mildly embarrassing on a page arguing that the standard is the point. It is also the honest state of the record, and the fix is not to write E5 and hope — it is to open A240, record the edition and the table, and promote the value. Until then the level stands.

Correction, 20 August 2026

This note was published on 19 August 2026 with a two-row table, both rows labelled plate. That was wrong: the 500 MPa figure is the cold rolled coil and sheet row, and the EN plate minimum is 460 MPa. The error was found on 20 August 2026 when the Forta datasheet was opened again to confirm the product form for a different grade.

The correction does not weaken the argument, it sharpens it — the original version blamed the whole gap on the standards, and most of the gap was the product form. It is also a demonstration of the failure this note is about, committed by the note itself, which is the reason it is recorded here rather than quietly fixed.

The habit

When you write a yield strength into a design note, write the standard next to it. Six characters. It is the difference between a number your successor can verify and a number they have to trust.

And if you are reading a value somewhere that does not name a standard, you have not found a specified minimum. You have found a number that resembles one.

What this note does not cover

• Why the two committees chose different floors. We have not traced the committee record and are not going to speculate.
• Other product forms. Bar, tube, forgings and castings have their own minima and their own gaps between systems. Wire rod and bar rows for this grade sit at 510 and 650 MPa on the same datasheet, marked as Outokumpu typical values rather than EN 10088-2 minima, which is a third kind of number again.
• Test orientation, which is a fourth coordinate and turns out to matter more on ferritic grades than on this one. It has its own note.
• Other alloys. 2205 is the clearest illustration, not a special case.
• Whether either figure is conservative in service. A specified minimum is a contractual floor, not a prediction of your heat.