ISO 7434 only got stainless in 2024, and it added A2 and A4

This page came out of a real enquiry. On 25 August 2026 a distributor in Wolverhampton asked us to price a hundred thousand pieces of M2 × 2 din 553 A1 or A2 Stainless Steel — Slotted Setscrew with cone point. Nothing about that request is unusual; it is how the trade writes. Reading it against the current documents is what turned up four separate places where the words and the standards have come apart.

Sources are the free clauses on the ISO browsing platform for ISO 7434:2024 and ISO 898-5:2012, and the free preview of ISO 3506-3:2025. The dimension tables and the hardness tables are behind the paywall and none of them is reproduced here.

DIN 553 is withdrawn, and the whole first page of Google still sells it

Search for the part and you get ten product listings and no explanation. That is not a complaint about the results; it is the finding. Two of those sellers say the thing out loud in their own catalogue text, which is more than most.

Bossard lists the part under “Standard: DIN 553 (Standard withdrawn) ISO 7434”. Würth is more precise still: “ISO 7434 replaces DIN 553 | There is some similarity | Nominal diameters for M1, M1.4 removed”. Some similarity, not equivalence, and a concrete difference named. The scope of the 2024 edition runs M1,2 to M12.

So a line item that says DIN 553 is naming a withdrawn document, and the live document it points at is not a rename of it. Everyone in the chain knows roughly what is meant, which is exactly why nobody checks the places where it stopped being true.

A2 is the first of two blocks

ISO 3506-3:2025 is the third edition of the stainless set screw standard, and it is the document ISO 7434 points at for grades. Its designation system, in Clause 4, is two blocks separated by a hyphen: the stainless steel grade, then the hardness class.

The first block is a letter and a digit. A for austenitic, D for duplex. The second block is a number that is one tenth of the minimum Vickers hardness, followed by H for Vickers. There are two of them.

The standard’s own worked example reads: A2-12H specifies a set screw in austenitic stainless steel of grade A2, soft, with a minimum hardness of 125 HV. Note the word soft, which the figure gives as the fastener condition for 12H; 21H is soft or work hardened.

An enquiry that says A2 and stops has named the corrosion behaviour and left out the hardness. For a screw whose whole function is to press a point into a shaft, the missing block is the one that describes the only mechanical property this family gets.

The grades in Figure 1 are A1, A2, A3, A4, A5 and A8 for austenitic, and D2, D4, D6 and D8 for duplex. Duplex has one hardness class, 21H. Low carbon grades, at or below 0,030 %, can carry an L after the grade, as in A4L-21H. So A1 is a real grade in the mechanical properties standard, but the grades ISO 7434 added to its own product standard in 2024 were A2 and A4, which is the narrower list a supplier is working from. What the A number is telling you in the first place is a separate page.

Two standards, three editions, and the torque test never leaves the socket

Hardness is not quite the only mechanical requirement in this family. There is one torque test, and it has a qualifier that has survived every revision.

Carbon steel or stainless, 2009 or 2025, the proof torque belongs to the socket drives. A slotted set screw is inside the scope of both documents (ISO 3506-3 says it applies to fasteners of any shape), and it gets a hardness class and nothing else.

That is defensible. The proof torque test asks whether the screw survives being driven to its tightening torque, and a slot fails long before the thread does, so the number would describe the screwdriver rather than the screw. It is still worth knowing before someone writes a torque figure on a drawing next to a slotted part.

What the carbon steel standard leaves out about the joint itself (the shaft never appears in it) is its own page, and that page is also where the four point types and what each does to a shaft are set out.

M2 × 2 is shorter than the shortest length the 2024 edition classifies

The foreword lists what happened to the lengths, and it is unusually specific: for M1,6, l = 2 mm has been classified as too short length; the lengths for M2 (l = 3 mm), M3 (4 mm), M3,5 (5 mm), M4 (5 mm) and M5 (6 mm) are classified separately as short standard lengths, with the nominal length calculated to get at least 2,5 full pitches. Regular standard lengths are calculated for at least 4 full pitches up to M3,5, and 3 above it.

The enquiry asked for M2 × 2. The shortest length the 2024 edition classifies for M2 is 3 mm, and the reason is thread count rather than preference: at 2 mm on a 0,4 mm pitch there is not enough full thread left to reach the number the committee worked back from.

This is not a refusal. Short set screws exist and get made, and the paywalled tables are where the actual limits live — this page does not have them and does not guess. What changed is who is deciding. Below the classified lengths, the part is no longer described by the standard on the line item, so the thread count, the point diameter and what counts as acceptable become things to agree in writing.

One of the sellers on that first page of results is listing M1.6 × 2 under DIN 553 / ISO 7434 in A1, the exact size and length the 2024 foreword calls too short. The catalogue and the standard have drifted apart and the catalogue did not notice.

The cone angle stopped having a tolerance

Another line from the same change list: the point angle with tolerance of ±2° has been changed to a reference angle of 90° or 120°, without tolerance.

A reference dimension is not a rating and not an acceptance criterion; it is a number printed so you can see where the geometry came from, and it is controlled by the other dimensions rather than on its own. The 1983 edition let you inspect a cone point by putting a tolerance on its angle. The 2024 edition does not, and the values of d_t, the point diameter, were added to Table 1 in the same revision. The control moved from the angle to the flat at the tip.

Why a characteristic marked reference is not something you can reject a part against is set out separately, and the standard that names the ends of threaded parts, including the cone, is ISO 4753, a normative reference of ISO 7434.

Stainless stops thirty degrees earlier at the cold end

Both mechanical properties standards state an application temperature range, and they do not agree.

The stainless one is a WARNING in the scope, and it puts the choice back on the user for anything outside that band. The carbon steel one is a NOTE, and it goes further: outside −50 °C to +150 °C, and up to a maximum of +300 °C, it tells you to consult an experienced fastener metallurgist.

What to settle before this goes on an order

• Write both blocks. A2-12H or A2-21H, not A2. The second block is 125 HV or 210 HV minimum and it is the only mechanical property a slotted set screw is given.
• Name the live standard. ISO 7434 for a slotted cone point. DIN 553 still communicates, but it is withdrawn, and the replacement is described by its own sellers as similar rather than identical.
• Check the length against the classification. Below the short standard lengths the part is no longer described by the standard on that line, and the thread count becomes something to agree rather than something to cite.
• Do not put a torque figure next to a slotted part and expect a standard to back it. The proof torque test is written for socket drives in every edition of both documents.
• Say what the shaft has to look like afterwards. A cone point expects a matching seat. That decision, and the other three point types, are on the set screw page.

This is not one of the six steps. It shows up across them, or after assembly. Where the decisions that lead here were made is in specifying a screw, which sets out the order and why doing it out of order is rework.

Common questions

Is DIN 553 still a current standard?

No. It has been withdrawn and replaced by ISO 7434 for slotted set screws with cone point. Distributors still label parts DIN 553 and everyone understands what is meant, but two of them say so in their own catalogues: Bossard lists the standard as "DIN 553 (Standard withdrawn) ISO 7434" and Würth writes "ISO 7434 replaces DIN 553 | There is some similarity | Nominal diameters for M1, M1.4 removed". The 2024 edition of ISO 7434 covers M1,2 to M12.

What does A2-12H mean on a set screw?

It is the two-block designation from ISO 3506-3. A2 is the stainless steel grade, austenitic, and 12H is the hardness class, where the number is one tenth of the minimum Vickers hardness. The standard’s own example reads that A2-12H specifies a set screw in austenitic stainless steel of grade A2, soft, with a minimum hardness of 125 HV. The other class is 21H, at 210 HV minimum. Writing only A2 leaves the hardness unspecified.

Which stainless grades does ISO 7434 cover?

The second edition, published 2024, added grades A2 and A4 with hardness classes 12H and 21H; the first edition from 1983 carried no stainless at all. The mechanical properties standard behind it, ISO 3506-3:2025, classifies a wider list (A1, A2, A3, A4, A5 and A8 for austenitic, and D2, D4, D6 and D8 for duplex), and its scope note says the designation system can be used outside its stated limits provided the requirements are met. So A1 exists as a grade while not being one of the two the product standard brought in.

Is there a proof torque value for a slotted set screw?

Not in the standards. ISO 898-5:2012 titles its clause 9.4 as the proof torque test for hexagonal socket and hexalobular socket set screws of hardness class 45H. ISO 3506-3 restricts it to socket set screws in both its 2009 and 2025 editions. A slotted set screw is inside the scope of both, since ISO 3506-3 applies to fasteners of any shape, and receives a hardness class only.

Why is M2 × 2 an awkward size to order to ISO 7434?

Because 2 mm is below the shortest length the 2024 edition classifies for M2. Its foreword sets out that for M1,6 a length of 2 mm has been classified as too short, and that M2 at 3 mm, M3 at 4 mm, M3,5 at 5 mm, M4 at 5 mm and M5 at 6 mm are classified separately as short standard lengths, with the nominal length calculated to give at least 2,5 full pitches. Shorter screws are made, but below that classification the thread count and the point become matters to agree in writing rather than to cite.

Does the cone angle have a tolerance?

Not since the 2024 edition. The foreword records that the point angle with a tolerance of ±2° has been changed to a reference angle of 90° or 120° without tolerance, and that the values of dt, the point diameter, were added to Table 1 in the same revision. A reference dimension is printed to show where the geometry comes from, not to reject parts against.

References

• ISO 7434:2024, Fasteners — Slotted set screws with cone point. Second edition, June 2024, six pages. Foreword and its list of main changes, Clause 1 and the normative references
• ISO 3506-3:2025, Fasteners — Mechanical properties of corrosion resistant stainless steel fasteners — Part 3. Third edition. Foreword, Clause 1, Clause 4 with Figure 1 and Table 1, and the contents page
• ISO 898-5:2012, Mechanical properties of fasteners made of carbon steel and alloy steel — Part 5. Third edition, June 2012, fifteen pages, confirmed on review in 2022. Foreword, Clause 1, Clause 3 and the contents page
• ISO 3506-3:2009, second edition, withdrawn. Contents page, read for the wording of its proof torque clauses

ISO 7434:2024 and ISO 898-5:2012 were read from the free clauses published on the ISO Online Browsing Platform, which carry the foreword, scope, normative references and contents but stop before the requirements. ISO 3506-3:2025 was read from its public preview; Clause 4 including Figure 1 and Table 1 falls inside that preview, and Figure 1 is an image, so it was rendered from the preview PDF rather than extracted as text. The dimension tables of ISO 7434, the chemical composition table of ISO 3506-3 and the hardness table of ISO 898-5 are all outside what is free, and none of them is reproduced here. This page gives no length limits for M2, no composition figures and no hardness values other than the two class minima printed in Table 1 of ISO 3506-3:2025. The four carbon steel hardness classes are not quoted here as current values; where they appear on this site they come from the 1998 preview of ISO 898-5 and are labelled as such on that page. Statements about DIN 553 go no further than that it is withdrawn, and the source for that is the catalogue text of two named distributors, Bossard and Würth, not the DIN document, which has no free preview. The listing of an M1.6 × 2 part under DIN 553 / ISO 7434 is a third distributor’s live catalogue entry, cited to show the drift rather than to judge the part. The enquiry quoted at the top is a real message received on 25 August 2026 and is reproduced as written.