Technical paper

The letter Z is telling you something

Why 5 kg spools of gasless wire can’t be qualified to AS/NZS 1554.1, and almost nobody selling it spells out what that means.

Graham Fry, Welder | Principal Welding Engineer

I get the same call a few times a year. A fabricator has done or won work, the drawings call up AS/NZS 1554.1, and someone has asked for a qualified procedure. They’re running gasless, it’s convenient, it’s portable. Usually 0.9 mm, sometimes 1.2. Can we qualify it? Well…

5 kg spool territory is almost exclusively occupied by T11 type wires. T11 wires get used incorrectly everywhere, unknowingly. Everything from small structural jobs to significant structures on mine sites. I say with confidence as I write this article, it is happening right now. With the convenience and portability of 5 kg and operator appeal of T11 wires in general comes a lack of mechanical properties. Like everything in welding, it’s a trade-off.

Can we qualify it? We often get asked. Well, we can run the test. What I have to tell them first is that in my experience the test usually doesn’t go their way, and that even if it did, the procedure was never going to be prequalified in the first place.

That conversation is always uncomfortable, because by the time it happens the wire has been in the shop for two years and it’s already gone into things that matter.

What the classification actually says

Take Lincoln Innershield NR-211-MP, because Lincoln publish more detail on this than anyone. The 2024 Australian consumables catalogue lists it as:

Read the designation left to right. T for tubular cored. 49 for the strength class, 390 MPa minimum yield, 490 to 670 MPa tensile. Then Z.

Z is the impact designator. In ISO 17632 that position sets the temperature at which the deposited weld metal has to achieve a minimum average absorbed energy, 47 J under System A or 27 J under System B. Every other symbol that can sit there is a temperature, including A and Y at +20 degrees C. Z means no requirement. Not “not tested”. Not “tested and low”. No requirement at all.

The ‘Z’ is printed on the datasheet. It has been sitting there in plain sight the whole time.

AWS A5.20 says the same thing in a different language. Lincoln reproduce the mechanical property table in their own Innershield welding guide, and for the T-11 classification the row reads 70 to 95 ksi tensile, 58 ksi minimum yield, 20% minimum elongation, and under minimum Charpy V-notch impact energy: Not Specified.

Lincoln then say it in plain English:

NR-211-MP is on that list. So is NR-212.

Three classification systems, one answer

If you think the Z is an ISO quirk, look at what the same wire is called everywhere else.

AWS replaced A5.20 and A5.29 with the open classification system in A5.36, and Lincoln print the cross reference in their own guide. Under A5.36, NR-211-MP becomes E71T11-AZ-CS3. The character after the A, which is the as-welded condition, is the impact designator. It’s a Z.

Two rows further down the same table, under Lincoln’s heading “All Position/Structural (with CVN toughness properties)”, NR-232 and NR-233 become E71T8-A2-CS3-H16 and NR-203MP becomes E71T8-A4-CS3-H16. Same position in the string. A 2 and a 4, which are temperatures.

Now take Fabshield 21B, the Hobart wire sold widely here through WIA and industrial supply. Its Australian data sheet declares:

The Hobart sheet for the same product adds the Canadian classification:

There it is again. AZ. Canada looked at this wire and wrote the same letter in the same position that ISO did.

So on one product you have three independent classification systems, written by three different bodies on three continents, and all three of them record the same thing: this wire has no impact requirement. Then the datasheet says it in English as well, “CVN Impact Values Not Required”, which is more than most of them do.

Credit to Hobart for printing it plainly. Nobody is hiding anything. It’s just that four separate declarations of “no toughness” apparently still aren’t enough to stop the wire ending up in a structure.

Why can’t it be prequalified?

AS/NZS 1554.1 Clause 4.3 sets out when a welding procedure is deemed prequalified. One of the conditions is that the consumables are prequalified in accordance with Clauses 4.3.2 and 4.6.

Clause 4.6.1.1 requires that consumables are matched to the steel type in compliance with Table 4.6.1(A), and adds that the impact test temperature of the consumables, as specified in the relevant consumable Standard, shall not be warmer than the design service temperature.

The project drawing says ‘Welding consumable must be a minimum of E49 XX’ or be classified as having a minimum tensile strength of 490 MPa. That’s the strength of the weld. Go to the FCAW column of Table 4.6.1(A) and the confusion starts. The prequalified consumable for AS/NZS 3678 Grade 350, Steel type 4, is ‘B-T490U’. Both have ’49’, so it’s good to go? You will notice that no T11, no T1 and no T8 ‘usability’ type classification appears in it at all, in either the A or the B system.

AS/NZS ISO 17632 A versus B and how they relate to the AWS system will be the subject of a separate article, so as not to confuse the current topic.

You must read the full designation and satisfy all of it for the consumable to be prequalified. The second part after the ’49’ is the ‘0U’. That 0U matters. It classifies minimum weld toughness, and it carries a numeric impact designator, 0, 2 or 3. 0U requires a toughness value or absorbed energy of 47 J, the ‘U’, at 0 degrees C, the ‘0’.

Note 2 to the table lets you go one way only: consumables with a higher impact grading than shown are also acceptable. Higher. Z is not higher than anything, it’s the absence of a grading, and not because the consumable manufacturer couldn’t be bothered to test it, but rather they already know the answer.

Then read Clause 4.6.1.1 against a Z wire. The clause asks you to compare the consumable’s impact test temperature to the design service temperature. A T49 Z wire has no impact test temperature. There is nothing to put on the left side of that comparison. You can do a test coupon to try and satisfy the clause, but in my experience you are going to be disappointed. The impacts do not meet the requirements for qualification to the standard, and they will not meet the basic internationally recognised requirement of 27 J at any temperature.

Worth noting how the T-8 wires behave under the same test, because it shows the mechanism working. Lincoln’s Australian catalogue classifies Innershield NR-232 as T55 3 T8-1 NO A H10 and NR-233 as T49 3 T8-1 NA UH10. Impact designator 3. That’s in the table, and those wires carry a published requirement of 27 J minimum at minus 30 degrees C. Same self-shielded process, same catalogue, completely different position under the standard. Wires are not wires. The T8 and T11 are different animals, and this article is about T11 wires.

Then people try to qualify it

This is where it gets expensive. The other route is Clause 4.6.2, qualification of the consumable by testing, which brings the transverse butt tensile and Charpy V impact requirements of Table 4.6.2 into play.

Now you’re asking a consumable that was formulated with no toughness requirement to demonstrate toughness. Self-shielded wire has no gas envelope, so the arc has to be protected chemically. That means aluminium, heavy deoxidisers and denitrifiers in the core, and a weld metal chemistry built to tolerate atmospheric nitrogen rather than exclude it. On Lincoln’s own published analysis for NR-211-MP the specification permits up to 1.8% aluminium in the deposit and the typical result runs 1.3 to 1.6%. That is not a toughness-friendly place to be, and nitrogen pickup doesn’t help either.

Hydrogen is the other one nobody talks about. Cigweld publish it for their WeldSkill gasless wire, which is the same classification: 15.0 to 20.0 mL of hydrogen per 100 g of deposited weld metal, on a good day. Lincoln’s ISO designation ends in H15. Whatever else that is, it isn’t low hydrogen, and on anything with restraint or thickness it changes your preheat conversation entirely.

Put all that together and the Charpys come back where you’d expect. I’ve watched more than one client, out of desperation, spend real money finding that out.

The thickness limit, which is a separate problem again

Toughness is only half of it. Lincoln’s Innershield guide carries this warning, and I’d frame it and hang it in a few shops I’ve been in:

The limits by diameter, from the US guide:

Note: the 1.1 mm is what is sold as 1.2 mm in Australia.

The Australian catalogue prints the same restriction rounded: 0.9 and 1.2 mm to 8 mm, 1.7 and 2.0 mm to 13 mm. Which is its own small problem, because the diameters actually offered on that page are 0.9, 1.1, 1.7 and 2.0. There is no 1.2 mm NR-211-MP in the range. The 1.1 mm, or 0.045 in, is what is sold as 1.2 mm in Australia, but the 1.1 mm wire a buyer walks out with isn’t named in the restriction at all. Lincoln also make the point that if you’re joining plates of different thickness, all of them have to sit inside the limit, and that the limit applies whether the weld is single pass or multiple pass.

Now look at which end of the range carries the tightest ceiling. The small diameters. Those are the sizes on the shelf everywhere, on 1 kg and 4.5 kg spools, feeding every small inverter in every shed, ute and now construction site in Australia. The wire with the 8 mm ceiling is the wire in the widest circulation, and the 1.7 and 2.0 mm sizes that get you to 13 mm are supposed to be industrial supply only.

Someone forgot to tell the local welding and fab shops.

Lincoln go further in their US Innershield guide and split the range into headed groups. NR-211-MP sits under “Sheet Metal to Thinner Plate/General Purpose (no specified CVNs)”. Two groups further down sits “All Position/Structural (with CVN toughness properties)”, which is where NR-232 and the rest of the T-8 family live. They have drawn the line themselves, in their own literature, and put the T-11 on the wrong side of it for structural work. That split does not appear in the Australian catalogue, where NR-211-MP and NR-232 both sit under the same heading, “Flux Cored Wire, Self Shielded, All Positions”. I applaud Lincoln for having the best technical literature on the limitations of the product.

Nobody agrees on how thick

Here’s the part that should worry a specifier more than anything else in this article. Take one AWS classification, E71T-11, and collect what the manufacturers publish as the maximum plate thickness.

Same classification. Same specification. The published ceiling runs from 7.9 mm to 20 mm depending on whose sheet you happen to be holding, a spread of two and a half times, and a third of the market doesn’t state one at all.

Fabshield 21B is the interesting one, because the same product is sold on two different numbers in two different markets. The American sheet says light structural under 19 mm. The Australian sheet says single and multi-pass welding of thin gauge to 20 mm thick. Lincoln, on the identical AWS classification, says 8 mm for the diameters most people buy.

Somebody is wrong. The specification itself is silent, which is how you get a range this wide, and it means the thickness figure on the sheet is a manufacturer’s recommendation and nothing more. It is not a code limit and it will not save you in a compliance argument. What will decide the argument is Table 4.6.1(A) or a consumable batch test in accordance with ISO 17632, and neither cares about thickness at all because the consumable never gets that far.

Read the applications list, it’s telling you

You don’t need any of the above to work out what a wire is for. Read the applications list on the datasheet. The manufacturer has already told you which market they built it for, and they’re rarely subtle about it.

• Fabshield 21B, Australian sheet: machine parts, tanks, gates, frames, sheds, general fabrication.
• Cigweld Shield-Cor X11, retail listing: gates, fences, steel frames, galvanised tanks, ornamental iron work.
• Bossweld Gasless 11: agricultural and rural maintenance, farm maintenance.
• Lincoln Powercraft T-11: perfect for use on small portable MIG welding machines, ideal for general fabrication, rural and DIY.
• Lincoln Innershield NR-211-MP: small diameters ideally suited to DIY welders.

Now turn to Innershield NR-203MP, a T-8 wire, in exactly the same Australian catalogue. Applications: general plate fabrication including bridge fabrication, hull plate and stiffener welding on ships and barges, storage tanks, structural welding, offshore welding in TKY joints. Conformances: AWS A5.20 E71T-8-JH8, ABS 3YSA H15, Lloyd’s Register 3YS H15, DNV-GL III YMS H10, Bureau Veritas SA3YMH, CSA W48 E491T8-A4-CS3-H8.

Five classification society approvals and a CSA impact designator of 4. That is what an industrial structural consumable’s paperwork looks like.

Gates, sheds and galvanised tanks is not a coy way of saying structural. It is the manufacturer describing gates, sheds and galvanised tanks. When the application list is domestic and agricultural fixtures, and the conformance block has no classification society approval in it, and the impact position in every classification string is a Z, the wire is telling you what it is in three separate places on the one page.

Ornamental iron work is not a load path.

The Americans wrote a code about this

None of this is new. Northridge was 1994. The SAC and FEMA investigation into the connection failures ran for the rest of the decade, and low toughness self-shielded consumables were one of the things it put in the frame, alongside backing bars, weld access hole geometry and the pre-Northridge connection detail itself. AWS D1.8 followed in 2005.

Under D1.8, filler metal used in the seismic force resisting system has to demonstrate a minimum of 27 J at minus 18 degrees C, by AWS classification or by manufacturer certification. Demand critical welds need more than that, 54 J at 21 degrees C, demonstrated across a heat input envelope. A wire classified with no toughness requirement doesn’t get past the first line.

We got to a similar place by a different route. AS/NZS 1554.1 simply never listed the classification.

What the market actually tells the buyer

Here’s what bothers me. Lincoln are the ones who publish the restriction properly, and they’re doing it on their premium industrial brand. Look at what the buyer gets everywhere else.

Several things in that table are worth sitting with.

Lincoln’s own Powercraft T-11 first. Same AWS classification as NR-211-MP, sold in 0.8, 0.9 and 1.2 mm, the retail sizes. The catalogue page carries no ISO designation, no plate thickness restriction, and no minimum property row, only typical results. Its key feature list offers ‘Versatile’ and ‘Single and Multi-pass* welding’.

Go looking for that asterisk. There is no footnote. Not on the page, not on the facing page, not anywhere in the catalogue. The one mark on the sheet that was going to tell the buyer something has nothing behind it, while the identical classification on the premium brand two pages earlier gets the full restriction spelled out. Same company, same standard, two very different pictures depending on which shelf the spool is going on. The deposit composition table on that page is labelled “Typical Results” and then prints the A5.20 maximum limits, which is not what a typical result is.

The Michigan sheet is worth a look too. The ISO string on it is a T14 and the AWS classification on the same sheet is E71T-11. Those two do not correspond. One of them is wrong and the buyer has no way of telling which.

Hyundai are an interesting case for the opposite reason. Their catalogue sheet prints Y in the impact position, not Z, and Y is +20 degrees C. Their technical data sheet for the same wire prints Z. If it is genuinely a Y, it still fails Clause 4.6.1.1, because +20 degrees C is warmer than any design service temperature you would sensibly nominate in this country. So the answer doesn’t change, but it does mean “look for the Z” isn’t quite enough. Look for anything that isn’t a 0, 2 or 3.

Then there’s that word “structural”, which turns up on datasheet after datasheet for a wire that isn’t listed as a prequalified consumable in Table 4.6.1(A). Nobody is lying. “Light structural work” isn’t a defined term, and no consumable manufacturer is obliged to tell you what your project standard requires. But when a wire is listed for civil construction, recommended for heavier structural work, or sold as light structural, the person buying it reasonably concludes it’s an industrial wire.

Then it goes into a mezzanine. A trailer chassis. A lifting frame. A tank stand. And nobody finds out until an inspector asks for the procedure, an expert witness reviews the documentation, or something cracks in the cold.

Where the wire is actually fine

T-11 is a good product used properly, and it deserves to be defended on that ground. Non-structural sheet and thin gauge work. It loves galvanised sheet where the slag system earns its keep. Site work in wind, less than 10 km/h, where a gas shielded process just won’t hold for non-structural work. Repairs and general fabrication that carry no code obligation and no impact requirement. This is left to user choice. AWS D1.3 covers sheet steel to 3/16 in and accepts any FCAW filler metal conforming to A5.20, A5.29 or A5.36, which takes E71T-11 in generically. That’s the territory for it when the application is considered. It runs beautifully in comparison, the slag peels, the operator appeal is genuinely the best in the self-shielded range, and that’s exactly why it ends up in the wrong place. Nobody misuses a wire that’s horrible to run.

Hobart put the AWS D1.1 position on T-11 at less than 1/2 in base metal in their own technical literature, and note plainly that T-11 wires do not require impact toughness testing. Their words, on their own product.

If you want a self-shielded wire that will stand up in a structural context, the T-8 classifications exist for precisely that reason, under A5.20 for the carbon steel grades and A5.29 for the low alloy ones, and they carry a real toughness requirement. They’re harder to run. They typically do NOT come in 1.2 mm, and are not readily available in 5 kg spools. That is the trade-off you’re making, and it’s the trade the code expects you to make, yet many using it are ill-equipped to navigate the judgement call.

What to do about it

Pull the datasheet for the gasless wire in your shop and find the impact designator. If it isn’t a 0, 2 or 3, that wire can’t sit on a prequalified procedure to AS/NZS 1554.1, and you need to know exactly what it’s currently welding.

Check the diameter against the thickness limit, not against what the wire feels like it can do. A 0.9 mm wire is capped at 8 mm plate regardless of how good the bead looks on 16 mm. And remember whose number it is: on the same classification the published ceiling runs from 7.9 mm to 20 mm, so the figure on your sheet is a recommendation, not a code limit.

Read the applications list before you read anything else. If it says gates, sheds, tanks and rural maintenance, and there isn’t a classification society approval anywhere in the conformance block, you have your answer in about four seconds. If you are unsure, the decision sits with the asset owner or lead contractor. Get that decision in writing.

If you’re a designer or a principal, name the consumable requirement in your specification instead of assuming the standard reference covers it. “To AS/NZS 1554.1” is expected to carry adequate detail in most specifications I read.

And if you’re a supplier, publish the restriction, inform the industry and DIYers. Lincoln do it on NR-211-MP and the industry is better for it. The rest of you have the same information sitting in your own classification string.

The wire isn’t the problem. The problem is a lack of education and awareness. The wire does exactly what its classification says it does, which is nothing in particular about toughness. The problem is that the only person in the chain who’s read the classification is usually the one holding the failed Charpys, after the fact.

What the ‘key welding person’ should learn from it

The key person, welding supervisor or welding coordinator, should read the standard, the data sheet, the drawing and in particular the notes. One piece at a time, understand what those ‘things’, in this case digits, mean. When in doubt, ask questions, no matter how dumb they seem. Find yourself a good mentor or peer who can act as a sounding board.