Buying metal - caveat emptor.

More a beginner's observation than a question perhaps.

A while back I started a topic on thread cutting with carbide inserts - thanks to advice from here and experimentation I thought I'd cracked it. Imagine my horror when I needed to make an M16x2 thread on the end of a bit of 3/4" EN1A bar and found I was back to square one, or worse.

After much faffing about with all the variables I sent a street urchin round to my mate Sherlock - he came came back with the response "When you have eliminated all which is impossible, then whatever remains, however improbable, must be the truth. It is the metal". I didn't really believe him, but chucked up a bit from another bar anyway.

Annoyingly he was was right yet again! Both (supposedly) EN1A from the same supplier, green blob on the end of the bar, but behaved quite differently. How can this happen?!

Watson.

Edited By Robin Graham on 19/01/2020 23:45:57

Edited By Robin Graham on 19/01/2020 23:59:42

I don't think anyone would call that top piece free cutting.

I just bought a UNC die and a bar of EN1A from GLR to avoid exactly this little snag, I still have faith.

Hmmm.

Was it the very end of the bar? Sometimes the end bit is not much good. Something to do with the production process.

Yes it was - I just turned off the wonky end of the rolled bar, perhaps about 3/16". Maybe that's it.

Thanks, Robin.

I tend to buy most of my steel by the bar length. Usually 3 metres long, this is cheaper and I like having a good stock.

As standard practise I cut the first 2"/50mm off the end and throw it in the scrap box.

As someone who spent a lot of my working life bughunting software, I can tell your mate Sherlock that 'eliminating the impossible' opens up a truly vast field of wasted effort in searching for a cause. You'll start by looking for the most impossible answer, and before long you'll've checked so many candidates that you'll never remember which one's next or whether it's sufficiently different to those you've already checked to justify another look. It's absolutely the worst possible way to investigate anything.

The two most likely reasons are:

i) work-hardening at the end of a bright-drawn bar

ii) inclusions of different steel types in a cheap gash melt.

Even buying whole lengths is no protection. I bought several full lengths of allegedly bright EN1A from a small local supplier approaching a decade ago and reckon there were at least three different types of steel in the bunch. Different heat treatments and conditions too. Judging by how they behaved when turning and cutting probably only 3 of 8 could be rightfully treated as close enough to EN1. First part I tried to thread was rather worse than Robins first example. Starting some way back down the bar too. The good bits threaded beautifully.

Fortunately most of that particular job was rough work. Bit about that long with a hole about that size sort of thing. But a couple of lengths were going into stock. Sorted good ones for that. Still not exactly the same tho'.

Dropping into stuck record mode we really need to stop buying steel by EN numbers and use more modern designations. 230M07 rather than EN1A for example. The EN specifications are just too sloppy to work with these days, especially as there is little control of material condition. EN3B is a good example. Nearest is probably 070M20 but if you get bright bar it could well be 080A15, which should be supplied with a test certificate to show the properties match adequately. That one doesn't matter much as folk like us will be hard pushed to notice any difference between 070M20 and 080A15 when machining.

There is a lot of overlap between steel grades. We are more concerned with condition, which generally defines how it works, than exact composition. If it waddles like a duck, quacks like a duck then it is a duck. But if, metaphorically speaking, you order duck and get bombay duck on the plate you are, I think, entitled to be a bit miffed. Which is about what happened to Robin.

It would be nice if we could find someone with plenty of experience in the modern supply chain able to give appropriate modern designations for materials that will work well for us and have consistent properties regardless of supplier.

Clive

Edited By Clive Foster on 20/01/2020 09:47:02

Edited By Clive Foster on 20/01/2020 09:47:35

Edited By Clive Foster on 20/01/2020 09:48:00

A bit of green paint and you believe it. Have a look at some of the offerings off ebay. every length has the end painted and that doesnt get done at manufacture. Buying ad hoc off all and sundry is prone to disaster and there are a few out their that want your money at any cost, even a dab of green paint.

Find a good supplier and stick with them. Cheap en1a, that will be the day

I bought some 100mm cast iron from a well-known supplier. It was for some cylinders for my loco. I started machining it, thinking, "This is funny cast iron swarf!" I cut off a small piece of the bar and put it in the vice. I hit it with a hammer. It bent through 90°. In all fairness, they quickly replaced the steel bar supplied with the correct material, but I do wonder sometimes with commercial purchasing, which involves Certificates of Conformity, how often the C of C is incorrect...

I I keep 6 different grades of steel in stock and use electrical tape to ID them.

As I sell the stuff I have to be careful as once cut it is nigh on impossible to tell them apart.

All full bars have a ring of tape around them as they come in. I use a ring of black tape to show imperial which makes it easier to see on the rack.

Different suppliers use different colour schemes so don't assume that if it is green it must be EN1A.

The colours my supplier uses (and by extension I use) are :-

Green = EN1A

Purple = EN1Apb (leaded)

Blue EN3B (BMS)

Yellow EN8

Brown EN24T

White EN16T

Regards.

David.

Most likely it's the bar end. Whilst Black Steel and Bright Mild Steel have exactly the same chemical composition, their physical properties are noticeably different. It's because rolling steel Bright Mild Steel involves reheating and work hardening. In the same way cutting bar steel adds an area of suspect behaviour.

Can't agree with Mick's suggestion the cause is 'inclusions of different steel types in a cheap gash melt', because it implies steel is made haphazardly in small quantities. Actually steel making is high-tech, with perhaps several hundred tons of metal heated well above melting point, chemically analysed and then purified in the furnace to meet a particular specification. Slag eroding off the furnace might cause inclusions, but unmelted ball-bearings are unlikely. It is possible for trouble to occur later when the steel is cooled and processed. Cooling too fast or too slowly will change properties, as will cutting fast and hot, rolling, or missing out stages such as Normalisation because many big customers prefer to do their own.

Another problem might be our determination in the UK to buy 'EN1A'. "Emergency Number" specifications were forced on the British Steel Industry by government because the War effort demanded consistent steel, rather than whatever Hadfield, Firth, Colvilles, Dorman, Consett, Guest and all the others happened to be pushing as their proprietary free-cutting mild-steel. Although chaps swore by certain Brands, defeating Nazi Germany required Industry to behave consistently. Standardisation did the job so well EN numbers survived: no-one went back to the pre-war buying methods.

As the Emergency Number system is long gone, I suggest 'EN1A' today isn't a consistent product. As far as I know, no-one in the world makes actual EN1A, though there are many steels like it. Therefore what you get is a similar steel, like 220M07, or 230M07, or SAE1113, or SAE1213, or Workstoff 1.0711/1.0715, or JIS SUM25 or whatever else the supply chain thinks best. Almost for sure it's made in the Far East in a modern steel plant.

I feel it's a good example of old-fashioned British engineering conservatism causing unnecessary confusion: EN1A entirely ceased to be in 1991, yet it's still what I have to ask for in my local stockist. I've had better results since telling them I want to machine the steel. Pretty sure I'd get a different mild-steel if I said it was to be welded. I'm fairly sure that today's metal 'quality' problems are more to do with people having the wrong specification rather than the metal itself being faulty. There are a tremendous number of different steels in use compared with the last century. Life isn't simple, especially machining unknown scrap or one of the less suitable alloys. I reckon the Golden Age of straightforward home engineering ended about 1970, since then Industry has moved ever further away from 'our' materials and techniques. Unlikely that anyone will make an IMLEC winner featuring stampings, ground and laser cut components, welded stainless steel, extruded & die-cast plastics, robotic assembly, a FEM analysed and computer optimised design, low carbon emissions, and half a dozen microcontrollers!

Simple answer to Robin's question, buy free-cutting steel, check how it cuts#, and avoid the ends!

Dave

# My stock box contains samples of 'EN1A' rods and strip bought at different times. Although it all machines reasonably well there are obvious differences between some samples. The differences are more noticeable with Carbide than HSS, requiring up to 20% more-or-less speed, depth-of-cut and feed-rate for best finish.

Some fine "gash melt" bar

Edited By Hopper on 20/01/2020 11:36:50

Like Chris I tend to buy small bar stock (<2" diameter) in the standard 10ft lengths. i haven't generally had a problem using bar ends on cold drawn steel bar. I have had occasional problems turning and milling material, even that sourced from the larger steel stockholders.

The tolerances on the composition of most materials are quite wide and I suspect get minimal testing. If you're making a low carbon steel are you going to be worried about chucking in some medium carbon alloy steel, like a bicycle frame? For critical applications one can buy certified materials which should be what they say on the tin and possibly come with a composition certificate. But it'll cost you.

It probably minimises, but doesn't eliminate, the problem by purchasing from trusted suppliers, but it depends upon usage. As an example the large bar of supposedly EN1A I turned recently to make a tapered former had some odd surface effects. I bought cheap from Ebay, but was fully aware that there might be some issues. However, the former will get used for making two chimneys and then be recycled in some form, so I wasn't worried about odd or poor finishes due to the material.

Andrew

I don't think us hobbyists buy enough material to influence the market, and we have little chocie but to accept what we are sold.

The stockholder I use when buying in bulk (3m bars, chopped to 3 x1m) still sells by EN numbers because that's what all his big customers use. If I asked for 230M07 he'd offer me EN1a (after he stopped laughing, probably).

That said, if I can be bothered to haul into West Brom I can get 'modern specs', but as all the stuff I have got locally is ok...

Neil

EN1A is so much more understandable that 230M07.

As for C of Cs, these have to be used with a degree of scepticism. I used to work in the aircraft industry where C of Cs are mandatory but it didn't stop a Californian steelworks from falsifying the paperwork.

Brian

Maybe, but mis-use of rebar is a famous problem, including steel recycled directly after a clean-up from burned down buildings.

What's the most likely cause of Brittle Steel? Excess Phosphorous, or maybe Sulphur, possibly due to bad Coke or not adding Magnesia to the melt. A chemical test would detect this problem, easy to spot and fix. Faulty heat treatment is a distinct possibility, one way of speeding up production is to spray hot rebar straight from the rollers with a carefully measured dose of cold water rather than applying a second heat treatment. Spraying just the right amount of water has the desirable effect of leaving a hardened skin around a soft but tough interior. Too much water will harden the entire rod leaving it brittle. It's always possible the clown responsible will cover up the mistake rather than admit he's blundered.

Not many of us buy Iranian Rebar, but the video illustrates why people pay extra for Certificates and Testing even when buying from reputable suppliers.

I haven't cut enough metal myself to confirm anything. Telling the difference between Pine and Teak is easy, identifying metals much harder, so what do I know! But I think it significant that manufacturers who depend on steel meeting a specification, whether sheet, plate, girders, railway lines or for machine making, all seem to have much less trouble with 'quality' than Model Engineers. Something is going on, but I doubt someone makes special nasty metal just to annoy home machinists. But the steel makers do produce millions of tons of alloys unsuitable for general purpose hobby work. My suggestion is we're randomly using varieties of steel of unknown specification, both in composition and the way it's been post-processed with heat-treatment, or by rolling. We see lots of variation because we buy metal in small quantities from different sources and spread the buys out over time. By professional standards, we are inconsistent.

Dave

Only if you think you know what EN1A is! Doesn't have much currency outside the UK, for example EN1A isn't mentioned at all in my copy of Machinery's Handbook.

The later British Standard system is logical, When the code is understood, designations like 230M07 give far information about the type of steel it is describing:

000-199 are Plain Carbon Steels with some Manganese
200-240 are Free-cutting versions of 000-199
300-499 are Stainless
500-999 are Alloy steels other than stainless

Depending on the group, the first 3 digits are meaningful, 220 means 2% Sulphur

'A' means the steel has a specified chemical analysis
'H' means the steel has specified hardenability
'M' means the steel has specified mechanical properties.

The last two digits indicate the Carbon content.

So the code 230M07 gives a clear indication of the steel's composition and it's purpose - it has particular mechanical properties. In a basic workshop using only a few different basic steels, EN1A is simple and helpful because no-one has to think. But those meaningless EN numbers are unhelpful to anyone looking for steels to meet advanced requirements.

Not a good idea to have practical men using a different system to the designers and purchasers because it causes lost in translation mistakes. I think it's only tolerated in the UK...

Dave

SOD:- Well, it was a long time ago, but I can remember prising out bits that looked a lot like broken gear teeth from a bit of what'd been held in RM stores as EN3B, that I was trying to turn. Describing the proper processing doesn't mean it always gets followed.

Edited By Mick B1 on 20/01/2020 14:46:14

Lets face it companies pay for C of C's because their quality controls system demands it. that's leaves suppliers with dodgy metal nowhere to sell the known / suspect stuff to except model engineers who won't set the lawyers on you when they find out it's rubbish.

Thanks for replies/discussion.

Mick B1 - the Holmes quote was tongue-in-cheek of course. Being relatively inexperienced I have a tendency to blame my technique and/or inappropriate tooling when things go wrong - I really didn't expect the metal to be the problem, hence 'whatever remains, however improbable, must be the truth'. And so it turned out.

Although the immediate problem was solved by using another bar I tried cutting the thread on the other end of the original piece, about 500mm in from the factory end, to test Hopper's suggestion that it might be an 'end effect'. No better. So it's just a crap piece of steel, for this purpose anyway.

This has all been quite enlightening for me. In retrospect I should have smelled a rat when I skimmed the bar to take off rust/mill marks. It didn't go well, but I put it down to my inexperience - I hadn't tried turning such a relatively slender 500mm piece before. But it went fine with the other piece using the same procedure.

In future I shall make some test cuts to get an idea of how the metal behaves before beating myself up.

Both pieces came from a local family-run ME business, not random eBay stuff.

It seems that global production of 'crude steel' (whatever that means!) is about 1.8 billion tonnes annually, so perhaps not surprising that not every bit is top-notch.

Thanks again, Robin