Quite some years ago I bought what I thought was 3/16″ copper sheet, from a scrap-yard, intending to use it for my steam-wagon’s boiler… before deciding I would be better having it made.
It still had a label telling me what it is, and that is now illegible, but must have described it as “CZ121” as I wrote that on the metal with a felt-tip pen.
I was told this is “Gilding Metal” – a type of brass but with a very low zinc proportion, though still not advisable for boiler-making. It has oxidised to a dull copper colour, but is still a bit brighter and redder than pure copper turns. However, that grade made be either wrong or superseded.
I started using a piece of it today, by sawing, drilling (step-drill and twist-drill) and milling, and although the cut surface and chips have a slightly brassy appearance and it does not seem as “sticky” as copper, I am not sure now quite what it is!
Does anyone know a way of identifying this stuff as either commercially-pure copper or a low-Zn brass easily, please, bearing in mind I do not have a metallurgy laboratory at my disposal? Would for example, comparing heating & pickling colours of samples help?
I followed that by examining suppliers’ publicity and discovered that firstly the brass is not sold as sheet, only various forms of bar, and secondly it looks much more like brass than this stuff I have.
The grade for gilding-metal I found seems to be CW, as well, but I don’t know if this is a more modern designation. I probably bought the metal a good twenty years ago, and it had a stick-on label giving stock-control information including specification..
I have just paused to go and look, and yes, CZ121 is what I had written on the sheet, so that seems wrong. Possibly the label was already damaged. So perhaps I have identified it correctly as “gilding metal” but quoted the wrong specification from the label, which is now too faded to read at least in artificial light.
At least the CZ does show this is some form of brass.
My question about the reaction to pickling goes back to when I once sectioned a worn-out injector, polished the exposed surface with emery then lightly pickled it. I can’t remember with what acid, but the cones and body did emerge with slightly different hues.
So I could try with this material is cleaning, polishing and pickling in the same acid in different pots, an off-cut of each of this suspect metal, copper water-pipe, brass and perhaps, phosphor bronze and compare the results. It will not identify the real nature of the metal if it is an alloy, but may show if it is or is not pure copper.
So if you have read the label wrong how far out are you, copper could be C101, but if it is a golden colour rather than copper and bends easily likely to be guilding metal.
It’s more like copper in colour but seems very slightly lighter. I am not sure about bending “easily”. Two of the three plates are a bit bent, but at 3/16″ thick they not something you can warp in your hands.
I will study the original label in daylight to see if the specification is still at all visible, but I don’t hold out much hope.
Diogenes –
Boiler material was the original aim, when I was going to make my boiler. These plates were intended as material for the outer and inner firebox tops, the smokebox tube-plate, and for the bolting-rings that hold the thing together as indeed the prototype was built.
The piece I have finally cut from it, with a bit more work to do on it, will be a ring that sits under the foundation-ring of the commercially-made boiler, on top of the ash-pan, to act as a locating-plate for the cladding.
””’
I am trying the pickling test.
I had a small quantity of hydrochloric acid, and diluted this – by double decanting so I “Always Added Acid”. I think I have over-diluted it as a test drop on limestone hardly fizzed.
Anyway there are now four pots outside, under an upturned bucket, holding separately small off-cuts of the Mystery Metal, copper pipe, brass and leaded gun-metal.
I prepared the surfaces by glass-paper then washing.
Nothing much seemed happening due to the over-dilution so I beefed the fluid in the pots up a bit with limescale remover containing sulphamic acid.
I had a small quantity of hydrochloric acid, and diluted this – by double decanting so I “Always Added Acid”. I think I have over-diluted it as a test drop on limestone hardly fizzed.
Anyway there are now four pots outside, under an upturned bucket, holding separately small off-cuts of the Mystery Metal, copper pipe, brass and leaded gun-metal.
I prepared the surfaces by glass-paper then washing.
Nothing much seemed happening due to the over-dilution so I beefed the fluid in the pots up a bit with limescale remover containing sulphamic acid.
And there they will sit until tomorrow….
What fun! But a complete waste of time if an actual answer is needed!
I suggest starting by measuring the density of a sample, and comparing it with tables. This might get the answer directly, more likely it will suggest several possibilities.
Identifying the alloy by old-school chemical analysis is possible, but not without suitable experience, equipment and reagents. The most important item of equipment needed is a chemical balance. (A precision scale able to weigh down into the microgram region.)
If the unknown metal is believed to be a Brass (many varieties), then it will contain Copper and Zinc. Select a reagent that dissolves Zinc, but not Copper.
Weigh a test-tube
Add reagent to test-tube and weigh again, giving the weight of reagent
Weigh the alloy sample. To speed the reaction, may be useful to granulate it.
Add sample to reagent, seal the container and leave to react.
When reaction is over,
weigh a second test-tube, decant the liquid into it, and find the weight of the liquid. The increase of weight is due to Zinc, and the quantity can be found by doing the maths.
Weigh the first test-tube with sample, and calculate the sample’s weight. In a simple Brass containing only Copper and Zinc, the remains will be pure Copper. If the ratio is 90:10, then Gilding Metal, whereas 65:35 is common brass.
Might be enough except ‘Brass’, is a family of alloys that might contain Iron, Aluminium, Tin, Phosphorous, Lead, Arsenic, Nickel, and other metals. The chemist would repeat the react and weigh process as necessary to identify these. Skill is important. Although the presence of Nickel is often revealed by a light-green colouration, the weight of the liquid might hint that it contains more than Zinc. Analytic Chemistry is skilled tedious work, requiring strict attention to detail.
When I did Chemistry at school the method described above were rapidly being phased out. Spectroscopy is much faster, more accurate band requires zero skill. Unfortunately, the equipment is off-the-scale expensive. Best bet I think is to find a friendly scrappy and ask him to scan it!
I don’t think Nigel’s backyard experiments will tell him much more than the sheet probably contains Copper, which we already know…
I think it unwise to substitute unknown scrap when it’s necessary to use a particular metal in a project for safety or functional reasons. Gilding metal is often decorative, so no harm in taking a punt with that or many other low stress requirements. Can’t imagine a boiler inspector being happy to find ‘might be Gilding Metal’ in a pressure vessel!
With great respect gents, there’s a lot of over-complifying going on here.
I bought the metal a long time ago thinking it was copper but it had a label on it suggesting otherwise. It was for boiler material but advised it was a type of brass I decided not use it for that and subsequently had the boiler built commercially.
So I am not looking for boiler-making material.
I never claimed wanting specific analysis as I do not have the equipment and knowledge either for chemical tests or determining its density, nor access to any.
Really I was more concerned with finding a simple way to determining whether it is copper or the copper alloy I think it is – Gilding Metal, a brass with very low zinc contents. I asked in case anyone could suggest a method I could use other than pickling samples and comparing the results.
So I did, using very dilute Hydrochloric Acid with a squirt of Sulphamic Acid domestic limescale remover.
The results are in the photo.
Left to right: the Gilding(?) Metal, a piece of brass, leaded gun-metal and two bits of copper plumbing pipe. Of the last I cleaned one with sandpaper but left the other dull. I sandpapered all the others, then washed them all in Swarfega then soap and water.
I removed the samples this morning, washed them in boiled water (still hot) then photographed them.
Both faces of the Gilding Metal emerged bright, but non-abraded reverse showed a faint, fine chevron pattern – I could not have photographed it.
The Brass, curiously, was more pink above the acid than submerged.
The Gunmetal stayed bright with find grey streaks on its drawn surface.
The copper pipe – well I expected it to be either bright or some sort of pink colour, not the fifty-first shade of grey. That was not a loose coating as it resisted rubbing.
I left the samples on the kitchen windowsill all day. This evening, as I have in front of me they are:
Gilding Metal: dulled a bit to a golden coppery hue with a satin sheen.
Brass: The portion had had stayed dry looks like, well, clean brass. The pickled area has a rather patchy pink tint to it
Gun-metal: Dulled slightly to a golden shade on the machined and abraded faces. The drawn external surface is patchier, rather varied with some pinkish patches.
Copper: the dirty brownish-grey areas are developing a green verdigris.
Conclusion as far this basic test goes:
The stuff I have always thought “Gilding Metal”, probably is that even if I had transcribed the original label incorrectly.
Whatever its true composition, it is not pure (commercial grade) Copper!
.
The second photograph is what I have just made with it: the ring to support the bottom of the cladding on my boiler’s firebox. Outside Diameter 8.25 inches, diameter over cladding 8 inches.
The inner firebox protrudes below the foundation-ring and that will nest in the big central hole; with the ashpan roof up against the whole caboodle . Apart from five tapped holes for stay-rods for other parts of the cladding, all those other holes are simply to reduce weight and heat transfer from the foundation-ting sitting on it. I may interpose a layer of thick gasket material to reduce that further. It is not secured to the boiler itself.
Having gone to all that effort, I do wonder if I over-complified it, but my first attempt, a spun flanged ring in 1mm thick sheet steel, looks too flimsy. Unless I am worrying too much about that. It’s hardly going to take a heavy load.
Cladding this boiler is turning out far harder than I had imagined due to its peculiar shape!
Nest task is to make new chassis mounting brackets and retainers for the boiler, partly to tidy up the many alterations / unforeseen-problem work-rounds to the existing ones, but mainly so they can be removed with the clad boiler from the chassis; in which operation they will have a dual role as part of the Lifting-Equipment, Special-to-Purpose, this boiler needs.
The nearest I approached that in my school metal-working class was a large tea-caddy spoon, made from about 16swg gilding-metal. I still have it.
What is noticeable is that the photographed jug, and my tea-caddy spoon, are a “brassier” colour than this metal I have, with its slightly pink tinge.
I took that photograph above just after deburring the ring with files and glass-paper. I only removed raised burrs and rounded the edges, as it does not need a fine finish – it will be hidden. The condition were early evening daylight under heavy overcast, so I let the camera decide the exposure for itself, but it does show the coppery tint.
The dark patch is verdigris confined to a very small area of the stock plate, probably from standing on a slightly damp floor and perhaps in contact with some other metal.
So I suspect this must be closer to the 95/5 proportions than some other versions.
The Wikiedia description has suggested an idea to me – that of the alloy being used as a silversmithing training material. I have a friend who does make jewellery as a hobby so I might take him a sample and see if he’s come across anything like it.
With great respect gents, there’s a lot of over-complifying going on here.
No Nigel, things should be simple, but no simpler. May not fit with what you want, but there it is. The real world is complicated. Though it may have been an interesting diversion, what you’ve done with chemicals proves nothing.
…
I never claimed wanting specific analysis as I do not have the equipment and knowledge either for chemical tests or determining its density, nor access to any.
Really I was more concerned with finding a simple way to determining whether it is copper or the copper alloy I think it is – Gilding Metal, a brass with very low zinc contents. I asked in case anyone could suggest a method I could use other than pickling samples and comparing the results.
I suggested a simple way: measure the density!!! Much more trustworthy than splashing chemicals about. Or know for sure by getting a scrappy to scan it with a spectroscope.
So I did, using very dilute Hydrochloric Acid with a squirt of Sulphamic Acid domestic limescale remover.
What on earth is this mixture expected to do? Generations of analytic chemists are rotating in their graves! Years of painstaking research and know-how ditched! And the sloppy description mean the experiment cannot be repeated by anyone else, which is a fundamental requirement in science.
Left to right: the Gilding(?) Metal, a piece of brass, leaded gun-metal and two bits of copper plumbing pipe. Of the last I cleaned one with sandpaper but left the other dull. I sandpapered all the others, then washed them all in Swarfega then soap and water.
I removed the samples this morning, washed them in boiled water (still hot) then photographed them.
Both faces of the Gilding Metal emerged bright, but non-abraded reverse showed a faint, fine chevron pattern – I could not have photographed it.
The Brass, curiously, was more pink above the acid than submerged.
The Gunmetal stayed bright with find grey streaks on its drawn surface.
The copper pipe – well I expected it to be either bright or some sort of pink colour, not the fifty-first shade of grey. That was not a loose coating as it resisted rubbing.
I left the samples on the kitchen windowsill all day. This evening, as I have in front of me they are:
Gilding Metal: dulled a bit to a golden coppery hue with a satin sheen.
Brass: The portion had had stayed dry looks like, well, clean brass. The pickled area has a rather patchy pink tint to it
Gun-metal: Dulled slightly to a golden shade on the machined and abraded faces. The drawn external surface is patchier, rather varied with some pinkish patches.
Copper: the dirty brownish-grey areas are developing a green verdigris.
Nigel’s observations add little value because the underlying method is far too simple. It’s suggestive in places. Pink suggests a copper alloy, the effect being produced by acid dissolving the alloy metal (probably Zinc), leaving unoxidised Copper exposed. Pure Copper develops a protective Oxide layer, and unless Nigel’s concoction removes it, there will be no pinkness.
Conclusion as far this basic test goes:
The stuff I have always thought “Gilding Metal”, probably is that even if I had transcribed the original label incorrectly.
Whatever its true composition, it is not pure (commercial grade) Copper!
We already know it’s not pure Copper, and nothing Nigel has done confirms Gilding metal. Rather the opposite, because Gilding Metal is a Brass, and should show pink. The evidence does not support Nigel’s conclusion.
.
The second photograph is what I have just made with it:
Nothing wrong with what Nigel made from his mystery metal because it doesn’t matter what the material is. Nigel is back on solid ground when it comes to metal bashing rather than DIY chemistry!
By the way looks like Copper to me in the photo, with no sign of Gilding Metal’s distinctive yellow. That might be a photographic issue; rendering colour correctly is another horribly complicated problem!
Pretty sure there was an article on measuring density in MEW a few years ago. Much simpler to get a strong indication from density than a casually designed experiment with limescale remover.
The average pair of “dealer” scales should be suffuicient to get a good idea, mine will read to 0.01gram. As it is a comparrison there is no real need for callibration.
CZ121 brass 8.47g/cm3
CZ101 gilding brass 8.80g/cm3
C101 copper 8.92g/cm3
So should be able to spot it with a 1cm3 sample but easier with say 10cm3
If Nigel were to draw his ring in Alibre it would give it’s weight for a range of metals which he could compare to it’s actual weight. Which would save having to waste material cutting samples.
I can’t sensibly determine its composition either chemically or by density.
The latter looks simpler but I do not have a chemical balance. Since the samples are of random shapes the only possible method is by specific gravity (I think, trying to recall school science lessons) using displaced water, but I don’t have an Archimedes Can (right name?) either.
I did not “waste” material by cutting samples. I used existing off-cuts, and apart from the two scraps of pipe they are still useable. The bigger problem is whether they are too small for further machining.
…..
For the intended purpose, no, the material is not critical.
More to the point though, this was Plan B but after all those hours, a sizeable square of Mysterium plate and a few units of electricity making it, I think Plan A (the spun steel ring made previously) is the better option and even that reveals unforeseeable problems!
I might be able to draw the ring in Alibre Atom, but even if I found the Mass tool, I still cannot weigh the object sufficiently closely.
I did not draw it at all in any way, not even a rough sketch. Like many other components on this project I made it from a few measurements of existing parts and a bit of arithmetic, written on scrap paper. I did make a formal drawing, and in CAD, for the cover-plate on the boiler top lagging, because it has a lot of holes in odd places.
.
I thought cladding a Belpaire pattern, taper-barrel locomotive boiler would be hard but this thing, essentially a big T-piece of two cylinders, is proving a nightmare. Easy for the traction-engine builders. They lag only the barrel!
I’ve also to re-make the boiler mountings on the chassis, and that may mean replacing a column that holds the injector and its water-valve spindle, hopefully without also having to make new pipes. If I fit a non-prototypical superheater, assuming that is physically possible though the flues are there for it, I’ve to consider whether I can hide its flow and return pipes in the cladding as hoped or need external pipes to elbows in the smoke-box shell…. and so it goes on.
Three steps forwards, two back… two steps then three back to one made five years ago last Maundy Thursday…
Given the volume of the finished part a pair of kitchen scales would probably be sufficient to weigh it.
The difference between it being brass or copper is around 40grams so a big enough difference to see.
Better still measure the rest of the sheet and work out it’s volume and then weigh it, the bigger the sample the larger the difference in weight which will be easier to see.
Regarding wasting material that was if you were going to make up 1cm3 samples. Buy working out the volume of what you have that is not needed.
Looking at your figures above, I doubt I could weigh the machined ring accurately enough for that, i.e. to tenths of a gram… Not on my kitchen scales though. They are a simple beam balance whose smallest loose weight is 1/2 oz.
I could try the stock sheets though, using the workshop hoist, my spring balance and a cat’s-cradle of that Physics Laboratory essential, the “light inextensible string”.
This balance is the type intended for parcels, baggage and the biggest one that nearly got away, so quite coarse (for coarse fishing?), with a range of 22kg X 0.25kg and no obvious way to set the needle to 0.
The metal is pretty obviously not “ordinary” brass. If it’s gilding metal that 0.12g / cm^3 difference from copper may not be enough to show with the balances I have; but I can at least try it.
.
This does remind me I have long wondered how to weigh the wagon itself. Short of waiting until it’s complete then loading it into my car and taking it to a weighbridge – but any weighbridges anywhere near me are private, and since they are for laden HGVs anyway I doubt it would barely register.
So……
I have a Salter ‘Dynamometer”, it says on its engraved brass dial, divided to 10cwt X 2cwt and qrs. I.e. a half-ton spring-balance.
I would be wary of dangling the complete (if ever…) vehicle from the overhead hoist but I can use that with the dynamometer and the small balance above as appropriate, to weigh the part-made chassis and other assemblies separately as they appear, and add them up.
You missed the point of my post above, 0.1g may be hard to measure on a small sample 1cm3 but you have a ring that is far larger so the difference is larger and easier to measure with equipment that does not resolve to such small increments.
The finest I can weigh anything to, is 0.25kg, and rather uncertainly, on a simple domestic spring balance intended only for tasks like weighing a parcel for posting. It is even labelled as “Not legal for trade”.
The sheets of the metal are much bigger and heavier than the ring, and plain rectangles certainly easier to measure for volume than the ring, so might give results that mean something, though only approximately. I do not for example, know if the accuracy of that spring-balance is linear across its full range, and enough for a test like.
I took up the challenge, weighing and measuring the three stock plates and the ring; all 22’6kg of it.
Calculating the volume of metal in the ring was rather laborious, with all those ‘oles – five separate values to determine!
Anyway, the result I obtained was 8.53gm/cm^3.
Which seems rather light, against your figures. It doesn’t seem to match anything in Model Engineer’s Handbook, although that shows 60/40 brass is less dense still. The table does not list gilding metal.
So I probably had transcribed the original, now unreadable label already on it; but given the impossibility of weighing these plates very accurately, having to average widths across sawn sheets, and rounding errors I would not be surprised if that figure is plus or minus 0.2 wrong.
I took up the challenge, weighing and measuring the three stock plates and the ring; all 22’6kg of it.
Calculating the volume of metal in the ring was rather laborious, with all those ‘oles – five separate values to determine!
Anyway, the result I obtained was 8.53gm/cm^3.
…
So what exactly this stuff is, remains a mystery.
Well done! Though the answer isn’t confirmed yet, 8.53g per cubic centimetre is correct for 60:40 Brass.
Gilding Metal (10% Zinc) has a density of 8.75g per cubic centimetre, suggesting Nigel has one of the Brasses containing more rather than less Zinc. Not Cartridge Brass, Tombac, or Gilding Metal. A more accurate density measurement would reduce the uncertainty considerably.
To get closer, it’s necessary to refine the measurement:
Measure the volume by displacing water rather than doing sums on dimensions: the latter is error prone unless the shape is very simple.
Invest in, or make, a better set of scales. Making them isn’t that difficult, but if I didn’t already own one, I’d head for the kitchen section of my local supermarket.
A chemical or spectroscopic analysis is needed to nail what it is for certain – 8.53g per cubic centimetre is about right for Aluminium Bronze, Nichrome and a few other candidates.
A nearby auction house offers to use XRF (X-Ray Flourescence) on precious metal items consigned for sale to establish whether they are e.g. pure or plate. Would such a machine be any use on engineering materials? I guess a pawnbroker might also have similar tech.
I’d be surprised if a kitchen balance or bathroom scales from a supermarket would be of any finer resolution than my parcel-weigher, unless an electronic type perhaps limited in resolution only by the display’s significant-figures’ setting. At least there’d be no trying to estimate the pointer’s position between divisions.
It might be more difficult to measure the displaced water’s volume than the metal’s mass, unless I weigh the water. Errr, what do I use for the displacement can. Perhaps the steam-wagon’s own water-tank using the hand-pump outlet, just below the top, as the overflow!
.
Potentially the most accurate form of balance while also the easiest to make, is the steel-yard. Neglecting its own weight, when suspended, is the load on the suspension always twice the object mass, or the object mass plus reference mass?
I experimented with that Salter Dynamometer as spring-balance, up to 10cwt capacity, to weigh the wagon so far. It proved nearly 2cwt already. I estimate full working order weight will be 3-4cwt, and me perched on it will add another hundredweight and a bit. Scaled up to a ton and a half – at least I am well under maximum load!
…
Adrian –
Interesting idea. Whether such equipment used in a sale-room would be set to analyse copper alloys or just look for the precious metals, I have no idea.
My balance type kitchen scales read to 10g divisions that are big enough to split further which is better than your spring balance hence why I said it should be possible with a kitchen scale and they are not digital ones which will read to 1g.
Easiest way is a simple suspended bar with the ring hanging X distance from the pivot and then just use a known weight that can be moved along the other side of the bar until it balances and measure it’s distance from the pivot. Then just a bit of maths to work out the missing weight of the disc. I have used this to work out postage on heavy castings that were too much for the scales I have
I’d be surprised if a kitchen balance or bathroom scales from a supermarket would be of any finer resolution than my parcel-weigher, unless an electronic type perhaps limited in resolution only by the display’s significant-figures’ setting. At least there’d be no trying to estimate the pointer’s position between divisions.
It might be more difficult to measure the displaced water’s volume than the metal’s mass, unless I weigh the water. Errr, what do I use for the displacement can. Perhaps the steam-wagon’s own water-tank using the hand-pump outlet, just below the top, as the overflow!
.
Potentially the most accurate form of balance while also the easiest to make, is the steel-yard. Neglecting its own weight, when suspended, is the load on the suspension always twice the object mass, or the object mass plus reference mass?
I experimented with that Salter Dynamometer as spring-balance, up to 10cwt capacity, to weigh the wagon so far. It proved nearly 2cwt already. I estimate full working order weight will be 3-4cwt, and me perched on it will add another hundredweight and a bit. Scaled up to a ton and a half – at least I am well under maximum load!
…
Adrian –
Interesting idea. Whether such equipment used in a sale-room would be set to analyse copper alloys or just look for the precious metals, I have no idea.
I fear we’re all guilty of not reading earlier posts, or at least failing to take the information in! Same applies to the video I attached, which covers various ways of measuring displaced water that Nigel hasn’t thought of yet. He hasn’t had his Eureka moment!
I’ve committed at least two misreadings. I missed that Nigel was using a Parcel-weigher, which should be fine. More important, I forgot Nigel told us right at the beginning that:
It still had a label telling me what it is, and that is now illegible, but must have described it as “CZ121” as I wrote that on the metal with a felt-tip pen.
I was told this is “Gilding Metal” – a type of brass but with a very low zinc proportion, though still not advisable for boiler-making.
I think we’ve all been derailed by whatever clown told Nigel that CZ121 is Gilding Metal! It’s not: CZ121 is close to 60:40 Brass plus about 2% Lead to improve machinability.
The evidence that the plate is CZ121 Brass consists of:
being labelled CZ121
colour inconsistent with Gilding Metal, which is very yellow
Density of plate measured at 8.53g/cubic centimetre. CZ121 is 8.47g/cubic centimetre, close!