16
$\begingroup$

I have this small gear which has sat on my desk for a couple years. I can't remember where it originally came from or what material it's made out of, so I'm hoping you can help me solve this mystery.

the gear in question

other face, showing galling

I initially thought this was made of stainless steel owing to the galling on one face and the I.D. as well as the colour, heavy weight and lack of major corrosion. However, on closer inspection it doesn't seem to be stainless steel - the dents and scratches indicate it's too soft, and there was some pale greenish-blue corrosion in the valleys of the gear teeth (since cleaned off). That (and the silvery colour) makes me think it might contain copper or nickel.

I'm not interested the relative proportion of metals in the alloy, only the general class of material - i.e. whether common elements are present or not in the object.

So far I've performed the following steps:

  • Visual analysis: matt grey, silvery where cut, free of major corrosion pits, major galling to some working surfaces. The teeth don't seem to be worn at all despite the galling.
  • Usage analysis: it's a gear, so I would expect an engineering alloy with good mechanical properties.
  • Magnetic test: Held ~1 mm away from a strong magnet, weak attraction was observed. An attraction force of ~3 g could be measured. A similarly sized block of iron would be attracted by at least a kg to the same magnet.
  • Density measurement: The weight is 144.70 g and measuring volume by the gravimetric water displacement method produces 16.03 ml. The resulting density is 9.03 g/cm^3

My overall conclusion from the above is that I still don't know what material this is made of. The density is much too high to be a steel alloy, but also higher than that of nickel and cupronickel alloys! The magnetic test is very weak but indicates it could contain nickel or iron (or cobalt!), albeit possibly just as an impurity.

I have common household (and some basic metalworking) tools and chemicals. I also have electronics equipment and tools that I could use to analyse the gear given a procedure.

I'm willing to damage the surface of the gear (e.g. sanding/filing/heating) but I want to keep the thing generally intact (i.e. no sawing in two or dissolving the whole thing in acid!)

What other tests could I perform at home to identify this material?

$\endgroup$
8
  • 2
    $\begingroup$ The trouble is that plenty of different common metals have a density around 9.056 g/mL. Example : Cobalt 8.90, Copper 8.92, Nickel 8.9. I have not mentioned rare earths, which will hardly be used as gears (Erbium 9.16). It may be an alloy. The usual way of analyzing it would be to dissolve partially in nitric acid and make a chemical analysis of the solution. But as you want to keep the gear intact, the analysis could not be done by chemists. The present Stack Exchange is no use for this sort of advice. $\endgroup$
    – Maurice
    Apr 22 at 18:20
  • $\begingroup$ Of course, I expect it is an alloy, functional parts aren't generally elemental. As you rightly say, dissolving in nitric acid isn't practical (for one, I don't have any nitric acid!). However, there are surely alternative chemical tecniques to rule in or out common elements/alloys. While I can't dissolve the whole gear, I could collect filings and dissolve/react with a more common chemical I do have (e.g. citric acid, hydrochloric acid, sodium hydroxide). I'm sure there are other chemical techniques I don't know about - perhaps some electrochemical test? $\endgroup$ Apr 22 at 18:34
  • $\begingroup$ Basically what I'm interested in is if there's an alternative to the standard chemical analysis (or zapping it with an XRF gun) that would be practical to do at home. I don't need to assay the thing, just to tell me what class of material it is. $\endgroup$ Apr 22 at 18:38
  • $\begingroup$ My guess would be some pot metal or another (en.m.wikipedia.org/wiki/Pot_metal). Which is pretty unspecific. So X-ray fluorescence or Rutherford Backscattering, neither of which is diy… $\endgroup$
    – Jon Custer
    Apr 22 at 18:54
  • 2
    $\begingroup$ And the third vector is to ask in Mechanics.SE This is a "straight cut" gear so is from a simple low-speed geartrain. One of the old-timers there might recognise it directly, and then work backward to a material. $\endgroup$
    – Criggie
    Apr 23 at 2:52

3 Answers 3

21
$\begingroup$

Xray Fluorescence

This may sound like some massive "Oh sure, I'll just hop over to the nearest government lab and ask for some beamtime!" overkill suggestion, but in fact handheld XRF devices are increasingly common at scrap metal operations nowadays - and most of these modern devices are able to determine relative metal concentrations to the point where they can distinguish between common alloys programed into the device's onboard library (which is exactly why scrap metal places want/need them).

This might not be particularly DIY, but if there is a metal recycling/scrapping place nearby you should be able to get someone to scan your gear for you.

$\endgroup$
4
  • $\begingroup$ I've definitely heard good reports of this for steel specimens, although I don't know to what extent it relies on having a "library" of known alloys. youtube.com/watch?v=cKR_LMF3XAY (starting around 7:45) and youtube.com/watch?v=KdfHVcU8U7U are relevant. $\endgroup$ Apr 23 at 6:08
  • $\begingroup$ @MarkMorganLloyd The devices determine relative elemental compositions (based on a reference library/database of individual elements' XRF spectra); additional composition libraries/databases can then allow the tool to say "This composition looks like ABC-123". For example, in your second video link at the 0:23 mark, on the screen the relative elemental compositions are shown, but down near the bottom it lists "Grade: SS304 Reference: SS/ 303-304" which is a specific industry-standard variety of stainless steel. $\endgroup$
    – DotCounter
    Apr 23 at 17:41
  • $\begingroup$ @MarkMorganLloyd Having a database of alloys is only needed to get responses like ‘This is probably SAE 4140 steel’, ‘This is probably C360 brass’, or ‘This is probably 6061 aluminum’. You can still get the individual elemental components without it, but then you need to know that, for example, a 61.5/35.5/3.0 ratio of Cu, Zn, and Pb with the rest being iron and traces of other elements is C360 brass. Some people can memorize the (absurdly long) lists of all the alloys they may encounter, but most cannot, so having the device give you an ID on the alloy is very useful. $\endgroup$ Apr 24 at 4:09
  • $\begingroup$ And the best part of these devices is that they are designed to be used on solid/bulk samples, rocks, soil, whole pieces of metal, etc. In other words, where dissolving the sample in nitric acid is undesirable for any number of reasons. $\endgroup$
    – Ben Norris
    Apr 24 at 9:45
14
$\begingroup$

Your idea of checking density is good, but consider the accuracy of your results. For a small part, errors in measurement due to meniscus of water in a cylinder, or sticking to the container or gear, can easily throw it off by a few percent. You might treat the results as more like 9.0, ±1 or 2%. In that case, here are some alloys and their densities, abstracted and abbreviated from the list at Machine Mfg, in that range:

Tungsten High-Speed Steel   8.7
Beryllium Bronze            8.8
Cadmium Bronze              8.9
Chromium Bronze             8.9
Manganese Bronze            8.8
Cr Ni Mn Stainless Steel    8.5
Copper                      8.9
Tin Brass                   8.8
Tin Bronze                  8.8
Nickel                      8.9
Nickel Chromium Alloy       8.7
Cupronickel                 9.0

Since it was only weakly attracted by a magnet, tungsten steel and pure nickel are ruled out, since they are strongly ferromagnetic (old Canadian 5¢ pieces were fairly pure nickel, and readily jump to a magnet).

Since it was somewhat attracted, bronzes and brasses seem unlikely, and most of those alloys have a yellowish cast (no pun intended), unlike the gear in your photos.

The remaining likely candidates, are one of the numerous stainless steels, some of which are only weakly magnetic, and a cupronickel alloy. Cupronickel is salt-water resistant and is used for gears.

Another nondestructive test would be for hardness, such as the Brinell, Knoop, Leeb, Rockwell and Vickers tests. You could perform one of those tests, and compare the value against those in the literature, such as at Machine Mfg and Nickel Institute.

You could also test chemically for nickel, copper and other elements, but state you don't find using dangerous or expensive reagents practical. You could dissolve a bit of metal electrolytically in distilled vinegar, by mounting the gear as the anode, and compare the flame tests and spectra with an inexpensive spectroscope, such as one of these, or at a high school lab. Of course, that would give just the merest idea of a qualitative analysis. (Why vinegar and not salt water? To reduce the intense yellow of sodium somewhat, though there's always some.)

My SWAG would be a cupronickel alloy -- please let me know if you do find out the material.


Thanks to Machine Mfg for useful online references!

$\endgroup$
8
  • 1
    $\begingroup$ I am actually quite confident on the density value. I used the gravimetric displacement method with accurate (and calibrated) scales, not a method that relies on judging a meniscus. I have however re-checked and realised that I did forget to take into account the effect of temperature on the density of water, so I have remeasured and the correct value for the density should be 9.03 g/cm^3. I'm sure there's still error from the true value, but I'd put it at a lot lower than 2%. $\endgroup$ Apr 22 at 21:30
  • 1
    $\begingroup$ Your suggestion to test the hardness is a good one though. A proper test against any of those standards isn't feasible (unless I buy a hardness testing machine), but I have tested it's ability to scratch and be scratched by some known materials. It could easily scratch a UK 50p coin enough to catch a fingernail (cupronickel) but was also scratched by the coin (to a lesser extent, only enough to feel with a fingernail - more like burnishing). It could easily scratch copper, and soft aluminium, and plain mild steel, but not stainless spring steel $\endgroup$ Apr 22 at 21:37
  • $\begingroup$ Could it be nickel-plated? Nickel-plated brass is a fairly common material in my field (electrical engineering) and I would call it weakly attracted to a magnet--the plating is thick enough for some magnetic activity. It would also form that pale greenish corrosion that was mentioned--both nickel and copper compounds tend to be greenish, and nickel in particular forms pale greenish compounds. $\endgroup$
    – Hearth
    Apr 23 at 16:07
  • $\begingroup$ If it is nickel plated, it's a very deep plating - when I filed away the edge of one tooth the material underneath was the same colour as the surface, just a bit shinier. $\endgroup$ Apr 23 at 21:49
  • 1
    $\begingroup$ @hifkanotiks bubbles stuck in places you can't see can still throw things off. But I think you've got 2 metals here. Look at the 2nd picture. There's an insert with a groove on the top face; the insert extends about half way down. The colour and pattern of scratches is different, and you may find a slight difference in the diameter of the bore. Depending on how it went together, the insert may be steel and the main toothed part another alloy, or vice versa (steel isn't a good material for plain bearings but it is for ball bearings). $\endgroup$
    – Chris H
    Apr 24 at 10:04
5
$\begingroup$

Thanks to @DrMoishe Pippik for the suggestion to perform hardness testing on the gear. I don't have easy access to a hardness testing machine, but tested against some household objects of known composition.

  • Stainless spring steel - much harder than the mystery gear, could cut a shaving off it. Mystery gear would not leave a mark on the steel.
  • Regular mild steel - mystery gear could leave a deep scratch, but was significantly abraded in doing so. It seems to have slightly less hardness than the steel.
  • Cupronickel 75/25% 50p coin - the gear scratched the coin enough to catch a fingernail, but the coin (even sharpened) would only lightly scratch (enough to feel but not catch a fingernail)/burnish the gear. It definitely has a greater hardness than the coin.
  • Hard copper (bronze) 2p coin - the gear scratched the coin enough to catch a fingernail, the coin would only burnish the gear with no mark that could be felt.
  • Soft aluminium - the gear would scratch the aluminium but the aluminium would not mark the gear.

I also compared the electrode potential of the gear to some other known references by half-immersing the two samples in a basin of 2% salt solution and measuring the voltage across the cell with a multimeter.

I tested with the following known materials:

  • Aluminium (foil): 540 mV potential to the gear.
  • Copper wire: 140 mV potential to the gear.
  • Nickel: 15 mV potential (this reading was unstable for some reason so I suppose should be ignored)
  • Cupro-nickel 75/25% coin - drifted slowly between +-5 mV.

From these two tests, I think I'm fairly confident that this is a cupronickel alloy gear. However, the density measurement (although close) still doesn't match up, and the hardness is greater than my other cupronickel sample.

My suspicion is that there is some other alloying element that both increases hardness and density over a "normal" cupronickel alloy.

$\endgroup$
1
  • 3
    $\begingroup$ You could try measuring the density of your cupronickel coin using same method, to make sure the density measurement is accurate. The C71500 alloy is at 8.94 g/cm³ which gets within 1% of your measurement. It is also slightly harrder than the C71300 coinage alloy. $\endgroup$
    – jpa
    Apr 24 at 18:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.