Needing more than a spark test?

graham-xrf

H-M Supporter - Gold Member
H-M Supporter Gold Member
Joined
May 27, 2016
Messages
3,449
I have some chunks of nice steel, 31mm thick and 120mm x 60mm (about 1.2" thick and 4.7" x 2.3").
The 31mm apart faces are parallel to within a tenth thou, so they are sometimes handy as a sort of fat parallel, but the perimeters are not as in a rectangular block. They have the profile of whatever hydraulic pump kit they were originally destined for.

They don't rust - so not just some nearly tool steel.
They are magnetic, so mostly iron in there.

It seems there is not much between the "spark test" on a grinder, and some expensive kind of analysis.
I can't afford one of those FLIR things which are only cost-effective if they are identifying exotic stuff all all day long.
Trying for density once let me identify a nice chunk of titanium, but I did have some other clues.

Is there some lower cost way of figuring out what is in the alloy?
 
A good question. As far as I know, no there isn't. The problem with identifying steels is that the alloying elements are usually not found in significant percentages and that many are common to an multitude of alloys so a simple qualitative test isn't sufficient.

My very first chemistry course lab in collage was qualitative analysis. Back then, this was usually done by dissolving a sample in acid and adding various reagents to create a signature reaction. My recollection is that the transition elements of which the major steel alloying metals are members were particularly difficult to identify.

One test that might be used is a flame test. It consisted of dipping a small platinum or nichrome loop into a dissolved sample and the placing the loop in a flame from a Bunsen burner. Different elements will color the flame differently. Unfortunately, it doesn't work with all elements.

Once an element was identified, then you have to analyze it for the particular amount in the alloy. My first professional job was as an analytical chemist designing analytical procedures for various materials. Many of the elements will interfere with each other, making an analysis difficult. A method known as polarography was used to perform some of these analyses. Another tool that we used was emission spectroscopy. that provided a relatively quick means of identifying materials. It consisted of placing a dissolved sample in a carbon cup and firing high voltage through the sample. The resultant light was captured and its spectrum recorded on a photographic plate. Unfortunately, the cost of an emission spectrograph is beyond the reach of any except large laboratories.

All this testing provided job security for chemists in the past. Which is why x ray fluorescence is the tool of choice today.
 
Here's a description of a "DIY" XRF setup: DIY XRF. It uses americium capsules from smoke detectors as the X-ray source. But you'd still need to get your hands on a PMT and scintillator crystal, and be willing to spend a lot of time messing with it (getting it to the point of making quantitative measurements would be a heavy lift for a DIYer). Make sure you have some lead-lined shorts :D.
 
I have only 0.9 mirocuries of Americuim 421 here on my desk, and possibly 2 microcuries down in the garage.
I have 2 photomultiplier tubes with the green flourescent back screens laying about in the cellar. Three if you count the cheap Russian night-sight. Honestly - I think there may be easier ways of making a pulse of X-Ray.

You are right about it being a DIY heavy lift attempt. Pity one needs X-Rays.

I am OK to dissolve a bit of this stuff, and try some flame test. The only platinum wire I have is a bit thin, it being from platinum-rhodium thermocouple, but it may do. I have sulphuric (OK - sulfuric) acid. It kind of pre-supposes all the alloy metals in there can end up as some sort of sulfate, no matter the valency.

Maybe if I can at least discover some of the alloy metals in there, then try a thin piece for melting, and see what happens if I heat-treat a bit to see if it self-hardens or something. It may narrow down the possibles. Given that it does not rust, there might be nickel, chromium, molybdenum maybe.
 
@RJSakowski : The flame test looks somewhat limited when we are after the steel alloys.
Is it realistic to conduct a chemical test for these, just assuming they are there, even if they prove not to be?

A first test for iron can yield a percentage, which then immediately reveals the remainder proportion being all the alloying elements?

Would the Curie temperature of the mix be a tell-tale as to which steel it might be?
If it corresponds to a known steel mix, the thing is revealed. I have visions of heating a lump of it, with a thermocouple stuck into a drilled hole, until a magnet lets go. Hmm.. we need the magnet material coupled in a way that the magnet does not die first!

In any case, any one of the XRF hand-held fancy kits is likely to cost a whole lot more than my entire machine, and all it's bits!
 
I don't think that the Curie temperature will tell you that. This is more of a decalescence thing. You have to have a good eye and a precise temperature reading to do this. One could possibly do it with a muffle furnace and an accurate pyrometer.
 
@RJSakowski : The flame test looks somewhat limited when we are after the steel alloys.
Is it realistic to conduct a chemical test for these, just assuming they are there, even if they prove not to be?

A first test for iron can yield a percentage, which then immediately reveals the remainder proportion being all the alloying elements?

Would the Curie temperature of the mix be a tell-tale as to which steel it might be?
If it corresponds to a known steel mix, the thing is revealed. I have visions of heating a lump of it, with a thermocouple stuck into a drilled hole, until a magnet lets go. Hmm.. we need the magnet material coupled in a way that the magnet does not die first!

In any case, any one of the XRF hand-held fancy kits is likely to cost a whole lot more than my entire machine, and all it's bits!
Graham, Last things first, Yes, a quick check on eBay shows xrf testers in the $6K range. Well beyond reason for anyone not using one for major business.operations. I think a business opportunity could be possible for someone doing testing for hire on mail-in samples.

I'm not sure that you would be able to distinguish different steel alloys by their Curie temperature. If it were possible, a more accurate test than simply looking for the temperature of loss of ferromagnetism. The is a tool called a differential scanning calorimeter which I understand can be used for measuring the Curie point. I haven't used it myself but about twenty years ago, one of our epoxy suppliers used into determine the extent of curing in an epoxy. It looks like the eBay prices are somewhat better than xrf but it really wouldn't tell you much about the composition of the steel.

A colorimetric test could be developed for measuring the percentage of iron. Again, I'm not sure it would be precise enough to identify an alloy. When looking at the composition of steel alloys, all of the components have a range with iron usually listed as balance. That said, iron forms some highly colored complexes the intensity of which can be measured with a colorimeter or by comparison to a color chart. To do so, a sample of known weight has to be dissolved and and diluted to a known concentration. Then the sample is compared to a color chart or, for more accurate values, the absorbance of a certain wavelength of light by the sample is measured. That's the simple version. In reality, checks for possible interferences need to be done, standards of known concentration have to be made and host of other possible issues have to be dealt with.

There are other means of measuring the amount of iron and other metals in an alloy. I had developed a titration procedure for measuring manganese which could also be used to measure iron concentration. Potentially metals like cobalt and vanadium could be measured in that way as well. But these are better suited to an analytical laboratory than a machine shop.
 
No joy, folks :(

Way back when I was a chemist, I learned about an instrument called an Atomic Absorption spectrometer. It was used for quantitative analysis of dissolved metals. For any given metal, the light source was a specific "hollow cathode" arc lamp containing that metal. Then there was a gas fueled burner, which had a long line of small holes - maybe 3" or 4" - through which the light was sent. The dissolved sample was slurped/atomized into the gas stream to "color" the flame. The length of the burner was necessary for sensitivity.

Light passing through the line of flames was measured with "some kind of" photocell detector. The amount of light passing through the flames would be reduced by whatever target metal was present in the flames, because the atoms would absorb some of the light from the source. The difference in readings between no-sample and sample, times a suitable absorption factor, times the concentration of sample dissolved in the solution under test, would give the concentration of the target metal in the solution.

It was a very persnickerty and involved test method, to say the least! And not that many metals could even be tested.


As far as a DIY "eyeball" flame color test, you're more likely than not to have a small quantity of sodium in the solution. Sodium gives a bright yellow light, which can overwhelm and mask the color(s) that might be there from other metals.
 
Wow! Between @RJSakowski and @hman , I seem to have unearthed the experts who are steeped in a combination of chemistry, physics, and element analysis at atomic level. I have images of periodic table pinned up next to the Starrett Tap Sizes chart!

OK - I guess since the steel was free, I can use it on whatever it seems handy for. You are right that the costs of knowing are unjustified.

As it happens, I do have enough alpha emitter material, to get up about 8 uCuries, and I already posses some photomultipliers, but they unfortunately only the type used for night vision cameras. The rest of the stuff I don't have. HM folk, especially those deep into CAD, CAM, and CNC kit, would probably find this well within their capabilities.

For those who might want to try..
You need to raid some smoke detectors. 8 or 10 of them, old or new.

The rest of the assemblies you can vary to your taste. You might want to turn up a mounting, or adapt some metal tubing.

I see one can get various kinds of scintillation crystal, and the Cesium Iodide type is $51.50 + $14.50 postage to UK
--> CsI crystal eBay

A new Hamamatsu Photomultiplier is about $90, but then, there is the mountain to climb to build the rest of the instrument, physical mountings etc. The information for the photomultiplier electronics is all there for some little printed circuit boards You can use other photomultipliers.

--> Hamamatsu Photomultiplier

--> PmTAdapter This appears to be a little opto-isolated highvoltage, low current supply for the photomultiplier.

and the little PIC microcontroller board --> ThereminoMCA

The software appears to be free.
The data for XRF measurements for almost every element is attached.
copperSpectrum.png

Thanks guys, for your response.
 

Attachments

  • XRFSpec_ENG.pdf
    3.9 MB · Views: 13
Last edited:
It would be an interesting project.

For an x-ray source, you need a gamma emitter. Apparently Americium 241 is the emitter of choice in ionizing smoke detectors and it is also used for XRF sources. It emits an 59.5 Kev gamma ray along with the alpha radiation.

There is some danger associated with amassing quantities of Americium 241 so caution should be exercised. If it were me, I would store the sources in a lead container and make sure that the XRF sources were well collimated in use.

This is all stretching my knowledge bank. I last dealt with anything like this over fifty years ago. What is the technology used in an x-ray spectrometer? Bruker, one of the leading manufacturers of XRF equipment, has a manufacturing operation in Madison, WI and a former colleague used to work there. I may have to have a talk with him.
 
Back
Top