Newbie: Field Alloy Analysis?

Hi all.
I Googled and FAQed but was unable to discover how to quickly, cheaply and accurately :0) determine the ratios and identities of each element in a given aluminum sample. Clearly there must have been a way to do this before the advent of radio-isotope and X-ray fluorescence analyzers.
Does one just whip out a pocket mass spectrometer / gas chromatograph?
When can I buy one at Radio Shack?
Thanks
--Winston

  Quoted Text Here  
There was, it was called qualitative chemical analysis. It required you to dissolve a sample of the material in a suitable reagent and run a series of chemical tests looking for precipitates or color changes in the solution to detect various elemental constituents. Tedious, time consuming, and you still don't know the ratios. That requires quantitative analysis, which is even more tedious and time consuming.
That's why hanging onto those tags from the manufacturer is so important. Saves you a hell of a lot of time and work to just read off the specifications.
Gary

Winston wrote:
  Quoted Text Here  
You have stumbled on the semi-universal truth.
In looking for "quick, cheap, and accurate" you will be saddened to know that you will have to choose two of the three, and whichever you do choose, will exclude the third as an option.
A quick, cheap method won't be accurate. A cheap, accurate method won't be quick. A quick, accurate method won't be cheap.
For what method are you sorting Al. Casting? If so, try to stick to objects that appear to be cast. Welding? Cast and extruded parts will probably do.
A couple vinegar packs from a fast food joint can help you to ID magnesium content. Scrape a bare patch on the object (expose fresh, clean metal) and apply a drop or two of vinegar. Magnesium will react with a purple tinged stain and bubble a bit. If a object seems too light for its size, or will not break when struck with a hammer, it may be Magnesium based.
Cheers Trevor Jones

Thanks, Gary and Trevor
I have *heard* of titration, but as you say, Gary it does not appear practical for me. My raw material will be almost completely recycled cast material of unknown provenance rather than carefully documented overruns.
> A quick, cheap method won't be accurate. > A cheap, accurate method won't be quick. > A quick, accurate method won't be cheap.
Sounds reasonable, unfortunately.
> For what method are you sorting Al. Casting?
Yes. I bought a 6 cyl. head from my local auto breaker that I hoped to melt. I see that others have improved their success by mixing 'fresh' metal 50/50 with recycled metal. I can do that but I would still like to determine and control the mix rather than guess. Your comment about buying cast parts as raw material makes intuitive sense. I will assume that cleaned cast parts should result in acceptable raw material. Also, thanks for the tip about magnesium sorting.
--Winston

  Quoted Text Here  
Winston,
This has been a challenge to me for YEARS - I have accumulated a pile of ingots og unknown composition over the years by melting down castings from various bits of machinery but I recently got one of therse off eBay:
http://www.syamkorea.com/english/products/metal%20spectroscope%20metascop.htm
Yet to fully get to grips with it but it shows GREAT promise.
Andrew Mawson Bromley, Kent, UK

(Snip) Andrew Mawson wrote:
  Quoted Text Here  
Wow! Please keep us posted on how this unit works for you.
--Winston

  Quoted Text Here  
of
from
http://www.syamkorea.com/english/products/metal%20spectroscope%20metascop.htm
Winston,
The theory of the instrument is quite simple - there is a buzzer type spark generator that arcs on the sample, and light from the arc passes into the instrument though a window and falls on a defraction grating. This splits the light up into a line spectrum, and the twiddle knob on the side lets you move each line under a graticule so you can measure its wavelength. It is calibrated in abritary units and you need to make a graph of knob readings against know items ( such as the double lines of the sodium 'd' spectrum). I haven't yet made up a graph as other things keep getting to the top of the pile.
In my study I have two LARGE spectrophotometers that I had intended to use in a similar way - redundant now !!!
Andrew Mawson Bromley, Kent, UK

(Snip)
Andrew Mawson wrote:
  Quoted Text Here  
Andrew,
Is there a photocell to sense intensity of reflected light at the graticule centerline? If so, I can see how a meter connected to that cell could be calibrated in any convenient measure of relative intensity. One could vary the angle of the diffraction grating to place any frequency of light on the centerline. A vernier drive to control this angle could be calibrated in wavelength (or absorption frequency?).
Is this how that tool works?
I wonder if some smart young fellow could place a linear color CCD in the path of the diffracted spectrum and connect that up to a computer so that it can do all the hard work. Just for fun, one might sacrifice an old flatbed scanner to output a monochrome version of the spectrum to print on to a grid prepared with frequency graduations.... Hmmm.
Thanks
--Winston

  Quoted Text Here  
http://www.syamkorea.com/english/products/metal%20spectroscope%20metascop.htm
spark
the
splits
you
is
readings
spectrum). I
  Quoted Text Here  
the
use
path
can do
  Quoted Text Here  
scanner
prepared
That's what I'd intended to do with the photspectometers I have - just digitise the output and graph it against wavelength.
The portable instrument relies on the observer classifying the brightness of lines relative to each other and thus is less accurate.
Andrew

Andrew Mawson wrote:
  Quoted Text Here  
(Snip)
I see. Thanks, Andrew.
--Winston

Winston wrote:
  Quoted Text Here  
Sure seems like it'd work. In my first programming job long ago, the company took gamma ray spectra of nuclear powerplant workers & the like, and matched the observed spectrum against known radionuclides. It seemed to work pretty well. If you knew what to look for, you could generally pick out (by eye) people who lived in brick houses by the thorium signature in their spectrum, left there by the radon in their houses.
You wouldn't need a color CCD - the diffraction grating does the color separation.
- ken

(Snip discussion of using CCD to analyze spectrum of energy diffracted by a grating in an effort to analyze alloys quickly cheaply and accurately.)
Ken Rose wrote:
  Quoted Text Here  
Cool! :0)
  Quoted Text Here  
Good point, Ken. Ironically, color scanners are more available than monochrome units. EBAY shows more than 40 likely candidates each under US$5.00. I wonder if there would be any advantage to the monochrome CCDs extended sensitivity into the near IR?
--Winston