Identifying aluminum alloys

Hi,
I imagine this is a common question, if it's addressed in a FAQ or other reference material please point me to it. That said, I have a need to identify unknown aluminum alloys. I do not need to identify a specific alloy, only the major group.
For example, if I am given an aluminum casting to repair, it is essential to know which alloy group - Al-Si-Mg, Al-Si-Cu, Al-Zn, etc - was used, in order to select the correct process and filler material for welding. The same is true for wrought alloys.
Some alloys are not weldable by conventional (TIG, MIG) means, it might be asking a bit much to identify those specific alloys without sophisticated testing. In some cases, I have the luxury of obtaining a small amount of material to test. Common "unweldable" alloys such as 2024 and 7075 reveal themselves very quickly in practice.
I know that there is at least one testing kit available, it is sold on www.tinmantech.com. The proprietor apparently doesn't believe in giving anything away; where I am able to compare prices for identical items sold there vs. other sources, he charges a 60-100% premium. I imagine that the test kit, at $700, is no exception. Not that I begrudge him a profit, but there is the concept of "value" to consider...
I am not adverse to hitting the books, and I took the usual college chem classes for a sci major. Given time, I can wade through most anything. Now, with that background, here is my question: Can I assemble my own testing kit, and if so, how?
Thanks.
G.
p.s. Sorry for fake email, but I get enough spam already.

  Quoted Text Here  
http://www.modernmicroscopy.com/main.asp?article &print=true
See if this reference works. Let us know if it does.

I love to recommend the use of X-Ray Fluorescence testing, which is nondestructive and gives immediate answers, usually qualitative. For aluminum alloys it should be performed preferably in vacuum. The equipment is available with many foundries, metal centers, junkyards and laboratories. A test is usually inexpensive. An aluminum casting or extruding facility will probably operate a Spark Optical Emission Spectrograph and should be able to provide tests for occasional customers. I would personally refrain from employing spot testing kits, it is old technology and inadequate. Hardness testing and metallographic examination could also help, but they need expert personnel and suitable equipment.
http://www.welding-advisers.com
snipped-for-privacy@aol.com (Bob K 207) wrote in message
  Quoted Text Here  

Since you took some chem, see if a local college bookstore has a text for the course Qualitative Analysis, where simple tests are performed for a variety of metals--most often, the metal ion, which would mean first dissolving a sample of the aluminum. Nothing fancy like the spectroscopic methods mentioned, and not quantitative, but cheap and simple--standard "bench" or "wet" chemistry techniques. Some of this is probably on the web, as well. The real problem might be in procuring the various reagents, as Sigma and Aldrich et al don't generally sell to individuals. Didn't know 2024,7075 were unweldable. Check this info and your main query with sci.engr.joining.welding--there's a few real brite sparks over there. TIG can be finicky indeed, but I have used a flux-coated alum rod stick on, I think, DC reverse, which welds like a sunuvabitch--not pretty, but good, strong welds. Seemed to me these rods were not too particular, and would weld anything aluminum: oxides, no oxides, clean, greased, whatever. If the repair is to get someone thru the day, and not to persnickety specs, these might help. ---------------------------- Mr. P.V.'d formerly Droll Troll
  Quoted Text Here  
to
order
Now,

  Quoted Text Here  
Why do it yourself? If knowledge of the grade of aluminum casting is that important spend the $25 to ship (overnight) 10 g sample to a testing lab and get a full analysis for about $100 and get results by E-Mail, tag this onto the price you charge for weld repair - after all it is part of the job.
Ed