Why do you need to test it after you’ve already tested the starting material? Just wondering. Is it acceptable to test a random sampling of the final products, which will be consumed by the QC process, to serve as a representation of the quality of the whole batch? Or are the individual parts too expensive to waste?

@tedd A company that I use to work for used a XRF hand held, for analysis of scrap and remelt and it paid for itself in two months.
@cockswain The requirement might be one hundred percent of product show proper content. Not just a sample.
If your customer requires the analysis of final product ( new requirement ) you maybe able to renegotiate the cost of the product.
Have a talk with sales and get the customer to ( indirectly ) pay for the equipment.

@cockswain Honestly I have no idea why we have to test them again. ONE of our consumers came to us cuz another supplier of theirs screwed up and they want us to ensure them we won’t make the same mistake. Personally I think the testing we do now is just fine, but we’re humoring them and trying to find other ways. The individual parts cost upwards of 50,000 dollars, so random sampling would be ill advised. Plus it sounds like the company wants to know EACH piece is good.

@Tropical_Willie Yah handheld XRF was immediately the first thing that came to mind. But my boss is asking us to try and think of other methods. Conductivity or density are the only two I’ve come up with thus far.

@tedd Is it a porosity issue too? Or just purity of Aluminum?

How about looking at the emissivity of the piece. You can set the part to a known temperature and use an imager in the 2um to 8um range. I’m not sure how sensitive it will be on small changes in alloy concentrations. Call FLIR and ask. They know their stuff.

@Tropical_Willie Purity of Al and the different metals that make up the alloy. We ensure Al up to 99.9999%, and our 1% throw ins of Si or Cu or whatever are usually guaranteed to similar levels. XRF has always worked for us in the past, so I dunno why we’re looking for more now.

@worriedguy Looking into that further now.

It looks like we’re probably going to combine conductivity and xrf scanning. But I’m doing a bit more research on emissivity to see if it would help us.

@worriedguy what more can you tell me about emissivity? How small a scale can we go on the numbers? um?

@tedd Typically we set the emissivity to match the material we are studying. I do thermal imaging so my wavelengths of interest are different from yours. I heard from one of the FLIR engineers they were using the 2–8 um range to detect differences in materials so they could image something that would not ordinarily be imaged. You can look at a perfectly smooth surfaces of steel, aluminum, lead for example and see they are different by the difference in black body radiation they emit. Their emissivities are significantly different. You might get lucky and find a wavelength combination that works for your alloy.

Oooo! How thick is the material? Do you have at least 4 cm to work with? Passively, of course. You might try some kind of pulse echo or SAW device to see if the speed of sound in the material is correct. We have the capability of measuring down to 0.2 ns resolution so we can measure slight concentration differences in air and liquids. The method would work for solids too.