NDT method for surface cracks & voids 1

[prdave00]

I have a part produced from a CoNiCr alloy. I'm looking to identify a method to check for microscopic cracks and voids. I've looked at fluorescing dye penetrant testing to limited success (i.e., even on a surface where cracks were identifiable under a light microscope, the dye did not facilitate easy identification of cracks). The surface is about .06" wide. Any thoughts? Hopefully something that can be easily brought in house.

 

[Duwe6]

Try "Zyglo" post-emulisifiable stuff or any other high-sensitivity "Level 3", or higher sensitivity dye-penetrant. I've found things using a Level 3 penetrant that could not be seen at 400X magnification. Start with Level 3, if you don't have any decent results, go to a Level 4. But if you can see the cracks using a microscope, you are having a basic PT problem. Follow a qualified Procedure [one approved by a L/III]. Just "spray & rinse" doesn't really work. Dwell time and developing times are critical.

Also, proper cleaning is of paramount importance for tiny indications. Try vapor-degreasing or a caustic soap wash. Both have worked will on 'tough' parts.

 

[prdave00]

Thanks Duwe6. I'll check on post-emulisifiable Zyglo Liquid Fluorescent Penetrant product. Can you refer me to a commercial caustic soap solution or would a 5-10% sodium hydroxide solution work? I use Liquinox liquid detergent (an Alconox brand) for a lot of critical cleaning, but this is just to "gently" remove organic residues. I imagine a caustic soap wash would also be good to prepare a surfaces for microscopy work such as SEM imaging.

Have you any experience with Eddy current testing? Besides the tedious (and expensive) task of using an SEM to go over a large population of parts, and assuming I can't find a suitable dye penetrant test option, I was considering this as an option. I used it in the past, but I'll need to freshen up on it.

 

[Duwe6]

Trust me, properly performed dye-penn PT will find ALL defects open to the surface.

Easy-to-obtain caustic soap is dishwasher detergent. Electrasol, etc. Like anything else, wash HOT 110-140°F and use a low-mineral content rinse. Distilled/deionized water preferred. Keep an eye [and stopwatch] on the parts while drying, and when all visible water is gone, continue drying for at least double the time it toof for the water to 'flash' off.

Also, are you aware that you cannot follow red visible-dye with fluorescent? There is something in the red dye that 'quenches' the fluorescence. Is it possible that that is the reason you cannot get the cracks to show up during PT?

If this gives you good results [lots of 'rejects'], you might give me a star.

 

[metengr]

Stick with fluorescent penetrant (Zyglo) as mentioned above. This is the most sensitive NDT. Fluorescent penetrant testing should not be done in-house, you need a qualified procedure and qualified examiner.

 

[prdave00]

Duwe6: Thanks for the follow up. I think you deserve a star for pointing me in the right direction. Besides it looks like metengr concurs with your suggestion. You did make me feel a little silly about the caustic soap just being detergent, but I'll live. Now I know.

metengr: I was actually on the phone this morning with a supplier for Magnuflux who "reprimanded" me for even thinking about doing this in house. I think I'll follow your and his suggestion. The call actually worked to my benefit since he was able to refer me to a lab.