Crack testing of powdered metal parts

Ron Rompen

Trusted Information Resource
We currently have a customer failure (cracked parts) and our investigation has concluded that the parts in question were cracked during compacting. The cracking was horizontal and did NOT extend to the surface of the part until undergoing assembly at the customer site.

Our current test method is to take samples during the compacting process, sinter them, and then do microscopic analysis of cross-sections. As you can imagine, this is a slow process (3 hours is a QUICK turnaround for sintering and analysis) and has an inherent potential for error.

Does anyone know of an alternative method which could be used to examine compacted powdered metal parts for cracking? I have already reviewed ultrasonic and eddy current testing, and found that neither of them are suitable at this stage.

Thanks in advance.


Could you ring the parts like a bell? I did this for some carbon ceramics composites and had a frequency response which showed cracks and weighing to show voiding. Ring each part, microscopy on 1 per batch. Of course I am naive in this application.


Quite Involved in Discussions
Acoustic spectral analysis is used for 100 % in-line inspection on P/M parts by some manufacturers. I had to dig into this a while back when we had issues with P/M parts supplied by a vendor who did NOT have this capability and the parts had voids the size of peppercorns. The vendor blamed heat treat and shot peen for the defects. :cool:

So in essence, yes, it is like tapping it and "listening" for the characteristic overtones. The disadvantage is that you can only interpret the response if you have done a correlation between the output spectrum and the defects.

I'll see what documentation if any I have at work this coming week.
Top Bottom