Measurement of Surface Roughness 2

[schradea]

We have a part that requires a smooth surface finish on a sharp radius (.005 / .010) of a brass part. This radius is used as a sealing surface, so it must be very smooth to make the part bubble tight.

Our quality department has taken issue with the specifying of a 16 finish for this surface, because they don't have an objective method for measuring it (i.e. profilometer). I've suggested that a visual inspection could be performed instead, but they're of the opinion that this is bad practice, because it doesn't give an objective result.

I have two questions:
1. Is there a standard for visual inspection of surface finishes? I think if I can say, "inspect visually per ____", that would likely be sufficient.
2. Is there a good tool for measuring the smoothness of small, nonflat surfaces?

Thanks in advance.

 

[KENAT]

Well, you can get samples of different surface finishes produced by different processes - that might be a start, but it's still not a very definitive test. Similarly you could get a 'gold standard' sample of your specific part that parts have to be better than but it's a bit weak.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[btrueblood]

"Our quality department has taken issue with the specifying of a 16 finish for this surface, because they don't have an objective method for measuring it (i.e. profilometer)."

Isn't that a bit backwards, specifying a design constraint based upon a lack of tooling or experience in the QC dept.? How do they objectively measure any other finishes?

Make them buy one. Or trust the supplier, if they have one. A profilometer and a lathe head to sweep the indicator along a circular lay path, done. May need to change the callout to specify a circular lay.

 

[TVP]

There are a number of different non-contact methods that capable of measuring surface roughness on small areas, areas with complicated geometry, etc. There are various names for the instruments such as optical profilers, white light interferometers, laser confocal microscopes, etc. Here are some links for additional information:

http://www.bruker-axs.com/white_light_interferometric_profilometers.html

http://www.zygo.com/?/met/profilers/

http://www.keyence.com/products/microscope/microscope/vk9700/vk9700_applications_1_1.php

 

[KENAT]

BTW, in case there's some confusion, bruker surface profilometers & optical profilers were formerly Veeco, Digitial Instruments, Wyko & Sloan.

For sure, not specifying something because Quality can't verify it is a bit back asswards. Specify based on function.

However, maybe they can sub contract independent inspection, if they don't want to trust the vendor/the vendor can't inspect it either.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[schradea]

Thank you all for your responses. I agree that our process is a bit backwards. We make a concerted attempt to ensure that our designs can be manufactured, inspected, etc. as we're designing, which I think is good in principle. Unfortunately, it sometimes results in the quality department thinking that it's our job to do their job for them.

We're looking into some of the tools specified, but haven't come to a good conclusion yet.