Metal Cutting Tools & Verification of PVD & CVD Coatings

[Douglas E. Purdy]

I am new to the Metal Cutting Tools Industry (i.e. drills, endmills, reamers, inserts, etc.) and I have been given, or have found names to tests for verifying coating adhesion (i.e. Ball Crater Test, Daimler-Benz Adhesion Test) but I have not been able to find written test procedures (Industry Standard or Non-). Due to the apparent limited number of providers of this coating, I have not approach them yet. I am only looking into what a manufacturer can do to verify adhesion of the coating. Any help in this matter would be greatly appreciated.

Thanks,
Doug

[tomvehoski]

Doug,

I used to be the Quality Manager for Multi-Arc (now Ionbond I think) in Troy, MI in 1997. We did both PVD and CVD coatings. If I remember the name right, it was the Daimler-Benz test we used on a daily basis.

Basically we did a Rockwell C indent on the coated part and observed the resulting indentation under the microscope. Good adhesion would show very little loss of coating around the indent. Poor adheasion would cause the coating to spall away from the area around the indent. We had a diagram that showed acceptable/unacceptable amounts of spalling, but it was still very subjective.

Our corporate lab in NJ also had the ability to do a "scratch test" for adhesion. I don't remember the details, but it was more detailed than the Rockwell C indent. I believe a stylus was dragged across the part with a known force - somehow it could measure the point at which the coating was ripped off.

We used a ball crater test for checking the thickness of the coating. Basically you rotated a carbide ball bearing against the part with a diamond paste. It would cut a ball crater into the part. Something like this (side cross section view - if it works here):

R1
________________......................... ___________ Coating surface
------------------\......................../---------------
------------------ \....................../-------Coating--------
--------------------\____________/ Substrate surface ----
| R2 |
| Substrate |

By knowing the diameter of the ball bearing and the radius of R1 and R2 circles, you can figure the thickness geometrically. We had a camera on the microscope that snapped a digital picture. All we had to do was outline the two circles you see when looking at the top view and the computer did the math. 2-4 microns was our standard thickness for PVD TiN. We also had an X-Ray method for thickness checks, but it was calibrated off of ball crater samples. The substrate could also throw off readings. It was much faster and non-destructive, but we had to be sure the parts were the same chemistry every time if we were to use it.

I understood that they were also looking at a similar program that would evaluate the percentage of spalling around a rockwell indent. I don't know if they ever succeeded in getting it to work or not. The development of the software and hardware was done in Germany.

The rockwell indent worked good as a basic test since it was quick and relatively cheap. We did have issues going through indenters - TiN, CrN, TiC and so on are very rough on diamonds. I did some experimenting on chipped vs. new indenters (trying to recycle chipped diamonds that were no longer good for measuring hardness to save $$$, we did hundreds of rockwell indents a day on hardened tools - lots of diamonds lost). It did appear that a chipped indenter would cause more spalling than a new one.

Drawbacks to this test were:

1. Destructive - not all parts have an area where you can put an indentation
2. Usually could not get close to the working edge, where it counts.
3. A fingerprint (or other dirt) on the part will cause localized adhesion problems

I don't believe there were industry standards on these tests. Everything we had came out of Germany. I might still have the work instructions in my archives - I'll talk a look over the weekend and see what I can find.

Hope this helps. Let me know if you need anything else.

[tomvehoski]

A couple more thoughts as I recall...

Thickness can be directly related to adhesion. TiN is a very brittle substance. The thicker it is, the more likely it is to spall under the pressure of a rockwell indent.

Part geometry is important in PVD. Trying to remember my physics here, but the electrostatic field at a sharp point, such as the point of a drill, will be higher than on a flat surface, such as the shank. This will attract more coating at the point. Thickness is usually measured on the shank, since you can't ball crater a point. 2 microns on the shank might mean 4-6 on the point. This extra thickness could cause spalling that would not show on the shank, where the rockwell indent might also be done. The best way to get a true thickness on the cutting edge is with a cut/mount/polish operation, which is destructive and time consuming.

Cleanliness was the most important issue with adhesion. Poor cleaning = poor adhesion.

This is probably more than you wanted to know, but I hope it helps.

Tom

[Douglas E. Purdy]

Tom,

Thanks for the informative response. I would love to see the procedures that you may have archived.

As for the Daimler-Benz Adhesion Test, is the magnification the same for this as the Rockwell Hardness Test?

Thanks,
Doug

[tomvehoski]

I think the official name for the Rockwell test is Daimler-Benz. I know it sounds familiar. I believe we used 50X magnification, but I don't recall for sure. We normally set it up to get the closest view of the indent, but still show the surrounding area. I believe our procedure said something to the effect that <20% spalling was OK. I think I did have a scan of something Germany sent over, which is probably in a work instruction I know I have.

[tomvehoski]

I could not find the work instructions in my files. I may have them on a CD somewhere and will take another look.

I did find these links with a quick google search that may help you out.

http://www.badgermetal.com/coatings/...otos.htm#tialn


This is the company that makes the analysis software:

http://www.soft-imaging.com/

Hope this helps,

Tom

[Douglas E. Purdy]

Tom,

Again thanks for the information! The pictures in the badger link and your remembrance of 20% spalling, does not jive in my mind. Could you give more detail of acceptance criteria?

Thanks,
Doug

[tomvehoski]

The 20% criteria would mean that if >20% of the coating, measured in surface area, around the indent did not adhere it would be rejected. I don't recall if we actually had a spec for distance from the center of the indent that counted. Most of the examples in the link show excellent adhesion, with the exception of the CrC ones. The difficult part was determining exactly what percentage did not adhere in marginal cases. Normally there was little or no problem with adhesion. If we saw a large problem, we would normally have an engineer make a decision, or just recoat (if possible). Adhesion was usually very good or very bad, with minimal borderline cases.

From your e-mail you stated you did a ball crater test. This may not give good results for adhesion, since several factors could affect the look of the crater including:

1. Surface finish of the base material - a rough finish will give a rough ball crater outline and may look like bad adhesion.
2. If you rotate the ball too fast, it may bounce and cause a rough appearance.
3. Size of the diamond slurry can affect the appearance. I think we used either a .25 or .5 micron.

If you can give me some more details of the product, application, and so on I may be able to give you a little more insight.