bend test for thermal-sprayed zinc

[HgTX]

Can anyone think of a reason one should be failing bend tests but passing pull-off adhesion tests?

Also, I wonder if we're failing or just doing the test wrong. The obvious fails look just like the failing picture in SSPC CS-23 or AWS C2.18. But the rest look just like the "marginal" picture, except that a sharp knife will pick off the coatings at every crack location. Does "lifting" mean doesn't take much with the knife to knock it loose, or we can pick it off in any way? Does "knife" mean dull putty knife, or sharp utility blade?

Hg

Eng-Tips policies: faq731-376

 

[HgTX]

I should add that this is flame spray.

Hg

Eng-Tips policies: faq731-376

 

[unclesyd]

We flame sprayed a lot of Zinc at one time to protect drip coolers and ice house condensers as well as ancillary components and I only remember two or three bending test on a sample traveler coupon. If we were bad we were real bad. If I recall correctly the failed miserably so we tried he ball peen hammer test, someone said it was a Navy test. The only kind of blade I've seen picking at the Zinc was a normal pocket knife. I called a friend who used to do flame spraying and he said the he used a pocket knife blade to check the coating. He stated that some cracking was was OK as long as it couldn't be lifted, disbonded.

A little note as I know your are probably held down by spec's we went to 100% Al flame spray on all components because we were having so much trouble with the Zinc. We tried the 85/15 material
but it was no better in our application.

Is the substrate metal or concrete?

 

[HgTX]

Substrate is steel. We were using an extension that "bends" the flame to get into pockets, and gave up and went to regular gun, and that helped. But the pull-off adhesion tests from the very same method that failed the bends were really good.

Here's something to wonder about--coating thickness is measured from the peak of the profile to ensure coverage. But if you take a micrometer to the pieces that flake off, they're considerably thicker because they include profile depth. So, for the purpose of determining mandrel size for bend test, should one be including or excluding the profile depth from consideration?

Hg

Eng-Tips policies: faq731-376