Blisters After Cure

Q. We have a problem with blisters in the coating after cure. It is very bad some days and not too bad on other days, but it never goes away completely. The blisters are mostly on the top of the parts, although they show up on the sides sometimes, too. We use a spray washer with an alkaline cleaner, a rinse, iron phosphate, a tap-water rinse and then a seal rinse. We measure our chemistry three times a shift and are in range. Our chemical supplier has not been able to identify any root causes. What should we look for and how should we look for it? B.A.

A. Like any defect, identification of the precise cause is difficult without testing and an examination of the system. However, blisters do have some common causes that you can look into. A blister is caused by some type of moisture or chemistry on the substrate under the film. You could have some kind of contamination that the washer cannot remove, but it sounds more like something is dripping on your parts near the washer exit.
The first thing I would do is test the surface under the blister to see if you can identify any specific chemistry. You can have the area lab-tested using scanning electron microscopy-energy dispersive analysis (SEM-EDS) to look for things that should not be there. For example, you may find alkaline salts in the analysis. Alkalinity is a common cause of blisters and adhesion loss, and it should not be present when the part exits the washer. If you find alkalinity in the blister area, go to the washer and test for carryover of alkalinity around the chain area or conveyor shroud and make changes to stop the carryover. You can use phenolphthalein to indicate the presence of alkalinity in water samples from the chain or other washer surfaces. At a pH of about 9.3, the ions in the phenolphthalein will begin to turn pink; a higher alkalinity generates a brighter pink.
Alkalinity is one possible cause of the blisters. Other mineral salts, oil or synthetic lubricants are others. The SEM-EDS may find things that do not belong there. Use the test data to figure out where to look in the system for the root cause.