How to Identify Substrate or Paint Defect

Q. Over the past several years we have had a coating defect anomaly randomly show up on painted parts. The defect shows an area of the coating surface that is disrupted in such a way that it resembles a crater. However, after examining numerous defects, no substrate abnormality has been found under the paint on the part. Abnormalities like gouges, scale and substrate craters have not been detected in the defect area. Additionally, the adhesion in the defect area has been excellent. If it were not for the coating disfigurement and color shade difference on lighter parts, no one would be aware that a cosmetic defect was present.

Material is 11-gage hot rolled, pickle and oiled, temper rolled with a rust preventative. The steel is laser cut, formed, resistance welded and finally, welded to print. Laser oxide is manually removed from the cut edges. Finished parts are then solvent degreased, wiped down with a solvent soaked towel, then a dry towel. Parts are sanded to remove weld spatter, rust, scratches, and are then solvent degreased as previously described.

After this, parts are hung on a monorail conveyor where they go through pretreatment consisting of a heavy-duty cleaner, five rinses, iron phosphate, two rinses and two RO water rinse stations. Parts are dried for 20 minutes at 275°F followed by an automated powder paint booth applying a polyester TGIC powder. Cure is by IR for 60 seconds and convection oven for 60 minutes at 400°F.

There is some feeling that the root cause is related to the HRP&O substrate. This problem has never been experienced with cold rolled steel and when it occurs, seems to happen only with a specific batch of fabricated samples. The previous and next batch of differing parts through the paint line will not exhibit the condition, but may affect up to about 50 percent of one batch of parts. It is believed the defect could be due to a subsurface contaminant related to the steel manufacturing process, possibly trapping rust inhibitor and oil.

A. Your disciplined approach to problem solving should be commended, but would also be assisted by including analytical tools into your investigation that would allow greater focus on root cause identification.

I would first try to categorize the contaminant as organic (i.e., oil) or inorganic (i.e., oxide). Working with an independent lab, you could request a defective location be sectioned to determine what is beneath it. This could be done with a section and mount in plastic, which could be inspected at high magnification and analyzed by SEM/EDS to provide a visual and elemental chemistry in the region between the paint and the steel. SEM/EDS will find carbon and oxygen in almost all analysis, but the relative amounts in comparison to a good location would be of interest, along with the finding of any unusual elements that may align with a lubricant or oxide.

Another technique would be to use a razor blade to remove the coating and characterize the substrate directly beneath the defect with SEM/EDS. Is there visual evidence of a phosphate coating at elevated magnification? Are there elements besides carbon, oxygen and iron at the surface? What do those elements correspond to? For example, is there evidence of sulfur, chlorine or phosphorous that could be residue from an extreme-pressure lubricant additive? How does that compare to the surface of the as-received steel?

A third technique would be to take the paint chip removed from the steel and use Infrared Spectroscopy (IR) to characterize the paint surface that was in contact with the steel in the area of the defect. The IR can remove the background spectra of the polyester coating and document unexpected findings like alcohols or organic acids that may be evidence of boundary additives from lubricants or amines that would be from a rust preventative.

Ideally, this will also be compared to a known, quality substrate to better understand the difference between paints. When this type of analysis is complete, you will have a better understanding of the defect and its root cause.

Originally published in the August 2017 issue.