Powder Coating Adhesion Problems and Tests (Cross-Cut)

When a chassis bracket comes back from the field, the first places to look are the corners and bends. If the coating flakes off there, the problem usually did not start at the gun but in the steps before it. Adhesion describes how well the coating “grips” the metal, and it is a measurable quantity. Below, I explain what causes adhesion to fail, how the cross-cut test — which can be done in 10 minutes in the field — is performed, and how the result is read with the ISO 2409 classification table.

Why adhesion is lost

Adhesion loss cannot be traced to a single cause; it is usually several weak links stacking up. The most common root causes are:

• Inadequate pretreatment. Oil, grease, drawing lubricant or fingerprint residue gets between the metal and the film. Powder applied before a conversion coating (iron/zinc phosphate or nano-ceramic) has formed finds no surface to grip mechanically. Salts left after poor rinsing also draw moisture and cause blistering from below.
• Undercure. If the powder melts and flows but crosslinking is not completed, the film stays soft and cannot bond chemically to the metal. That means a surface that lifts easily in the cross-cut test.
• Excessive film thickness. Going to 150 µm on a job targeting 60–90 µm increases internal stresses; at bends, the film separates from the metal.
• Wrong surface profile. On an overly smooth surface with no roughness at all, mechanical keying is weak; with overly coarse blasting, the film cannot adequately cover the peaks.

Most of these causes are invisible. The part comes out of the oven looking fine, the color is right, and by hand and eye there seems to be no problem. That is why adhesion must be verified by testing, not by eye. When hunting for the root cause, a practical order is to review the pretreatment and conversion coating step first, then the curing temperature and time.

Cross-cut test (ISO 2409 / ASTM D3359)

The cross-cut test is the most common and fastest way to measure adhesion. The logic is simple: the film is divided into small squares by intersecting cuts; then tape is pulled off and you see how many squares have lifted. The equipment is inexpensive and the result can be read in a few minutes.

Choose cut spacing according to film thickness

ISO 2409 ties the distance between the blade teeth to film thickness. Rule of thumb:

• 0–60 µm, hard substrate: 1 mm spacing.
• 0–60 µm, soft substrate: 2 mm spacing.
• 61–120 µm: 2 mm spacing.
• 121–250 µm: 3 mm spacing.

Since powder films are usually above 60 µm, a 2 mm six- or eleven-tooth blade is most often used in the field. It can also be done with a single blade and a ruler, but a multi-tooth cutter gives faster and more repeatable results.

Step-by-step procedure

• Clean the surface; the area to be cut must be free of oil and dry.
• Holding the blade perpendicular to the surface, make 6 (or 11) parallel cuts at a steady speed, in a single pass, down to the substrate.
• Rotate the part 90° and repeat to create a grid pattern. The cuts must reach the metal; merely scratching the film gives a false result.
• Remove loose particles with a soft brush.
• Apply the adhesive tape specified in the standard (peel strength of roughly 5–6 N/25 mm) over the grid, pressing it down so no air bubbles remain.
• Wait about 90 seconds, then pull the tape off in one motion within 5 seconds at an angle of about 60° to the surface.
• Inspect the grid under good lighting and classify it according to the table below.

Mandrel bend and impact test

The cross-cut test measures pure adhesion but gives limited information about the film’s flexibility. If the part is subject to bending or impact in the field, two more tests come into play.

Mandrel (cylindrical) bend test: The coated sheet is bent around a rod of a given diameter. Cylindrical mandrels range from 2 to 32 mm in diameter; with a conical mandrel, the effect of different radii is seen in a single bend. After bending, the outer surface is checked for cracking or film lifting. Polyester and polyurethane systems are generally more flexible than epoxy; the exact limit again depends on the resin and the TDS.

Impact test: A standard-weight head is dropped onto the coated surface from a set height; both direct (front) and reverse impacts are applied. Cracking or peeling of the film at the point of impact is assessed. Reverse impact is generally more demanding, because the film is stretched as the metal deforms.

These two tests also catch the brittleness that appears when the film is cured too hard (overcure). So if the mandrel/impact result is poor, the problem is not always adhesion; sometimes it may be the curing regime.

When deciding which test to use and when, go by the part’s real load. For a fixed, flat panel, the cross-cut test is usually sufficient. For parts that are bent or pressed during assembly or take impacts in transport, add the mandrel and impact tests as well. Whenever possible, take the sample from a panel that has gone through the same pretreatment and oven cycle as the real part; the result from a test coupon that came out of a different cycle does not represent production.

Interpreting the results

In ISO 2409, the cross-cut result is rated in six classes from 0 to 5; 0 is the best, 5 the worst. The table below gives the classes together with the roughly corresponding ASTM D3359 (0B–5B) values.

ISO 2409 class	Condition of cut edges/intersections	Area detached (approx.)	ASTM D3359 equivalent
0	Cut edges completely smooth; no squares detached	%0	5B
1	Small flakes detached at intersections	Less than 5%	4B
2	Noticeable detachment along edges and at intersections	%5–15	3B
3	Squares partly/wholly detached along the edges	%15–35	2B
4	Large detachment in ribbons	%35–65	1B
5	Heavier detachment not covered by any of the above	More than 65%	0B

In practice, class 0 or 1 is expected for most industrial powder coating work; architectural and automotive specifications often require class 0 (Gt0). Class 2 is borderline acceptable, while class 3 and above is generally grounds for rejection. However, these values are not absolute: the acceptance limit is set by the part’s service specification and the powder manufacturer’s technical data sheet (TDS). The same class may be accepted on a decorative indoor panel but rejected for an outdoor pole.

What to do if the test fails

Separate a low adhesion result along two axes. If detachment occurs between the metal and the film (bare metal is visible under the peeled area), the root cause is almost always pretreatment: degreasing, conversion coating or rinsing. If detachment occurs within the film/between coats, look at curing and thickness. Keeping regular cross-cut records also makes it easier to trace back what went wrong in which batch. These records also serve as a reference when diagnosing other visual defects.

Adhesion is not an isolated value but an indicator of the line’s health. When test results start to deteriorate, usually the concentration of the pretreatment baths has dropped or the oven profile has drifted. That is why making the adhesion test a fixed step of routine quality control is the cheapest way to catch defects on the line rather than in the field.