What ASTM D3359 cross-hatch grade should a powder-coated part have to pass?

What does an ASTM D3359 rating actually prove, and what does it not?

It proves that the coating stays attached to the substrate when a sharp edge and a peel load are applied together, which is the failure a pretreatment or cure problem produces. It does not measure how strong the bond is, so it cannot rank two coatings that both reach the top class.

The scope of ASTM D3359 says the methods establish whether adhesion is at an adequate general level; they are not meant to distinguish between higher levels of adhesion. That is exactly what an incoming-inspection test needs: a fast, cheap, binary-feeling check that catches a skipped conversion coating, a contaminated machined face or an undercured film. It is the wrong tool if you want to compare two powder suppliers, prove a design margin, or set a number for a structural bond. For that, ASTM D4541 pull-off testing reports a tensile value and the layer where the failure occurred.

Two consequences follow for the acceptance criterion. First, a 5B result on a passing lot tells you nothing about how close that lot was to failing, so the sampling plan and the test location carry more weight than the grade. Second, the grade is only comparable between two labs when everything around the cut is the same: tape, cutter, timing, part temperature and where on the part the lattice sits.

How do you choose between 5B, 4B and 3B as the acceptance grade?

Choose by what happens to the part after it leaves the coater: 5B for faces that see fasteners, gaskets, clips or cosmetic inspection, 4B as the general industrial floor, and 3B only for hidden or primer-only surfaces where flaking at a cut edge does no harm.

The classes describe how much of the lattice the tape removed. Per ASTM D3359, 5B means the cut edges are completely smooth with no squares detached, 4B allows small flakes at intersections affecting less than 5 % of the area, and 3B allows flaking along edges and at intersections affecting 5 to 15 % of the lattice. A properly pretreated and fully cured powder on clean aluminum, zinc or steel usually reaches 5B in practice, so the question is really how much of the gap below 5B the part can tolerate.

• 5B minimum. Cosmetic faces, faces under a screw head or a clamp, sealing faces next to a gasket, and anything outdoors. Fasteners and gaskets are field versions of the cut-and-peel test: a film that lifts at a lattice edge is likely to lift at a torqued washer. Architectural powder specifications from the fenestration industry likewise call for no loss of adhesion on their cross-hatch test.
• 4B minimum. Functional industrial housings and brackets with no cosmetic grade, where a little edge flaking at an intersection is not a defect in service. This is the most common industrial default and a reasonable class for machined faces that were coated after machining.
• 3B. Rarely appropriate for a powder applied over a conversion coating. A lot sitting at 3B often has a pretreatment or cure problem that can show up later as underfilm corrosion, so treat 3B as a trigger for investigation even where the drawing tolerates it.
• 2B and below. Not an acceptance grade. Removal at that level points to a missing or contaminated pretreatment, a wrong cure, or an incompatible primer.

It is legitimate, and often smarter, to set two grades on one part. Cast skin, machined faces and sharp edges do not behave the same, and the guide on pretreating zinc die castings for powder explains why machined zinc and cast skin pretreat differently. Write the higher grade against the surfaces that matter and say so with a surface flag on the drawing.

How should the drawing or purchase order callout read?

Name the edition, the method, the spacing, the minimum class by surface, the sample, the timing and the tape, so an inspector at either end produces the same result from the same part.

An example note for a single-coat powder on an aluminum housing, written in drawing style:

FINISH: POWDER COAT, COLOR AND GLOSS PER APPROVED LIMIT SAMPLE.
FILM THICKNESS PER ASTM D7091, 60–100 µm ON SURFACES MARKED C.
ADHESION PER ASTM D3359-23 METHOD B, 2 MM CUT SPACING.
ACCEPTANCE: 5B MIN ON SURFACES MARKED C; 4B MIN ON ALL OTHER COATED SURFACES.
TEST THREE PARTS PER CURE LOT AFTER PARTS REACH ROOM TEMPERATURE; ZERO FAILURES ALLOWED.
TAPE: AS LISTED ON THE PURCHASE ORDER; RECORD LOT NUMBER AND PHOTOGRAPH EACH LATTICE.

The tape line is not bureaucracy. Recent editions of the standard no longer name a tape product; the tape’s adhesion strength is agreed between supplier and user, and the standard warns that tape adhesion can vary from batch to batch and with age, so two inspectors using different tapes can rate the same coating differently. Name one tape on the PO, and keep a roll of the same lot with the retained samples if the part is important.

Common mistakes to avoid when writing the note:

• Citing “ASTM D3359” without a method letter, so the inspector chooses between an X-cut and a lattice.
• Citing ISO 2409 on one document and ASTM D3359 on another. The ASTM standard itself says the two are similar but not technically equivalent, and the ISO classes run in the opposite direction (0 is best), so “class 1” means opposite things under the two documents.
• Setting a grade with no surface definition, which lets the test be run on whichever face looks best.
• Using the tape test as a cure check on its own. A film can pass 5B while still undercured; pair it with a solvent rub or the cure check from the powder maker’s data sheet if cure is the concern.
• Leaving the post-coat thread or bore condition out of the same note. The guide on surface finish before or after coating covers how to sequence those requirements.

Before you commit the note to a drawing, check that the finishing source actually runs the test you name. On a Saola order, powder coating and pad printing are applied by partner plants that Saola manages, on parts machined at its lead CNC plant or sourced as castings and stampings, as described on the surface finishing page; the quote names where each step is done. The inspection and test documents Saola lists on its quality page include first-article inspection reports, dimensional inspection records and salt-spray results, with the salt-spray standard and duration confirmed in the quote; a cross-hatch adhesion record is not on that list. So if your purchase order requires an ASTM D3359 result, put it in the RFQ: the surface finishing page asks buyers to say so before tooling when a specification calls for a test standard outside its menu. Get the answer in writing in the quote.

What should the RFQ say about powder coat adhesion?

Send the adhesion requirement with the drawing, not after the first lot: method, class by surface, sample size, tape, conditioning and which records you expect to receive with each shipment.

• The base material and how the surface is produced: cast skin, machined, stamped or a mix, flagged on the drawing.
• The film thickness range and the method used to measure it.
• The adhesion method, edition and spacing, and the minimum class for each flagged surface.
• Whether test panels are acceptable, and how many production parts per cure lot must be tested.
• Dry only, or dry plus wet adhesion after a named humidity exposure.
• The tape product, and whether lattice photographs must accompany the shipment record.
• Any cosmetic limit sample, and who signs it. Saola finishes to the cosmetic standard the customer supplies, whether a boundary sample, a written specification or marked-up photos, and asks for one before production when none is supplied, per its quality page.

The general structure of a complete package is in the guide on how to prepare a machining RFQ, and the first article inspection guide covers where the adhesion result belongs in the first-article record. When the drawing is ready, send it through the quote form.