Key takeaways
- Cut spacing is set by dry film thickness: 1 mm up to 60 µm, 2 mm from 61 to 120 µm, 3 mm from 121 to 250 µm, and no valid ISO 2409 test above 250 µm.
- A 1 mm lattice carries 2.4 mm of cut edge per square millimetre against 0.8 at 3 mm, so the same coating can read Gt1 at 2 mm spacing and Gt3 at 1 mm.
- ISO 2409 gives a grade, not an acceptance limit: the purchaser has to state Gt0, or Gt0 to Gt1, and the area of the part that grade applies to.
- Textured powder, wrinkle finishes and films above 250 µm are outside the method; specify ISO 4624 pull-off with a stress in MPa instead.
Two panels leave the same powder line with the same 70 µm film and both read Gt1. One holds 4 MPa on a pull-off test and the other holds 9 MPa. That gap is not a defect in the standard. It is the difference between a test that ranks and a test that measures, and knowing which one the drawing asked for is the whole skill.
ISO 2409 cross-cut adhesion is the most widely quoted coating test in sheet metal and the most loosely applied. The standard fixes the blade spacing to the dry film thickness, and a supplier who uses 1 mm spacing on a 90 µm film has made the test roughly three times more severe than the document allows. The part then fails, the coating takes the blame, and the true cause is a blade chosen out of habit.
What follows is the test as written, the arithmetic that shows why spacing decides the grade, and the boundary where a cross-cut stops being the right question.
What the cross-cut test actually measures
ISO 2409:2020 specifies a lattice of six parallel cuts in one direction and six at 90°, cut through the coating to the substrate, followed by a controlled tape pull and a visual classification of how much coating lifted. It is a qualitative ranking of adhesion. Nothing in the document produces a strength value, and no result from it can be compared with a pull-off figure in MPa.
The method survives because it is cheap and portable: a multi-blade tool, a roll of specified tape and good light. On a flat panel it takes under two minutes, so it can run on every batch rather than on a laboratory sample taken once a quarter.
The grade scale has six steps. Gt0 means the cut edges are completely smooth and no coating detached. Gt1 allows up to 5% of the lattice area to detach, Gt2 covers 5% to 15%, Gt3 15% to 35%, Gt4 35% to 65% and Gt5 more than 65%. The percentage refers to the area inside the lattice, not to the part, and that distinction matters when the lattice is only 5 mm across: one flake of coating can move a small lattice from Gt0 to Gt2.
Two things the standard deliberately leaves open are the acceptance level and the test location. Both belong in the purchase specification. A drawing that says "adhesion to ISO 2409" without a grade has specified a method and no requirement, and a supplier can meet it with a Gt3.
The cut spacing and grade table
| Parameter | ISO 2409:2020 requirement | Ganesh practice and limit |
|---|---|---|
| Dry film 0–60 µm | 1 mm cut spacing | nominal 60 µm work, tested at 1 mm |
| Dry film 61–120 µm | 2 mm cut spacing | the standard 60–80 µm powder range; every batch at 2 mm |
| Dry film 121–250 µm | 3 mm cut spacing | two-coat and high-build work, spacing agreed at enquiry |
| Dry film above 250 µm | outside the scope of the method | no cross-cut claim is made; pull-off to ISO 4624 |
| Cuts per direction | six, at 90° to each other | template-held tool, not freehand |
| Cut depth | through the coating to the substrate | checked on a scrap part at line setup |
| Tape | specified pressure-sensitive tape, 25 mm wide | one brand per line, recorded on the batch card |
| Removal | at approximately 60°, within 5 minutes of cutting | timed by the inspector, not left to the end of shift |
| Gt0 | no detachment; edges completely smooth | acceptance for outdoor enclosures and canopy panels |
| Gt1 | up to 5% of the lattice area detached | acceptance for general industrial powder parts |
| Gt2 | 5% to 15% detached | reject unless the drawing says otherwise |
| Gt3 to Gt5 | 15% to over 65% detached | reject; the batch is re-tested before dispatch |
| Test location | a representative area of the coated part | flat face, one laser-cut edge, one hole |
| Textured finishes | not addressed by the method | excluded; alternative method named at enquiry |
Why the spacing decides the grade
A lattice of six cuts at spacing s has a side of 5s and a total cut length of 12 × 5s, which is 60s. The area enclosed is 25s². Cut length per unit area is therefore 2.4/s mm per mm²: 2.4 at 1 mm spacing, 1.2 at 2 mm and 0.8 at 3 mm.
At 1 mm spacing a panel carries three times the cut edge per square millimetre that it carries at 3 mm. Every cut is an initiation site, because the blade lifts the film locally and the tape then peels from those lifted edges. Three times the edge gives roughly three times the coating available to detach at the same underlying adhesion.
There is a second effect, and on thick films it dominates. The blade does not remove a line of zero width; it cuts a groove with a damaged zone either side, of the order of 0.1 to 0.2 mm on a 90 µm polyester film. At 1 mm spacing, 0.2 mm of damage on each side leaves 0.6 mm of intact coating in a 1 mm land, so 40% of the land is compromised before the tape is applied. At 2 mm spacing the same damage leaves 1.6 mm of 2 mm, or 20%.
Put the two effects together and a panel that reads Gt1 at 2 mm can read Gt3 at 1 mm with nothing about the coating changed. The part did not get worse. The test did.
The depth problem on thin substrates
On 0.8 mm and 1.0 mm CRCA the blade has to cut through the film and stop at the steel. A worn tool, a slightly convex panel or a heavy hand takes the blade into the substrate, and the burr it raises changes both the cut geometry and the tape contact. Check the tool on a scrap part at every setup: six clean lines, no bright steel swarf, and a lattice that lifts as a set of squares rather than a continuous strip.
What the percentage means in square millimetres
The grade is an area fraction, so the same grade tolerates different amounts of damage at different spacings. A six-cut lattice at 2 mm spacing is 10 mm × 10 mm, or 100 mm², and Gt1 at 5% allows 5 mm² to lift — a little more than one 2 mm square. At 1 mm spacing the lattice is 25 mm², and 5% allows 1.25 mm², which is barely one square plus its neighbour.
So a Gt1 at 1 mm spacing is a quarter of the absolute damage of a Gt1 at 2 mm, on a test that is already three times more severe per unit area. The two effects compound. When a supplier reports "Gt1" with no spacing, the number could describe 1.25 mm² of lifted film or 5 mm², and there is no way to tell which from the certificate.
What a grade change tells you
A supplier whose results drift from Gt1 to Gt3 across three batches is not describing a coating that has weakened by a measurable amount. More often it is a different operator, a dulled blade, an unauthorised tape, or a film that has moved from 118 µm to 124 µm and crossed the 120 µm boundary in the table. Ask for the film thickness with the grade. A Gt3 at 130 µm tested at 2 mm is a real failure; a Gt3 at 130 µm tested at 2 mm against a 3 mm requirement is a test error.
Where ISO 2409 stops
Three boundaries are worth writing into a purchase specification, because each one turns a valid result into a meaningless number.
Film thickness. The classification applies up to 250 µm. A high-build epoxy at 300 µm, a two-coat system at 280 µm or a tank lining at 400 µm cannot be graded Gt0 to Gt5 at all: no spacing in the table is valid, and the multi-blade tool will not cut cleanly to the substrate through that depth. The substitute is a pull-off test to ISO 4624, which reports the tensile stress at which the coating separates and where in the coating stack the break occurred.
Surface texture. A wrinkle, hammer-tone or coarse textured powder has a surface the blade cannot score uniformly and a tape contact area that varies across the lattice. Results move by two grades between operators on the same panel. A drawing that calls for Gt1 on a textured finish is asking for a measurement that cannot be reproduced, and the fix is to change the test rather than tighten the requirement.
Geometry. The multi-blade tool has to sit flat. On a 20 mm flange, a tight bend radius or a curved canopy panel only part of the tool contacts the surface and the cut depth varies along the lattice. Test a flat coupon from the same batch with the same cure and say in the inspection plan that the coupon is the acceptance piece.
There is also a scope point that catches procurement teams: the cross-cut is not a corrosion test and not a cure test. A part can read Gt0 and still fail a 480-hour neutral salt spray because the film is thin at a laser-cut edge or the pretreatment is wrong beneath an intact film. Adhesion and corrosion resistance come from different mechanisms and need separate evidence. Where the question is how long a coating lasts outdoors, the test is ISO 9227 salt spray with a scribe, plus a dry film thickness record and a pretreatment record.
The standard is also silent on how many parts to test. One panel per oven load is common practice for a single batch. A first article should be three, from the start, middle and end of the load, because the cure profile varies across a rack.
The cross-cut on the Ganesh line
Coated parts run through the seven-tank pretreatment line, the 18 ft × 14 ft electrostatic powder booth, a wet paint booth and a curing oven. Dry film thickness is measured with a gauge on production parts, and cross-hatch adhesion is checked on the same parts rather than on coupons alone.
The test locations that matter here are a laser-cut edge and a punched hole. Edge film builds to roughly half the face thickness, so the tape starts lifting at the edge first, and a part that reads Gt0 on the flat face can read Gt2 across a dross line. The edge coverage note covers why that happens, and breaking the cut edge before coating does more for the grade than any change to the acceptance criterion.
The plants hold ISO 9001:2015 and ISO 14001:2015, and enclosure, canopy and lift programmes run through the same paint shop. That means a 1.5 mm CRCA door and a 3 mm canopy panel come off the same line settings and are judged against the same grade.
What to put on the drawing
- The full reference with the year: "cross-cut to ISO 2409:2020", not "adhesion test".
- The cut spacing, or the dry film thickness that determines it.
- The acceptance grade and the surface it applies to: "Gt0 to Gt1 on face, edges and around holes".
- The test frequency: one part per batch, or one per oven load.
- Whether a coupon is acceptable where geometry prevents a flat cut.
- The tape specification and removal angle, if a house standard exists.
- The alternative method and acceptance stress for any film above 250 µm.
- The cure record, because a Gt0 result on an undercured film is a failure that appears three months later.
Batch records for coated parts travel with the finished coated assemblies route, so the grade, the thickness and the cure data can be read together. The standards and compliance hub collects the other test methods that sit alongside this one, and the powder coating failure guide covers what a Gt3 usually means in service.
When the cross-cut is the wrong test
Cross-cut is the right test for a 60 to 120 µm powder or wet film on flat steel or aluminium, inspected at the start of production and audited on sample parts. It is the wrong test in four situations.
For a high-build or multi-coat system above 250 µm, use ISO 4624. For a textured finish, use pull-off or an impact test and state the method. For a part already installed on a customer's site, a cross-cut leaves a scar that has to be touched up, and a pull-off leaves a larger one; neither is a good field check on a finished enclosure, and the inspection should have happened in the plant.
For galvanised steel with a heavy zinc layer, the blade may not reach steel at all and the result describes the coating-to-zinc interface rather than the coating-to-substrate one. That is a legitimate result, but it is a different one, and it should be labelled as such.
Send the substrate, the film thickness range, the finish type and the acceptance grade with the drawing. The quotation will then state the spacing, the test location and the tape it assumes, and the batch record will match the drawing at audit.
Ready to specify the test properly? Request a quote with the current revision and the coating requirement, and the engineering team will confirm the adhesion method before pricing.
Frequently asked questions
What cut spacing do I use for a 90 µm powder coating?
2 mm. ISO 2409 sets 1 mm for films up to 60 µm, 2 mm for 61 to 120 µm and 3 mm for 121 to 250 µm. A 90 µm film tested at 1 mm spacing is three times more severe than the standard and will fail parts that meet the specification.
Is Gt0 a pass?
Only if the drawing says so. ISO 2409 defines the grades but no acceptance level. General industrial powder work is usually accepted at Gt0 to Gt1, outdoor enclosures at Gt0. The grade is meaningless without the spacing and the test area, so state all three.
Can I cross-cut a coating thicker than 250 µm?
No. Above 250 µm no spacing in the table applies and the multi-blade tool will not cut cleanly to the substrate, so the result is not a test to ISO 2409. Use ISO 4624 pull-off and specify an acceptance stress in MPa with the allowed fracture location.
Why do two inspectors get different grades on the same panel?
Usually tape. The standard calls for a pressure-sensitive tape of defined adhesion, rolled down at a set pressure and peeled at about 60° within five minutes of cutting. Packaging tape lifts more coating. Blade wear and cut depth on a curved or narrow surface cause the rest of the spread.
Does a Gt0 result prove the coating will not corrode?
No. Cross-cut measures how well the film is attached; corrosion resistance depends on pretreatment, film thickness at edges and the barrier the film provides. A part can read Gt0 and still fail a 480-hour neutral salt spray because the laser-cut edge carries 40 to 60% of the face film thickness.
Where on the part should the test be done?
On a production part, on the flat face, across one laser-cut edge and around one hole. The face is the easiest place to pass and the least informative. If the geometry prevents the tool from sitting flat, test a coupon from the same batch and cure and say so in the inspection plan.
Sources and further reading
- ISO 2409:2020 — Paints and varnishes, cross-cut test · reviewed 3 Nov 2026
- ISO 4624:2023 — Paints and varnishes, pull-off test for adhesion · reviewed 3 Nov 2026
- ISO 9227:2022 — Corrosion tests in artificial atmospheres, salt spray tests · reviewed 3 Nov 2026
Need an adhesion record with the batch?
Send the drawing with the film thickness, the substrate and the acceptance grade. We confirm the spacing, the test locations and the tape before the first panel is coated.
Request a quote