# Adhesion testing

Source: https://coatmentor.com/learn/adhesion-testing/ · Updated 2026-10-04

Pull-off adhesion (ISO 4624, ASTM D4541) and tape tests (ISO 2409, ASTM D3359): when each is used and how results are reported.

*Figure: Where a pull-off test can break (ISO 4624). A letter is a layer, and a pair with a slash is the interface between two layers. Y is the glue.*

Adhesion tests show whether a coating holds to its substrate and to the coats between. Pull-off tests measure a stress in MPa (or psi). Tape tests rate a pattern of damage.

## The why: the weakest link breaks first

A coating system is a chain of bonds and layers. A pull-off test loads the whole chain until one link fails. The break shows which link was the weakest and at what stress.

The stress is the force divided by the area of the dolly. The failure mode is as important as the stress, because it points to the cause.

- A break at the steel surface points to contamination, soluble salts, a poor profile or bad conditions at priming ([surface cleanliness](https://coatmentor.com/learn/surface-cleanliness/)).
- A break between two coats can point to dirt or a missed recoat window between them ([cure and recoat](https://coatmentor.com/learn/cure-and-recoat/)).
- A break inside one coat points to a weak coat, such as a wrong mix or a poor cure.
- A break in the glue says nothing about the coating. It shows only that the glue joint was weaker than the coating.

## Pull-off in the field

1. Choose the tester. ASTM D4541-22 has five tester types (Methods B to F), and results from different types are not comparable.
2. Abrade the dolly face and a glossy surface, then clean both with solvent.
3. Glue the dolly on in a thin, even layer and remove the excess. Let it cure as the data sheet of the two-part glue says. Its temperature limits apply.
4. Cut round the dolly if the method requires it. ISO 4624:2023 allows you to leave out the cut below 150 µm if the parties agree. Report any cut.
5. Fit the tester, raise the load at no more than 1 MPa/s, then record the stress and the failure modes.

ISO 4624 dollies are 20 mm across (nominal). ASTM D4541 reports the failure as adhesive, cohesive or glue, and it has two protocols. Protocol 1 tests to fracture, and it is the default. Protocol 2 is a pass/fail test at a set load.

## The failure codes

ISO 4624 names each break with a letter code. A letter alone is a cohesive break inside a layer. A pair with a slash is an adhesive break between two layers.

| Code | Where the break is |
| --- | --- |
| A | Inside the substrate |
| A/B | Between the substrate and the first coat |
| B | Inside the first coat |
| B/C | Between the first and second coats |
| -/Y | Between the final coat and the glue |
| Y | Inside the glue |
| Y/Z | Between the glue and the dolly |

Estimate the share of each break to the nearest 10 %. Write the result like this: "18 MPa, 40 % A/B, 60 % B".

## A worked example

A pass/fail test to ASTM D4541 needs at least 4 MPa. Three dollies of 20 mm give these results.

- Dolly 1 breaks at 6.2 MPa, 70 % B and 30 % B/C. The coating itself broke, so the result is valid and it passes.
- Dolly 2 breaks at 3.1 MPa with 20 % glue failure and 80 % A/B. The glue failed over 5 % of the area while the stress was below 4 MPa. The test is inconclusive. Repeat it with a new dolly.
- Dolly 3 breaks at 4.8 MPa with 100 % glue failure. The glue broke above the minimum, so the coating is at least that strong. The test passes, reported as "greater than 4 MPa".

Check: the dolly area is π × 10² = 314 mm². Dolly 1 needed 6.2 N/mm² × 314 mm² = 1,948 N, which is about 1.9 kN.

## Tape tests: cross-cut and X-cut

A tape test is simple and empirical. The blade cuts through to the substrate. Tape then lifts any coating that is not bonded, and you rate the area that came off. ISO 2409 says that the test is not a means of measuring adhesion. For a measurement, use pull-off.

- **Cross-cut** (ISO 2409, ASTM D3359 Method B): a lattice of cuts. ISO 2409 uses six cuts each way. The spacing is 1 mm up to 60 µm on hard substrates. It is 2 mm from 61 to 120 µm and 3 mm from 121 to 250 µm. It does not suit films above 250 µm. ASTM D3359 Method B suits films up to 125 µm (5 mils).
- **X-cut** (ASTM D3359 Method A): two cuts that cross. It is mainly for films thicker than 125 µm (5 mils). For films above 250 µm, ISO 16276-2 also gives an X-cut procedure.

Rate the result against the class table. The two scales run in opposite directions.

| ISO 2409 class | ASTM D3359 class | Area affected |
| --- | --- | --- |
| 0 | 5B | None. The cut edges are smooth. |
| 1 | 4B | Not above 5 % |
| 2 | 3B | Above 5 % up to 15 % |
| 3 | 2B | Above 15 % up to 35 % |
| 4 | 1B | Above 35 % up to 65 % |
| 5 | 0B | Worse than 65 % |

Both tape tests damage the coating. Repair the test area afterwards.

## What the inspector records

- For pull-off: the stress, the failure codes with their shares, the glue, the dolly diameter, the pull rate and any cut.
- For tape tests: the cut spacing, the class and the film thickness.
- The place of each test and its repair.

ISO 16276-1 and ISO 16276-2 give site procedures. They apply only when the contract names the test and the value or rating.

## Exam traps

- Trap: "A glue failure means the coating failed." The glue failed first. The stress is only a lower limit of the coating adhesion.
- Trap: "Low stress with glue failure is a fail." In an ASTM D4541 pass/fail test, 5 % or more glue failure below the criterion makes the test inconclusive. Repeat it.
- Trap: "A/B is a break between the first and second coats." A/B is between the substrate and the first coat. B/C is between the coats.
- Trap: "Use a cross-cut on a 300 µm coating." ISO 2409 does not suit films above 250 µm.
- Trap: "Method A is the cross-cut." In ASTM D3359, A is the X-cut and B is the cross-cut.
- Trap: "Class 5 is the best result." On the ISO scale class 0 is best. On the ASTM scale 5B is best.

## How the exams treat it

Note: Destructive tests are outside the scope of CIP 1. The CIP 2 guide names adhesion tests, and the CIP 2 practical has a pull-off adhesion station worth 15 points. A FROSIO training body lists cross-cut, X-cut and pull-off stations in its course. See [ISO 4624](https://coatmentor.com/standards/iso-4624/), [ASTM D4541](https://coatmentor.com/standards/astm-d4541/), [ISO 2409](https://coatmentor.com/standards/iso-2409/) and [ASTM D3359](https://coatmentor.com/standards/astm-d3359/).

**Use it:** [Standard: ISO 4624 / ASTM D4541](https://coatmentor.com/standards/iso-4624/)

**116** practice items in the members area mention this topic (CIP 1: 30 · CIP 2: 57 · FROSIO written: 29).

## Related

- **Standards:** [ISO 4624](https://coatmentor.com/standards/iso-4624/) · [ISO 2409](https://coatmentor.com/standards/iso-2409/) · [ISO 16276](https://coatmentor.com/standards/iso-16276/) · [ASTM D4541](https://coatmentor.com/standards/astm-d4541/) · [ASTM D3359](https://coatmentor.com/standards/astm-d3359/) · [NACE SP0892](https://coatmentor.com/standards/nace-sp0892/) · [ASTM D7234](https://coatmentor.com/standards/astm-d7234/)
- **Defects:** [Flaking and delamination](https://coatmentor.com/defects/flaking/) · [Delamination (intercoat adhesion failure)](https://coatmentor.com/defects/delamination/) · [Peeling](https://coatmentor.com/defects/peeling/) · [Laitance](https://coatmentor.com/defects/laitance/)
- **Inspection and testing:** [Dry film thickness measurement](https://coatmentor.com/learn/dft-measurement/) · [Holiday (discontinuity) testing](https://coatmentor.com/learn/holiday-testing/)

## About this page

**Sources:** ISO 4624:2023 · ASTM D4541-22 · ISO 2409:2020 · ASTM D3359-23 · ISO 16276-1:2025 · ISO 16276-2:2025.

**Also published openly:** [Elcometer 510 pull-off tester instructions, nature of fracture codes](https://www.southampton.ac.uk/assets/imported/transforms/content-block/UsefulDownloads_Download/ED337B88B6D74F43B030846FDB393223/510.pdf) · [Gardco: PosiTest CH cross-hatch kit, ISO 2409 blade spacing](https://gardco.com/Products/Adhesion-Testers/Cross-Cut-Testers/PosiTest-CH-Cross-Hatch-Adhesion-Tester/c/p-100604).

Written from the published standards and coating practice, in our own words, and checked on 2026-10-04. Found an error? It goes in the [corrections log](https://coatmentor.com/method/). CoatMentor is an independent study aid. It is not affiliated with ISO, AMPP, SSPC, NACE International, ASTM or FROSIO.