Pull-off adhesion testing

Pull-off adhesion testing bonds a loading fixture, commonly called a dolly, to a coating and applies tensile load approximately perpendicular to the surface. Separation occurs at the weakest plane in the dolly–glue–coating–substrate assembly. The result can therefore be limited by glue, a coating layer or the substrate rather than the coating-to-substrate interface. Pull-off results are not interchangeable with tape, knife or scratch-test ratings. DeFelsko: test methods for coating adhesion

Select the applicable method and question

ASTM D4541-22 addresses portable pull-off testing of coatings on metal substrates. It distinguishes testing to fracture from a pass/fail protocol that checks whether a specified load is sustained. Apparatus type, glue, fixture size, coating cure, loading rate, environment and scoring can affect the result; comparisons require controlled conditions and identification of the actual method. This test can damage the coating and require spot repairs. ASTM D4541-22

ASTM D7234-22 addresses coatings on concrete. It emphasizes both strength and failure mode, sufficient coating cure and agreement on instrument and fixture. If determining cohesive strength of concrete or a cementitious surfacer is the objective, ASTM C1583 may be more suitable. Do not select a concrete procedure simply because the same tester can operate on metal and concrete. ASTM D7234-22

ISO 4624:2023, edition 4, dated June 2023, includes three arrangements: two dollies for rigid or deformable substrates, one-sided single-dolly testing for rigid substrates, and an arrangement using a dolly as the painted substrate. Substrate deformation matters. A single-sided field pull on a thin flexible panel does not automatically reproduce a supported laboratory assembly. Specify the applicable standard, edition and arrangement before comparing results. ISO 4624:2023

Stress, area and instrument configuration

Nominal pull-off stress is force divided by the specified loaded area:

stress = force / area
circular dolly area = π × diameter² / 4
1 N/mm² = 1 MPa

Original arithmetic example: an ideal 20 mm circular loaded area is approximately 314.16 mm². A 1570.8 N force corresponds to 5.00 MPa. At the same force, a 10 mm diameter area gives 20.00 MPa because its area is one quarter as large. This calculation illustrates the importance of the selected dolly diameter; it does not predict that physically testing the same coating with different dollies will yield equivalent strength. Material behavior and test geometry also change.

As a bounded equipment example, the Elcometer 510 datasheet lists maximum ranges of 25 MPa with a 20 mm dolly and 4 MPa with a 50 mm dolly, with stated accuracy of ±1% of full scale. These are instrument specifications, not coating acceptance limits or complete uncertainty statements. Check the actual configuration and calibration range. Force, hydraulic-system pressure and calculated coating stress must not be confused. Elcometer 510 datasheet

Prepare and bond the dolly reproducibly

Dolly contamination can produce a weak glue interface before the intended coating system is tested. DeFelsko's preparation study distinguishes degreasing, abrasion and removal of loose debris. It studied disposable aluminum dollies and calls for manufacturer-specific preparation instructions; it does not establish one abrasive treatment for every dolly material or reusable fixture. Prevent handling contamination after preparation and identify the dolly type in the record. DeFelsko: dolly preparation

For its PosiTest AT-A, DeFelsko instructs users to check that abrasion, cleaning agents and glue do not change the coating. Its manual cautions that coating abrasion can introduce flaws. Mix and cure the bonding adhesive according to its manufacturer's instructions and support the fixture during cure where needed. An adhesive's handling time is not automatically sufficient cure for a tensile test. PosiTest AT-A instruction manual, preparation and application

Coating cure and glue cure are separate conditions. For concrete coatings, D7234 notes that required coating cure time should come from the coating manufacturer and may need extension for site temperature or humidity. Record the actual conditioning rather than applying an assumed universal waiting period. ASTM D7234-22: coating cure

Scoring, alignment and loading

Cutting around a dolly isolates the intended test area, but it can also introduce microcracking. The PosiTest AT-A manual makes the decision dependent on the controlling standard, specification or contract. Record whether the coating was cut and the extent of isolation. Before pulling, verify dolly size, pull rate, limit and any hold time in the instrument. Keep hands clear of the loaded coupling and actuator as directed by its instructions. PosiTest AT-A manual, isolation and testing

Use an arrangement that applies the intended concentric tensile load without unintended bending. Maintain the agreed apparatus and procedure across a comparison series. Changing scoring or loading conditions to obtain a higher result changes the test; it does not demonstrate an improvement in the coating. ASTM D4541-22: comparability of results

Interpret both fracture surfaces

Examine the dolly face and the corresponding surface. Record mixed failures by estimated area and identify the affected layers or interfaces:

DeFelsko describes these distinctions and recording fracture percentages. The stress number and fracture description belong together; “adhesion = 5 MPa” alone loses the essential location information. DeFelsko: adhesion testing and failure modes

Glue failure below a required level does not automatically prove the coating failed that requirement. Conversely, glue failure above a specified minimum may support a lower-bound statement for that individual test when the controlling procedure permits it. Elcometer explicitly discusses this distinction. Apply the standard's validity and retest rules rather than converting manufacturer summaries into universal percentage thresholds. Elcometer 510 datasheet: assessment of the adhesion test

Acceptance and inspection record

Define required strength, permitted failure modes, sampling locations, replicate count, aggregation and retest rules before testing. A test stopped at 5 MPa without fracture demonstrates survival of that test load under the specified conditions; it does not determine ultimate fracture strength. Retain individual results and distinguish fractures, load-limit terminations and indeterminate tests rather than averaging them as equivalent measurements.

Record substrate and coating system, cure and environment, method and edition, instrument identity and verification, dolly geometry, glue and cure, surface preparation, scoring, loading rate, peak stress or sustained load, termination reason, fracture percentages and photographs. State the acceptance criterion separately and arrange repair of damaged test locations.

Sources were reviewed November 1, 2026. The cited ISO edition is June 2023. The DeFelsko manual footer identifies IATA-Z.v.7.0-0225, copyright 2025. The Elcometer PDF carries copyright 2026 without a separate revision date. Product ranges and instructions remain specific to the cited equipment.

Inspection and quality documents