Pull-Off Adhesion Test Result Evaluation: Standards, Failure Modes and Acceptance Criteria

The pull-off (tensile) method is one of the most widely used quantitative techniques for measuring the adhesion strength of coatings, paints, and overlays on metallic and non-metallic substrates. Accurate evaluation of pull-off test results requires not only correct strength calculation but also systematic failure mode classification in accordance with standards such as ISO 4624, ASTM D4541, and GB/T 5210. This article explains how to prepare samples, perform the test, interpret failure patterns, and judge whether reported adhesion values are valid and compliant.

Technical Overview

The pull-off method, also known as the tensile adhesion test, quantifies the adhesion of a coating system by applying a perpendicular tensile force to a test stub (dolly) bonded to the coating surface with a high-strength adhesive. The load is increased gradually until detachment occurs, and the maximum stress at failure is calculated as the peak force divided by the bonded area of the dolly, expressed in MPa or psi. The method is governed by internationally recognized standards including ISO 4624, ASTM D4541, ASTM C1583, EN 1542, and GB/T 5210, and is routinely performed by accredited third-party laboratories for quality control, product qualification, and failure analysis.

A critical point that distinguishes professional result evaluation from simple data reporting is failure mode analysis. The measured stress value is only meaningful when interpreted together with the exact location and distribution of the fracture surface. A test in which the adhesive glue itself fails, rather than the coating or the coating-substrate interface, provides only a lower-bound indication of the true adhesion strength and must be flagged accordingly in the test report.

Testing Samples

Metallic Substrates with Organic Coatings

Steel panels, aluminum alloy sheets, galvanized steel, and fabricated metal components protected by primers, liquid paints, powder coatings, or epoxy systems. These samples are the most common subject of pull-off testing in the industrial, automotive, and marine coating sectors, and are evaluated primarily against ISO 4624 or ASTM D4541.

Thermal Spray and Metallic Coatings

Zinc, aluminum, and zinc-aluminum sprayed coatings, as well as ceramic or cermet coatings applied by flame or plasma spraying on structural steel and machine parts. The relatively high thickness and inherent porosity of these coatings demand careful dolly preparation and adhesive selection.

Concrete and Masonry Surfaces

Floor coatings, traffic membranes, protective liners, cementitious repair mortars, and polymer-modified overlays applied to concrete substrates. Testing is typically conducted per ASTM C1583, ASTM D7234, or EN 1542, and result evaluation must account for the weak, heterogeneous nature of the concrete surface layer.

Plastics, Composites, and Multi-Layer Systems

Coated plastic housings, fiber-reinforced panels, and multi-coat paint systems in which intercoat adhesion between primer, basecoat, and clearcoat is the property of interest. For these samples, evaluation focuses on identifying at which interface the separation occurred.

Testing Items

A complete pull-off adhesion evaluation covers the following items:

• Pull-off adhesion strength of the coating, reported in MPa or psi to the precision required by the standard (typically 0.1 MPa).
• Failure mode classification according to the standardized nomenclature, identifying whether failure occurred within the substrate, at the substrate-coating interface, cohesively inside a coating layer, between coats, or within the adhesive.
• Percentage distribution of each failure type over the fracture surface, for example a 60% interfacial failure at the primer-steel interface combined with 40% cohesive failure of the primer.
• Statistical treatment of replicate results, including mean value, range, standard deviation, and the number of valid determinations.
• Validity assessment of each individual result based on the location of fracture and the condition of the dolly and specimen.

Testing Methods

Specimen Preparation and Adhesive Bonding

Test surfaces are cleaned to remove dust, oil, and loose material without damaging or stressing the coating. Aluminum dollies of defined diameter (commonly 20 mm for ISO 4624 or 14 mm to 50 mm depending on expected strength and tester capacity for ASTM D4541) are abraded, degreased, and bonded to the coating using a two-component epoxy or acrylic adhesive with a cohesive strength significantly exceeding the expected adhesion value. Specimens are conditioned and cured under controlled environment, typically 23 ± 2 °C and 50 ± 5% relative humidity, for the full curing period recommended by the adhesive manufacturer. Where required by the standard or specification, a circumferential cut is made through the coating down to the substrate around the dolly to isolate the test area; the influence of this cut on crack initiation must be considered during evaluation.

Test Execution

A portable or stationary pull-off adhesion tester applies a steadily increasing tensile load through a self-aligning actuator perpendicular to the specimen surface. The loading rate is controlled within the range specified by the applicable standard (for example, not exceeding 1 MPa/s per ISO 4624) until detachment occurs. The maximum applied force at the moment of fracture is recorded automatically or read from the calibrated instrument. At least five valid determinations are required for ASTM D4541, while ISO 4624 calls for at least six valid results, with more replicates recommended for heterogeneous or field surfaces.

Calculation of Pull-Off Strength

The pull-off strength is calculated by dividing the maximum tensile force at failure by the cross-sectional area of the dolly. For example, a fracture load of 3,140 N on a 20 mm diameter dolly yields a nominal stress of approximately 10.0 MPa. Reported values are rounded as specified in the standard, and the calculated stress is always presented together with the observed failure mode rather than as a standalone number.

Failure Mode Classification

Evaluation follows the standardized failure nomenclature of ISO 4624, in which each interface and layer is assigned a letter:

• Failure within the substrate, indicating that the measured value is governed by substrate cohesion rather than coating adhesion.
• Adhesive failure between the substrate and the first coat, the most direct measure of coating adhesion.
• Cohesive failure within a coating layer, indicating that internal strength of the coating, not interface adhesion, is the limiting property.
• Adhesive failure between two coats, relevant for multi-layer systems and intercoat compatibility assessment.
• Cohesive failure of the adhesive glue, meaning the bond between glue and coating exceeded the glue strength and the result represents a lower bound.

Each failure location is quantified as an estimated percentage of the dolly area, producing a complete fracture map for every replicate.

Validity and Data Evaluation

Results are considered valid only when fracture occurs at or below the glue-coating interface or within the coating system. If the adhesive fails over a substantial portion of the area, the recorded value is reported as a minimum, expressed as "greater than" the measured stress. Individual outliers caused by misalignment, partial bonding, or edge cutting defects are identified through fracture surface inspection and excluded from the mean with justification. The final evaluation compares the mean pull-off strength and dominant failure mode against the acceptance criteria of the product specification, contract, or qualification standard.

Applications

Pull-off adhesion result evaluation supports decision-making across a wide range of industries:

• Protective and marine coatings: verifying that steel structures, ships, and offshore facilities meet specified adhesion levels before and after exposure testing such as salt spray or cyclic aging.
• Pipeline coatings: qualification and routine quality control of fusion-bonded epoxy and liquid epoxy systems against project-specific minimum adhesion requirements.
• Architectural and infrastructure coatings: acceptance testing of concrete floor coatings, bridge overlays, and repair mortars where pull-off strength also reflects surface preparation quality.
• Automotive and appliance finishing: evaluation of pretreatment effectiveness, primer compatibility, and intercoat adhesion in multi-layer paint systems.
• Thermal spray coating shops: verification of sprayed zinc and aluminum coatings on steel structures in accordance with structural steel protection standards.
• Failure analysis and dispute resolution: comparing the adhesion performance and failure modes of complaint samples against retained reference panels in an objective, quantitative manner.

FAQ

What does it mean when the glue fails instead of the coating?

When fracture occurs within the adhesive, the true adhesion strength of the coating was not reached. The measured value is reported as a lower-bound result, conventionally written as "greater than" the recorded stress, and the failure mode percentage for glue failure is documented. If a higher strength value is needed, a stronger adhesive or alternative method must be used.

How many test points are required for a valid evaluation?

ASTM D4541 specifies a minimum of five measurements, and ISO 4624 requires at least six valid determinations for each surface or coating system being evaluated. Field surfaces with high variability typically warrant additional replicates to obtain representative statistics.

Can results from different dolly sizes be compared directly?

Not without caution. Larger dollies tend to produce lower average strength on heterogeneous surfaces because failure is triggered by the weakest point within a larger stressed area. Comparison is only appropriate when dolly diameter, adhesive, coating thickness, and test conditions are identical.

How is a pass or fail conclusion drawn from the results?

The mean pull-off strength of the valid determinations is compared with the minimum acceptance value defined in the product standard or project specification. The dominant failure mode is reviewed simultaneously; a system meeting the numeric requirement through substrate failure, for example, demonstrates that adhesion exceeds the strength of the substrate itself rather than a precise interface value.

How does pull-off testing differ from cross-cut adhesion testing?

Cross-cut or tape tests per ISO 2409 and ASTM D3359 provide a rapid qualitative rating, whereas the pull-off method delivers a quantitative stress value with defined failure location. Pull-off evaluation is preferred whenever numerical adhesion data is required for specification compliance or engineering calculation.