Pull-off adhesion strength analysis is the most widely used quantitative method for evaluating the bond strength between coatings, platings, or thermal spray layers and their substrates. Performed in accordance with internationally recognized standards such as ASTM D4541, ISO 4624, and GB/T 5210, the pull-off method measures the tensile stress required to detach a coating, reporting results in megapascals (MPa) along with a detailed failure mode classification. This article provides a comprehensive technical overview of the pull-off method, covering its testing principle, sample requirements, test items, procedures, quality control applications, and frequently asked questions.
Technical Overview
The pull-off method, also referred to as the detachable adhesion test or tensile pull-off test, is a destructive testing technique used to quantify the adhesion strength of a coating system to its substrate, or the cohesion between successive coating layers. In this method, a test dolly (loading fixture, typically 20 mm in diameter) is bonded to the coating surface using a high-strength two-component epoxy adhesive. After the adhesive has fully cured, a portable adhesion tester applies a steadily increasing tensile load perpendicular to the coating surface until failure occurs. The maximum stress at the moment of fracture is recorded as the pull-off strength.
The pull-off strength is calculated as the maximum tensile force divided by the effective test area of the dolly, expressed in MPa (N/mm²) or psi. Because the measured value represents the weakest point in the entire system—adhesive, coating layers, or the coating-substrate interface—the method not only provides a numerical strength value but also reveals where failure actually occurs, which is essential for diagnosing coating system performance.
- Quantitative result: pull-off strength in MPa or psi
- Failure mode identification: adhesive, cohesive, or glue failure, reported with area percentages
- Portable instrumentation allows both laboratory and on-site (field) testing
- Governed by major standards: ASTM D4541, ISO 4624, GB/T 5210, ASTM D7234, and EN 1542
Testing Samples
Organic Coating Systems
Liquid-applied and powder coatings such as epoxy, polyurethane, acrylic, alkyd, chlorinated rubber, and zinc-rich primers applied to steel, aluminum, galvanized steel, and other metal substrates. These samples include multi-layer anti-corrosion systems commonly found on pipelines, storage tanks, structural steel, and marine equipment. The pull-off method is particularly suitable for coating systems with dry film thickness typically above approximately 250 micrometers, where adhesive penetration into the film is not a concern.
Metallic and Thermal Spray Coatings
Metallic layers including electroplated coatings, hot-dip galvanized layers, and thermal spray coatings such as flame-sprayed or arc-sprayed zinc, aluminum, and zinc-aluminum alloys used for corrosion protection, as well as thermal barrier coatings (TBC) applied in high-temperature service. For thermal spray coatings on rigid substrates, the laboratory tensile method described in ASTM C633 is frequently applied as a complementary technique, while portable pull-off testers are used for field verification of sprayed metal coatings.
Concrete and Cementitious Substrates
Protective coatings, waterproofing membranes, and polymer overlays applied to concrete structures such as bridges, parking decks, floors, and wastewater treatment facilities. Testing on concrete is performed according to ASTM D7234 or EN 1542, and often reveals whether failure occurs within the concrete substrate itself (a cohesive substrate failure), which indicates that the measured value is limited by substrate tensile strength rather than coating adhesion.
Testing Items
A complete pull-off adhesion analysis comprises the following evaluation items:
- Pull-off adhesion strength (MPa or psi): the maximum perpendicular tensile stress the coating system withstands before detachment
- Failure interface identification: determination of the exact plane at which fracture occurred (substrate/coating interface, within a coating layer, between coating layers, or within the adhesive)
- Failure mode area distribution: percentage of each fracture surface attributed to each interface or layer, reported per the conventions of ASTM D4541 and ISO 4624
- Validity assessment of each measurement: evaluation of glue failures, partial fractures, and alignment quality to confirm test validity
- Statistical evaluation: individual values, mean or median values, and scatter of replicate measurements
Testing Methods
Sample and Dolly Preparation
Test surfaces must be clean, dry, and free of grease, dust, and loose contamination. The coated specimens are conditioned, where practical, at a standard laboratory atmosphere of 23 ± 2 °C and 50 ± 5% relative humidity. The mating face of each aluminum or steel dolly is abrasively blasted and degreased immediately before bonding to ensure strong adhesion of the epoxy. Dolly diameter is selected according to the expected strength level: standard 20 mm dollies are suitable for most organic coatings, while larger dollies (e.g., 50 mm) may be used for low-strength or concrete coatings and smaller dollies for very high-strength systems.
Adhesive Bonding and Curing
A two-component epoxy adhesive with a cohesive and adhesive strength demonstrably higher than the expected coating adhesion is applied to the dolly face, and the dolly is pressed firmly onto the coated surface. Excess adhesive is removed, and the assembly is allowed to cure for the full period recommended by the adhesive manufacturer, typically 24 hours at ambient temperature. For coating systems with adhesion above approximately 20 MPa, high-strength adhesives are required; if the adhesive itself fails before the coating detaches, the result is reported as a lower-bound value limited by glue strength.
Perimeter Cutting
After curing, a circumferential cut is made around the dolly through the full thickness of the coating down to the substrate using a hole cutter or drilling guide. This step isolates the test area to the dolly footprint and eliminates the contribution of surrounding coating, which would otherwise artificially inflate the measured strength. Cutting must be performed carefully to avoid micro-cracking at the cut edge, which could prematurely initiate failure.
Pull-Off Loading
The dolly is coupled to a portable hydraulic, pneumatic, or mechanical self-reacting adhesion tester. In accordance with ISO 4624 and GB/T 5210, the tensile load is applied smoothly and continuously with a rate of stress increase not exceeding 1 MPa/s until fracture occurs. The peak force is recorded automatically or read from the pressure indicator and converted to stress using the effective dolly area. A minimum of three valid determinations—five or more in many project specifications—is performed on each sample or structure, spaced sufficiently apart to avoid mutual influence.
Data Recording and Evaluation
For each measurement, the recorded pull-off strength, the failure interface, and the estimated percentage of each failure type are documented. Failure modes are classified by identifying the fracture surfaces on both the dolly and the substrate, for example: adhesive failure at the substrate/first-coat interface, cohesive failure within a coating layer, adhesive failure between coats, cohesive failure of the adhesive, or mixed modes. The individual values and the average (or median) are reported together with the failure mode, because a numerically high value accompanied by glue failure may not represent true coating adhesion.
Applications
Pull-off adhesion analysis is applied extensively across industries as a quality control, qualification, and failure-analysis tool:
- Oil and gas pipelines: verification of adhesion of fusion-bonded epoxy (FBE), three-layer polyethylene systems, and field-applied repair coatings, with project specifications frequently requiring values of 10 MPa or higher
- Marine and offshore structures: acceptance testing of anti-corrosion coating systems in accordance with ISO 12944, where pull-off values of at least 5 MPa are commonly specified for high-corrosivity environments
- Bridge and infrastructure steelwork: confirmation of adhesion of zinc-rich primer systems and multi-coat protective systems after surface preparation and application
- Thermal spray metallizing: field and laboratory verification of bond strength of sprayed zinc and aluminum coatings on steel bridges and components
- Aerospace and power generation: evaluation of thermal barrier coatings, sealing coatings, and other engineered layers on turbine components
- Concrete structures: assessment of coating and overlay adhesion on floors, parking decks, and wastewater infrastructure per ASTM D7234 and EN 1542
- Coating qualification and dispute resolution: comparative testing in accredited third-party laboratories to evaluate coating formulations, surface preparation methods, or contractor workmanship
- Failure analysis: determination of whether field delamination originates from inadequate surface preparation, intercoat contamination, under-curing, or substrate-related weakness
FAQ
What is the difference between the pull-off test and the cross-cut test?
The cross-cut or cross-hatch test (ASTM D3359, ISO 2409) is a qualitative or semi-quantitative method that rates adhesion on a 0–5 scale by cutting a lattice pattern and applying tape, and it is suited to coatings up to approximately 250 micrometers thick. The pull-off method is a fully quantitative technique that reports a strength value in MPa and identifies the failure plane, making it the preferred method for thick anti-corrosion coatings, project acceptance testing, and engineering specifications.
What does it mean when the glue fails instead of the coating?
A glue (adhesive) failure means the epoxy bond to the coating was weaker than the coating adhesion itself. In this case the measured value is a lower bound: the true adhesion of the coating is at least as high as the recorded figure. When glue failure predominates and a higher value is required, a stronger adhesive, improved dolly surface preparation, or the laboratory tensile method of ASTM C633 for rigid specimens should be considered.
How is the failure mode reported?
Following ASTM D4541 and ISO 4624 conventions, the failure location and the approximate percentage of the test area are reported, for example a mixed result indicating 60% adhesive failure at the substrate/primer interface and 40% cohesive failure within the primer. This documentation is as important as the strength value itself because it identifies the weakest element of the coating system.
How many measurements are required?
ISO 4624 and GB/T 5210 require a minimum of three valid determinations per test area, and many project specifications require five or more to improve statistical confidence. Values from tests invalidated by glue failure, misalignment, or edge cracking are excluded and replaced with additional measurements.
Can the pull-off method be performed on curved surfaces such as pipes?
Yes. For pipes and other curved surfaces, dollies with a matching contoured radius or special alignment fixtures are used to ensure axial loading. Careful alignment is essential, because angular misalignment introduces peel stresses that underestimate the true adhesion strength.
What are typical acceptance criteria?
Acceptance values are defined by the project specification rather than the test standard itself. Commonly encountered requirements include at least 5 MPa for structural steel coating systems in aggressive environments per ISO 12944-based specifications, 10 MPa or higher for pipeline FBE coatings, and at least 10.3 MPa (1500 psi) for many thermal spray coating applications. Where the pull-off value exceeds the specification limit with cohesive failure inside the substrate or glue, the coating is considered compliant.