Adhesion Testing

An adhesion test quantifies the bond strength within a coating system by applying a controlled tensile stress perpendicular to the coated surface until detachment occurs. The test does not measure a single material property; it measures the weakest interface or layer within the complete system of substrate, primer, intermediate coats, topcoat, and the adhesive used to bond the test dolly.

The result is reported as a pull-off stress in megapascals (MPa) or, less commonly, in pounds per square inch (psi). This value represents the maximum stress sustained before failure, together with a description of where failure occurred. Both pieces of information are required for a meaningful result: a high pull-off value with failure in the adhesive (glue) is not equivalent to the same value with failure at the primer–substrate interface.

Adhesion testing does not assess long-term durability, fatigue resistance, or performance under service conditions. It provides a point-in-time measurement of bond strength under a specific loading condition.

Adhesion testing is applied wherever the performance of a coating or bonded layer depends on its attachment to the substrate. Common application areas include:

• Protective coatings on steel structures (bridges, tanks, pipelines, marine vessels)
• Industrial paint systems in manufacturing and maintenance
• Concrete repair and overlay systems (bonded screeds, renders, coatings)
• Fire protection coatings (intumescent systems on structural steel)
• Flooring systems (epoxy, polyurethane, and resin floor coatings)
• Thermal spray and metallised coatings
• Aerospace and automotive primer and paint systems

In each case, adhesion testing serves as a quality control or acceptance tool: confirming that the coating system meets a specified minimum bond strength before the asset is placed into service or handed over.

The standard pull-off adhesion test involves bonding a cylindrical test dolly (also called a stump or pull stub) to the coating surface using a structural adhesive, allowing the adhesive to cure, and then applying a tensile load perpendicular to the surface until the dolly detaches. The force at detachment is divided by the dolly contact area to calculate the pull-off stress.

The test can be performed with or without scoring (cutting) around the dolly before pulling. Scoring isolates the test area from the surrounding coating, ensuring that the measured stress relates to the defined area rather than being influenced by the tensile strength of the adjacent film. Whether scoring is required depends on the applicable standard and the coating system.

The load is applied using a portable adhesion tester that grips the dolly and pulls at a controlled rate. The rate of loading and the alignment of the pull relative to the surface both influence the result. A pull that is not perpendicular introduces shear and peel components that reduce the apparent adhesion value and change the failure mode.

For a detailed explanation of what is physically measured and how typical outputs should be read, see the related knowledge base article: Pull-off Adhesion Test Principles.

The reliability of a pull-off adhesion test depends critically on how the test area is prepared and how the dolly is bonded. If the adhesive bond between the dolly and the coating is weaker than the coating–substrate bond, the test measures the glue joint rather than the system of interest. This is the single most common cause of invalid or misleading adhesion test results.

In practice that means a clean, dry, contamination-free bonding area — lightly abraded on smooth or glossy coatings — an adhesive cured to a strength above the expected coating adhesion, and, where the standard requires it, scoring around the dolly down to the substrate to define the test area precisely.

For full guidance on preparation, bonding, and curing effects, see the related knowledge base article: Surface Preparation and Dolly Bonding.

The failure mode describes where detachment occurred within the coating system. It is at least as important as the pull-off stress value, because it identifies which interface or layer is the weakest plane.

Failure modes are classified as:

Adhesive failure (at an interface) — detachment occurs at the boundary between two layers (e.g., primer to substrate, topcoat to primer).

Cohesive failure (within a layer) — the fracture passes through the body of a single layer (e.g., within the primer itself), indicating that the layer’s internal strength is the limiting factor.

Substrate failure — the substrate material itself fractures (common in concrete, where the surface concrete may be weaker than the coating bond).

Glue failure — detachment occurs at the adhesive used to bond the dolly, meaning the test has not measured the coating system and the result is typically invalid.

Most real pull-off failures are mixed: the fracture surface shows areas of different failure types. Reporting should include estimated percentages of each mode across the dolly face.

For detailed guidance on identifying, classifying, and interpreting failure modes, see the related knowledge base article: Adhesion Test Failure Modes.

Adhesion testing is governed by several international standards that define the test method, apparatus, specimen preparation and reporting format, and they differ in detail on scoring, pull rate, number of tests per area and acceptance criteria. What they share is a reporting expectation: pull-off stress, failure mode, dolly size, adhesive type, surface preparation and any deviation from the procedure — incomplete reporting is a frequent audit finding. Identify the applicable standard before testing rather than applying a generic method.

For a detailed breakdown of standard requirements and reporting obligations, see the related knowledge base article: Adhesion Testing Standards and Reporting.

Portable adhesion testers are mechanical instruments that apply and measure tensile force. Like all force measurement instruments, they require periodic verification to confirm that the displayed force (and therefore the calculated stress) is accurate.

Verification typically involves loading the instrument against a reference device (such as a proving ring or calibrated load cell) and comparing the displayed value with the known reference. This checks the complete measurement chain, including the hydraulic or mechanical actuator, the pressure transducer or load cell, and the display or readout.

Factors that affect confidence in adhesion test results extend beyond instrument calibration:

• Dolly area accuracy (the stress calculation assumes a known contact area)
• Pull rate consistency (rate affects measured adhesion for some coating systems)
• Alignment of the pull relative to the surface
• Number of tests performed (single tests have high uncertainty; replicate tests improve confidence)

For verification procedures and guidance on assessing measurement confidence, see the related knowledge base article: Verification & Checks for Adhesion Testers.

Adhesion test results are influenced by many variables beyond the instrument itself. Common sources of error and variability include:

• Inadequate surface cleaning or preparation before bonding the dolly
• Insufficient adhesive cure time, especially in cold or humid conditions
• Glue contamination of the coating–substrate interface during scoring
• Misalignment of the tester, introducing peel or shear components
• Inconsistent pull rate between tests or operators
• Testing on areas that are not representative of the overall coating condition
• Reporting pull-off stress without recording or classifying the failure mode
• Using dollies of incorrect size or with damaged bonding faces

Many of these errors produce results that appear plausible but are systematically biased. The most reliable way to detect them is to examine the failure surface carefully after every pull and to compare replicate tests for consistency.