Corrosion Protection for Industrial Coating Contractors
Article 17: Coating Adhesion Testing and Failure Interpretation
Understanding Not Only the Test Result, but Where and Why the System Separated
The Big Idea
Adhesion testing does more than produce a number or rating. A useful test identifies how much force the system resisted, where separation occurred, how the test was performed, and whether the result represents the coating system.
What Is Coating Adhesion?
Adhesion is the ability of a coating to remain attached to the substrate or to another coating layer. A multi-coat system contains several possible bonding locations.
These locations may include:
- The bond between primer and substrate
- The bond between primer and intermediate coat
- The bond between intermediate and finish coats
- The internal strength of each coating layer
A coating can remain strongly attached to the steel while separating between coats. It can also remain bonded at every interface while breaking internally within one coating layer.
This is why the location and type of separation are as important as the reported test value.
Field Rule
Never report an adhesion number without reporting where the system separated.
Adhesive and Cohesive Failure
Adhesive Failure
Adhesive failure occurs at the interface between two different materials. Examples include separation between the primer and steel or between two coating layers.
Adhesive failure at the substrate may be associated with contamination, inadequate surface preparation, unsuitable profile, moisture, improper wetting, incompatible coating, or another condition at the interface.
Adhesive failure between coats may be associated with contamination, an exceeded recoat interval, improper cure, chalk, amine blush, zinc salts, an overly smooth surface, or incompatible coating layers.
Cohesive Failure
Cohesive failure occurs within one material rather than at an interface. The coating layer splits internally while part remains attached to each side of the separation.
Cohesive failure may indicate that the internal strength of that layer was lower than the strength of the surrounding bonds. The result must still be compared with the project acceptance requirement and expected coating behavior.
Glue Failure
In pull-off testing, the adhesive used to attach the loading fixture may fail before the coating system. This is commonly called glue failure.
A glue failure may show that the coating resisted at least the force reached before the adhesive separated, but it may not reveal the coating system's actual maximum pull-off strength. The result must be reported and evaluated according to the specified procedure.
Substrate Failure
On some substrates, the substrate itself may fail before the coating bond. This is more common with concrete, wood, weak metal surfaces, corrosion products, or deteriorated material than with sound structural steel.
Why Adhesion Testing Is Performed
Adhesion testing may be required to:
- Qualify a proposed coating system
- Evaluate a test patch or mock-up
- Verify new coating application
- Evaluate an existing coating before overcoating
- Investigate a suspected coating failure
- Compare preparation or application procedures
- Support a repair or maintenance decision
The purpose should be established before testing begins. A test method appropriate for evaluating an existing maintenance coating may not provide the information needed to qualify a new high-performance lining.
Three Common Field Test Approaches
Common field approaches include knife testing, tape testing, and pull-off testing. These methods apply different forces and produce different types of results.
Their ratings and results should not be treated as interchangeable.
| Method |
General Result |
Damage to Coating |
| Knife Test |
Qualitative or classified evaluation of resistance to lifting |
Destructive at the test cut |
| Tape Test |
Visual classification based on coating removed by tape |
Cuts and tape may damage the test area |
| Pull-Off Test |
Tensile pull-off strength and fracture description |
Normally destructive when tested to fracture |
Knife Adhesion Testing
Knife testing evaluates how readily the coating can be lifted from the substrate or separated between layers after controlled cuts are made.
The operator uses a sharp knife and follows the cut geometry and evaluation procedure required by the referenced method.
Knife testing can be useful for field evaluation because it requires limited equipment and can help identify the plane of separation.
However, the test involves operator judgment. Knife sharpness, cut depth, angle, force, coating thickness, coating hardness, substrate shape, and operator technique can affect the result.
Tape Adhesion Testing
Tape testing uses controlled cuts through the coating followed by application and removal of specified pressure-sensitive tape. The amount and pattern of coating removal are compared with the classification criteria of the test method.
The governing procedure determines the cut pattern, spacing, tape, application, removal, timing, and rating.
Tape testing is often more suitable for relatively thin coatings on smooth, flat substrates. Thick, tough, flexible, textured, reinforced, or highly irregular coating systems may be difficult to cut and evaluate reliably by this method.
Variables Affecting Tape-Test Results
- Cut depth and spacing
- Sharpness and condition of the cutting tool
- Coating thickness and hardness
- Tape type, age, storage, and adhesion
- Pressure used to apply the tape
- Time between applying and removing the tape
- Removal angle and speed
- Surface texture and curvature
Pull-Off Adhesion Testing
Pull-off testing uses a loading fixture, commonly called a dolly, bonded to the coating. A portable tester applies tensile force approximately perpendicular to the surface until the system separates or reaches a specified proof load.
The test can provide a numerical pull-off value, but the value is only one part of the result. The fracture surfaces must also be examined and reported.
Pull-off strength is influenced by the coating system, substrate, adhesive, dolly size, tester design, cure, temperature, humidity, scoring, loading rate, surface geometry, and test procedure.
Field Rule
Pull-off results obtained with different tester designs or different procedures should not be compared as though they were produced by the same test.
Preparing a Pull-Off Test Location
The selected area should be representative, accessible, sufficiently flat for the loading fixture, and large enough for the tester to operate without interference.
The coating surface and dolly face must be prepared according to the procedure and adhesive instructions. Preparation may include cleaning and controlled abrasion.
Abrasion should improve the glue bond without cutting through the coating or weakening the system being evaluated. Dust, oil, fingerprints, moisture, and cleaning residue must be removed.
Selecting and Applying the Adhesive
The adhesive attaching the dolly must be strong enough to exceed the expected coating-system strength. It must also be compatible with the coating surface and cure under the test conditions.
Too much adhesive can create a thick, uneven bond line. Too little can leave voids beneath the fixture. Excess adhesive around the dolly may interfere with scoring or accidentally bond the fixture to surrounding coating.
The fixture should remain stable and aligned while the adhesive cures. Follow the adhesive manufacturer's requirements for mix ratio, application, cure time, temperature, and handling.
Adhesive Cure Matters
A dolly should not be pulled simply because the adhesive feels hard. The adhesive must reach the required cure under the actual temperature and environmental conditions.
Testing too early can produce glue failure and a misleadingly low result. Cold surfaces can substantially extend adhesive cure.
Scoring Around the Dolly
Scoring means cutting through the coating around the perimeter of the dolly before applying the pull force.
Scoring isolates the test area from the surrounding coating, but it can also introduce damage, stress concentration, heat, vibration, or microcracking.
Whether scoring is required or permitted should be established before testing. Scored and unscored results should not be combined without recognizing the procedural difference.
Test-to-Fracture and Pass-Fail Testing
Test to Fracture
Testing to fracture identifies the force reached when the weakest part of the test assembly separates. It also exposes the fracture surfaces for evaluation.
Pass-Fail or Proof Testing
In proof testing, the load is increased to a specified value. If the system remains intact at that load, the test satisfies the defined criterion without deliberately pulling the system to failure.
The specification should identify the required protocol. The contractor should not decide during the test whether to stop or continue based on the displayed number.
Applying the Pulling Force
The tester must be aligned so the load is applied concentrically and approximately perpendicular to the surface. Side loading can introduce peeling or shear forces that change the result.
The load should be applied at the rate required by the test procedure. Pulling too quickly or too slowly can affect the measured value.
Reading the Fracture Surface
After separation, examine both the dolly and the corresponding test location. Together, these two surfaces show where the system failed.
The fracture may occur entirely at one interface or across several locations. A result might include a combination of glue failure, cohesive coating failure, intercoat separation, and substrate-interface separation.
Estimate and report the percentage of each failure type according to the required procedure. Photographs of the dolly and test surface can strengthen the report.
The Highest Number Is Not Always the Most Useful Result
A test in which the dolly adhesive fails at a relatively high load may demonstrate that the coating resisted at least that load. It does not necessarily identify the coating system's actual breaking strength.
Another test may produce a lower value but clearly reveal separation between two coating layers. That fracture location may provide more useful diagnostic information.
Numerical results must always be interpreted together with the fracture description and test conditions.
Coating Cure and Test Timing
Coating strength changes as the system cures. Testing too early may produce low values or internal failure that would not occur after full cure.
The specification should identify the required cure condition or test age. Record the application date, test date, curing conditions, and any exposure occurring between application and testing.
Choosing Representative Test Locations
Test locations should represent the coating work and the purpose of the evaluation. Testing only easy, flat, favorable areas may produce a result that does not represent the structure.
Location selection may consider:
- Different application areas and work shifts
- Different substrate conditions
- Separate coating batches
- Areas with different environmental exposure
- Shop-applied and field-applied coating
- Repairs and original coating
- Suspected failure and apparently sound control areas
How Many Tests Are Required?
One pull does not characterize an entire structure. Coating systems and field conditions naturally vary.
The project specification should define the number of tests, test areas, locations, acceptance criteria, treatment of invalid tests, and method of calculating results.
Do not average unrelated tests from different coating systems, substrates, instruments, procedures, or conditions.
Interpreting a Low Result
A low result does not automatically identify the cause. First determine whether the test was valid and where separation occurred.
Investigate:
- Was the coating fully cured?
- Did the dolly adhesive cure properly?
- Was the tester aligned and operated correctly?
- Was the required loading rate used?
- Was scoring performed correctly?
- Did failure occur in the glue, coating, interface, or substrate?
- Does the test area represent the surrounding work?
- Do nearby tests show the same behavior?
Field Rule
Repeating a failed test without investigating the cause does not turn a failure into acceptable work.
Adhesion Testing Existing Coatings
Adhesion testing can help evaluate whether an existing coating may be suitable for overcoating, but it is only one part of the condition assessment.
An existing coating may show acceptable adhesion in selected areas while containing widespread brittleness, hidden corrosion, excessive thickness, contamination, or weak layers elsewhere.
An overcoating evaluation should also consider:
- Coating type and history
- Number of existing layers and total thickness
- Rusting and underfilm corrosion
- Blistering, cracking, checking, and delamination
- Compatibility with the proposed maintenance coating
- Results from representative test patches
Repairing Adhesion Test Locations
Knife cuts, tape-test cuts, scored circles, and pull-off sites damage the coating system. Test locations must be repaired when required by the specification or service environment.
A repair procedure may require:
- Removal of adhesive and test residue
- Removal of loose or damaged coating
- Preparation of exposed steel
- Feathering adjacent sound coating
- Application of approved repair materials
- Restoration of the required system thickness
- Cure, inspection, and holiday retesting when applicable
Common Adhesion-Testing Mistakes
- Using an unspecified method: The result does not match the acceptance requirement.
- Testing before full cure: The coating has not developed its expected strength.
- Poor dolly preparation: The adhesive fails at the fixture.
- Improper scoring: The test area is damaged before loading.
- Side loading the tester: Peeling or shear forces distort the result.
- Ignoring the loading rate: Results cannot be compared reliably.
- Reporting only the numerical value: The failure plane is unknown.
- Calling glue failure coating failure: The wrong material is blamed.
- Comparing different tester types: Results from different methods are treated as equivalent.
- Leaving test sites unrepaired: The inspection creates future corrosion sites.
Adhesion-Test Documentation
A complete test report may include:
- Project, structure, and test-location identification
- Purpose of the test
- Coating system, coat sequence, and thickness
- Coating application and test dates
- Test standard, method, and protocol
- Tester manufacturer, model, and identification
- Dolly size and adhesive
- Surface and dolly preparation
- Whether the coating was scored
- Environmental and surface conditions
- Pull-off value or test rating
- Failure location and estimated percentages
- Photographs of both fracture surfaces
- Acceptance decision and corrective action
- Repair of the test location
- Name or initials of the person performing the test
Contractor's Adhesion-Test Checklist
- The purpose and required test method are defined.
- Acceptance criteria are established before testing.
- The coating has reached the required cure condition.
- Test locations represent the work.
- The test instrument is suitable and within its required calibration status.
- The correct dolly, adhesive, cutting tool, or tape is available.
- Surface preparation and adhesive cure follow the approved procedure.
- Scoring requirements are understood.
- The required loading procedure and rate are followed.
- Both fracture surfaces are examined.
- The value, failure plane, and failure percentages are documented.
- Questionable or low results are investigated rather than ignored.
- All test locations are repaired as required.
Key Takeaways
- Adhesion can be evaluated at the substrate and between coating layers.
- Adhesive failure occurs at an interface; cohesive failure occurs within a material.
- Knife, tape, and pull-off testing produce different types of results.
- Pull-off strength depends on the coating and the complete test procedure.
- Glue failure does not reveal the coating system's maximum strength.
- Tester type, dolly size, adhesive, scoring, cure, and loading rate affect results.
- A numerical value must be reported with its fracture location.
- Adhesion tests usually damage the coating and require repair.
Bottom Line
An adhesion test does not deliver one simple answer. A responsible evaluation considers the method, procedure, numerical result, failure location, coating cure, test conditions, and whether the selected locations represent the work.
Technical References
The following industry standards provide the formal procedures associated with the adhesion-testing methods discussed in this article:
These references establish formal testing procedures. The project specification, coating-manufacturer instructions, and referenced edition of each standard govern the actual work.
Knowledge Check
1. What is the difference between adhesive and cohesive failure?
View Answer
Adhesive failure occurs at the interface between different materials. Cohesive failure occurs within one material.
2. Why is a pull-off value incomplete without a fracture description?
View Answer
The value shows the force reached, but the fracture description identifies whether separation occurred in the glue, coating, between coats, at the substrate, or within the substrate.
3. Does glue failure show the coating system's maximum pull-off strength?
View Answer
Not necessarily. It may show that the coating resisted at least the force reached before the adhesive failed.
4. Why should results from different portable pull-off tester designs not be compared directly?
View Answer
Tester design and procedure can affect the stress applied and the resulting value. Comparisons require consistent equipment and test conditions.
5. Why can testing a coating too early produce a misleading result?
View Answer
The coating may not have developed its expected cure, internal strength, or adhesion.
6. What should happen to a destructive adhesion-test location?
View Answer
The damaged coating should be repaired using the approved coating-system repair procedure and inspected before the structure enters service.
Coming Next
Article 18: Coating Defects and Failure Diagnosis
The next article examines blistering, delamination, cracking, rusting, pinholes, solvent entrapment, dry spray, amine blush, underfilm corrosion, and the evidence needed to separate symptoms from root causes.
