AirSprayTech Academy Certificate Program
Commercial and Industrial Roof Coatings | Article 13 of 25
Adhesion Testing, Test Patches, and Pull Testing
Adhesion testing answers more than “Did it stick?” A useful test verifies
the complete preparation and primer process, measures or compares bond
performance, identifies the weakest layer, and documents whether the
proposed system is suitable across the roof's different substrates and
conditions.
Test the Proposed Process—not Just the Product
A test patch should duplicate the work that will be performed on the
roof: cleaning, drying, abrasion when required, primer or activator,
coating or membrane, reinforcement when applicable, application rate,
cure, and test procedure.
A strong result on an unusually clean hand-prepared spot does not approve
a production process that uses different equipment, cleaner, primer
coverage, or drying time. The test is valid for the conditions it
represents.
Three Different Adhesion Approaches
| Approach |
Best Use |
Reported Result |
Important Limitation |
| Field peel test |
Flexible roof membranes and cured liquid-applied test strips
|
Peel force, often pounds per linear inch, plus failure mode
|
Field procedures vary; a manufacturer method is not
automatically full ASTM D903 compliance.
|
| ASTM D4541 pull-off test |
Coatings on suitable rigid substrates, commonly metal
|
Pull-off strength and fracture location |
Tester type, dolly preparation, adhesive, scoring, substrate
flexibility, and loading rate affect results.
|
| ASTM D7234 pull-off test |
Coatings on concrete |
Pull-off strength and failure plane within coating, interface,
or concrete
|
Concrete strength, surface preparation, moisture, dolly bonding,
and scoring affect the result.
|
Field Peel Adhesion Tests
A common roof-coating field test creates a reinforced coating strip with
a free pull tab. After the required cure, the strip is cut to a known
width and peeled at the specified angle and rate using an approved force
gauge or spring scale. The average or required force is reported in
pounds per linear inch or another specified unit.
ASTM D903-98(2025) is a laboratory test method for comparative peel or
stripping strength of adhesive bonds using defined specimens,
pretreatment, conditioning, temperature, and testing-machine speed. A
simplified rooftop peel test should be called ASTM D903 only when it
actually follows the applicable standard. Otherwise, identify it as the
named manufacturer field-peel procedure.
Typical Field-Peel Sequence
-
Select and identify a representative test location.
-
Prepare the substrate using the proposed production procedure.
-
Apply the specified cleaner, primer, coating, reinforcement, and
coverage.
-
Leave a usable reinforced pull tab where the procedure requires one.
-
Cure for the specified time under documented conditions.
-
Cut the strip to the specified width without damaging the substrate.
-
Peel at the specified angle, direction, and rate while recording
force.
-
Record the average or required value and describe the failure mode.
Pull-Off Adhesion Testing
Pull-off testing bonds a loading fixture, commonly called a dolly, to
the coating. A portable adhesion tester applies force perpendicular to
the surface until separation occurs or a specified proof load is
reached.
ASTM D4541-22 covers pull-off strength of coatings using portable
adhesion testers. ASTM D7234-22 specifically addresses coatings on
concrete. These methods are not interchangeable without considering
substrate, equipment, procedure, and specification.
Critical Procedure Controls
- Tester type and verified calibration
-
Dolly material, diameter, surface preparation, and alignment
-
Adhesive type, mixing, application thickness, and cure
-
Whether the coating is cut or scored around the dolly
- Loading rate and perpendicular alignment
-
Coating thickness, substrate stiffness, temperature, and moisture
-
Repair of every test location after completion
Do not compare values produced by different devices or procedures as
though they are identical. The report must name the standard, tester,
method, dolly size, scoring procedure, units, and failure description.
Failure Mode Can Matter More Than the Number
The force value indicates how much stress was applied, but the fracture
location identifies the weakest part of the tested system.
| Observed Failure |
What It May Indicate |
|
Adhesive failure at the roof-to-primer or roof-to-coating
interface
|
Surface preparation, contamination, primer selection, wetting,
cure, or chemical compatibility may be inadequate.
|
| Adhesive failure between primer and coating |
Primer cure, recoat window, contamination, or system
compatibility requires review.
|
| Cohesive failure within the new coating |
The substrate bond exceeded the coating's internal strength at
that condition; compare with the specified acceptance criterion.
|
|
Cohesive failure within an old coating, membrane facer, asphalt,
foam, or concrete
|
The existing substrate layer is the weak plane and may be
unsuitable even though the new coating adhered to it.
|
| Glue failure between dolly and test coating |
The test adhesive or dolly preparation failed; the result may
not measure roof-system adhesion.
|
| Mixed failure |
Report estimated percentages and locations; do not reduce the
result to one number.
|
A high pull value that tears away weak concrete, delaminates an old
coating, or removes a membrane facer can reveal a serious substrate
problem. It is not automatically a passing result.
Where to Place Test Patches
One convenient test near the roof hatch does not represent a complex
roof. Select locations based on materials, exposures, repairs, and
preparation variables.
- Each different membrane or roof substrate
- Weathered and less-weathered exposures
- Field membrane, seams, flashings, and previous repairs
-
Each existing coating type, color, age, or apparent application period
-
Factory-painted, field-painted, galvanized, corroded, and repaired
metal
-
Oil-, grease-, salt-, exhaust-, or silicone-affected zones after
cleaning
-
Areas using different cleaners, profiles, primers, or tie coats
-
Representative concrete placements, repairs, profiles, and moisture
conditions
Establish Acceptance Criteria Before Testing
The specification or manufacturer should state the test method, specimen
construction, cure period, test conditions, minimum result, acceptable
failure modes, number and location of tests, retesting procedure, and
action after failure.
Do not invent a universal minimum. For example, some GAF liquid-applied
roof specifications require field peel adhesion of at least 2.0 pounds
per linear inch for identified systems and substrates. That value should
not be transferred automatically to another manufacturer, chemistry,
substrate, pull-off method, or warranty program.
Pass/Fail Requires Two Answers
First: Did the measured value meet the written
requirement? Second: Was the failure mode acceptable?
Both answers must support proceeding.
Cure and Environmental Conditions
Adhesion develops as primers, coatings, and test adhesives cure.
Temperature, substrate temperature, humidity, dew point, film thickness,
ventilation, sunlight, and moisture exposure can change the required
cure time.
Record application and testing conditions. Do not test early merely to
protect the schedule. If the field coating will be applied under cooler
or more humid conditions than the original test patch, additional
representative testing may be necessary.
Testing Primer and Preparation Options
When the approved documents allow more than one procedure, use separate,
clearly labeled test patches. Change only one variable at a time when
possible.
- Cleaner A versus cleaner B
- Primer versus an approved no-primer condition
- Different abrasion or metal-preparation standards
- Bleed-blocking primer systems over asphaltic surfaces
- Dedicated TPO, PVC, or EPDM surface treatments
- Tie-coat options over existing approved coatings
The selected procedure should be approved in writing and transferred
exactly into crew instructions, quality-control forms, and material
estimates.
When a Test Fails
- Stop coating work on the represented condition.
-
Confirm that the test specimen, cure, equipment, and procedure were
valid.
-
Identify the failure plane and inspect the exposed surfaces.
-
Investigate contamination, moisture, weak substrate, inadequate
profile, incorrect primer, coverage, and recoat timing.
-
Obtain written manufacturer or design-professional direction.
-
Revise the preparation or system only through an approved procedure.
- Install and test new representative patches.
-
Do not begin production until the revised process passes and is
documented.
Adhesion Test Report
- Project, roof area, date, and unique test number
- Roof-plan location and photographs
-
Substrate, existing coating, repairs, and observed condition
-
Cleaning, abrasion, profile, primer, and preparation procedure
-
Products, batch numbers, mixing, coverage, and reinforcement
-
Application and cure conditions, dates, and elapsed time
- Test standard or manufacturer procedure and edition
-
Tester, calibration status, dolly or strip dimensions, and units
-
Individual readings, average when required, failure modes, and
percentages
-
Pass/fail decision, approving party, corrective action, and
test-location repair
Testing Errors That Should Be Rejected
-
Calling every rooftop peel test “ASTM D903” without following the
standard
-
Testing only one clean area on a roof with several substrates and
repair materials
-
Changing cleaner, primer, or preparation after the test without
retesting
-
Testing before the coating or dolly adhesive has cured
- Reporting only the highest reading
-
Ignoring failure within the old coating, facer, foam, asphalt, or
concrete
- Comparing peel values directly with pull-off values
-
Using another manufacturer's minimum value as a universal requirement
- Leaving destructive test locations unrepaired
Technical References
Application rule: Use the current referenced standard,
project specification, approved manufacturer field-test procedure,
product data sheets, cure requirements, and written warranty criteria.
Test methods and acceptance values are system-specific and are not
interchangeable.
