Surface Preparation and Dolly Bonding

A pull-off adhesion test introduces a controlled tensile stress into the coating system through a dolly bonded to the surface with a structural adhesive. For the test to measure the coating’s bond to its substrate, the adhesive joint between the dolly and the coating must be stronger than every interface and layer within the coating system. If it is not, the adhesive fails first and the test result reflects the glue bond rather than the coating adhesion.

Three preparation variables govern whether this condition is met: the cleanliness and mechanical profile of the bonding surface, the selection and application of the adhesive, and the curing conditions (time, temperature, and environment) under which the adhesive develops its strength.

A fourth variable — scoring — defines the geometry of the test. When the coating is scored (cut) around the dolly circumference down to the substrate, the test area is mechanically isolated. This ensures that the measured stress relates to the defined dolly area and is not influenced by the tensile strength or shear resistance of the surrounding coating film.

Surface preparation for the dolly bond begins with cleaning the coating surface at the test location. Dust, oil, release agents, moisture, and any loose material must be removed. On smooth or glossy coatings, light abrasion (typically with fine sandpaper or an abrasive pad) creates mechanical keying for the adhesive. On rough-textured coatings, abrasion may not be necessary, but the adhesive must be applied in sufficient quantity to fill the surface profile and achieve full contact with the dolly face.

The dolly bonding face must also be clean and, ideally, lightly abraded. Dollies that have been reused should be cleaned of all residual adhesive and coating material. A contaminated or uneven dolly face reduces the effective bonding area and can create stress concentrations that initiate premature failure.

Adhesive mixing and application must follow the manufacturer’s instructions. For two-part epoxy adhesives, the correct mix ratio is critical; an incorrect ratio produces an under-cured bond with reduced strength. The adhesive should be applied as a thin, uniform film — thick enough to fill any surface irregularities, but not so thick that the bond line itself becomes a weak layer. Air bubbles trapped during mixing or application reduce the effective bonding area.

After placing the dolly, excess adhesive that squeezes out around the perimeter should be removed before it cures, particularly if scoring will be performed. Cured adhesive overflow that bridges the scored line can carry load outside the intended test area, inflating the measured pull-off value.

Scoring is performed after the dolly is bonded but before the pull. The cut should penetrate through all coating layers to the substrate, following the dolly circumference as closely as possible. Scoring tools include manual cutting guides, drill-mounted hole saws, and template-guided knives. Incomplete scoring leaves the test area partially connected to the surrounding coating, which can increase the measured stress and change the failure mode. Scoring that extends into the substrate (especially on soft substrates such as concrete) can weaken the substrate locally and promote substrate failure.

Curing time is the single most influential preparation variable. Adhesive strength develops over time and is strongly temperature-dependent. At room temperature (approximately 20–25 °C), most two-part epoxies used for adhesion testing require a minimum of 12–24 hours to reach full strength. At lower temperatures, curing slows significantly: at 5 °C, the same adhesive may require 48 hours or more. Testing before the adhesive has reached adequate strength is the most common cause of glue failure.

• Material-dependent effects: Coating surface energy and chemical composition affect how well the adhesive wets and bonds to the topcoat. Low surface-energy coatings (silicone-based, fluoropolymer, some anti-fouling systems) resist bonding and may require special adhesives or surface treatments. Porous coatings can absorb adhesive, starving the bond line. Very rough coatings require more adhesive to fill the profile, increasing the risk of air entrapment and thick bond lines.
• Environmental effects: Temperature during curing directly controls the rate at which the adhesive develops strength. Humidity can interfere with some adhesive chemistries and can introduce moisture at the bonding interface. Wind and airborne dust during the bonding and curing phase can contaminate the adhesive before it sets. Direct sunlight can cause uneven heating of the dolly and adhesive, leading to differential curing and internal stress.
• Instrument/technique-specific effects: Scoring tool condition and technique affect the quality of the cut. A dull or wobbling cutter can create ragged edges, pull coating material, or score too deep into the substrate. Dolly diameter determines the test area and therefore the stress calculation; dollies must be dimensionally accurate and free of edge damage. The choice of adhesive must be matched to the expected adhesion range: a standard epoxy may be adequate for coatings with moderate adhesion (2–8 MPa), but high-adhesion systems may require a higher-strength adhesive to avoid glue failure.
• Operator or setup issues: Inconsistent surface cleaning between tests introduces variability. Incorrect adhesive mix ratio — whether from inaccurate dispensing, incomplete mixing, or estimating quantities by eye — is a common and invisible error that degrades bond strength unpredictably. Placing the dolly off-centre relative to the scored circle changes the effective test geometry. Impatience with cure time is the single most frequent procedural error; testing “early” to meet a schedule produces glue failures that are then incorrectly attributed to poor coating adhesion.

Surface preparation and bonding quality matter in every pull-off adhesion test, regardless of substrate or coating type. However, the consequences of poor preparation are most visible in high-stakes applications: acceptance testing of protective coatings on infrastructure, quality audits for intumescent fire protection, and verification of concrete repair systems where low adhesion can indicate a structural deficiency.

Glue failures are frequently misread as evidence of poor coating adhesion. When the fracture surface shows adhesive (glue) on both the dolly and the coating surface with no coating detachment, the test has not stressed the coating system to failure. Reporting the pull-off value as a coating adhesion measurement in this case is incorrect and can lead to unnecessary rejection of compliant coating work.

The importance of dolly bond preparation is sometimes overlooked on the assumption that surface preparation only matters for the coating applicator, not for the adhesion tester. In reality, the dolly bond is itself an adhesive joint, subject to the same principles of cleanliness, wetting, mechanical keying, and cure quality that govern the coating’s bond to the substrate.

Seasonal variation in curing temperature is also a significant but under-recognised influence. A series of tests performed in winter (low temperature, extended cure) and summer (warm temperature, faster cure) on the same coating system can produce systematically different pull-off values purely because the adhesive bond strength differs, not because the coating has changed.

• Pull-off Adhesion Test Principles
• Adhesion Test Failure Modes
• Verification & Checks for Adhesion Testers

If surface preparation and bonding reliability are becoming the deciding factors, the next step is usually to review the wider setup requirements around valid pull-off testing.

• Adhesion Test Setup and Validity if you need to compare how much control the overall pull-off workflow should provide around preparation, curing and validity.