The Case of the Mysterious Hailstones [Premature Coating Failure]

Specification writers, contractors, and coating manufacturers, at one time or another, have been blamed when installed coatings fail to meet service life expectations. However, rarely does the wrath of Mother Nature receive the blame for premature coating failure. In this Cases from the F-Files, an interior roof coating system reportedly failed because of localized impact on the roof exterior from large hailstones during a severe thunderstorm. Can this really happen, at least in this case? Let’s find out.

Background

The facility was a 100,000-square-foot aircraft hangar supported by a steel framework and covered with steel panels. Shortly after erection, the interior roof panels and associated supporting steel were painted with an alkyd-based coating system. The original alkyd coating was reportedly applied at a dry film thickness (DFT) of 3.0–3.5 mils. After approximately 10–15 years of operation, the roof panels and support steel were power washed with water to remove surface dirt, then overcoated with a water-borne acrylic dry fall coating (also called dry fog coatings or sweep-up coatings). The dry fall overcoat was reportedly applied in a DFT range of 3.0–4.5 mils.

Approximately three years after the application of the dry fall waterborne acrylic overcoat, the dry fall coating system began to delaminate from the original alkyd coating. The falling paint chips contaminated the hangar workspace. Aircraft maintenance operations had to be suspended until the problem was resolved. The peeling paint was blamed on a recent severe thunderstorm during which large hailstones struck the exterior of the painted roof panels and reportedly caused delamination of the coating applied to the underside of the roof panels. The owner called in a third-party consultant to analyze the failure.

Site Investigation

A visual inspection within the hangar revealed obvious areas of distressed coating on the inner roof and associated network of supporting roof trusses. Both locations contained areas where the gray acrylic topcoat had spontaneously delaminated from a buff-colored alkyd layer. The areas of coating failure were distributed throughout the hangar. The exposed alkyd layer was covered with many small areas of rust that had developed at the steel substrate beneath the alkyd. Figure 1shows extreme delamination of the gray waterborne acrylic dry fall coating off the pre-existing buff alkyd layer applied to a corrugated roof panel. Figure 2 shows the same type of delamination of the light gray acrylic overcoat from the buff alkyd layer on a supporting steel member.

Fig. 1: Extreme delamination of the dry fall coating from the original alkyd on the interior roof panel Photos courtesy of the author.

Fig. 2: Extreme delamination of the dry fall coating from the original alkyd on supporting steel

Physical probing with the point of a knife blade was used to explore the degree of coating adhesion at failing and non-failing areas of the applied roof coating system. At failing areas where the acrylic was already delaminating from the alkyd layer, probing extended the coating loss to a point at least 6–8” beyond the original damage. At most areas tested, the acrylic topcoat could be cleanly separated from the underlying alkyd layer using only a moderate degree of physical probing.

In contrast, the adhesion of the alkyd layer to the underlying steel substrate was excellent. Aggressive scraping was required to remove the alkyd from the steel at all tested locations. Although pinpoint rusting had penetrated the alkyd layer, the overall bond to the steel remained strong. Figure 3 shows that the original alkyd layer was removed only after aggressive scraping with a knife blade.

Fig. 3: Adhesion of the original alkyd to the steel was excellent. Aggressive scraping was needed to remove the alkyd.

An electronic gage was used to measure the thickness of the applied coatings. The thickness of the original alkyd layer on the roof panels and supporting steel ranged from 3.5 to 6.5 mils, with an average value of 5.0 mils. The thickness of the waterborne acrylic dry fall overcoat ranged from 7.8 to 12.0 mils, with an average of 9.3 mils.

The field investigation revealed the following significant factors that caused the coating failure:

Excessive Coating Thickness

The DFT of the waterborne acrylic dry fall coating on the interior surface of the roof panels exceeded the manufacturer’s recommended DFT. According to the product data sheet, the recommended DFT for the coating was 3.0 to 4.5 mils, with an average of 3.8 mils. Measurements at the site revealed an average of 9.3 mils and a range of 7.8 to 12.0 mils. The average thickness of the acrylic topcoat was approximately 2½ times greater than recommended by the coating manufacturer.

Flexibility Losses

Dry fall coatings are formulated so that when they are sprayed onto a substrate, the overspray droplets dry before they contact the floor or other surfaces. The coatings are designed to dry after falling 8 to 10 feet, depending on the formulation and the environmental conditions. The use of dry fall coatings minimizes the amount of masking and covering of surfaces below that are not to be coated, and the dried coating can easily be cleaned up simply by sweeping. The characteristics of the coatings require application by spraying because the quick-drying characteristics of dry fall coatings would not be realized with non-spray application techniques.

Dry fall coatings are typically formulated with a high pigment-to-binder ratio to facilitate the formation of dry particles and rapid cure. These properties also result in a reduced cohesive strength, leading to a loss in flexibility at high film builds. The water-borne acrylic dry fall coating applied within the hangar probably has an acceptable degree of flexibility when applied in the recommended thickness range of 3.0 to 4.5 mils. Exceeding these limits caused a corresponding loss of flexibility within the cured coating film.

Internal Coating Stress

Coating films also develop internal stress as they cure from a liquid coating to a protective film. When coatings are applied at a normal (recommended) thickness, the stress levels are uniformly distributed during the curing process. When recommended thickness levels are exceeded, the concentrations of resulting stress can lead to adhesion loss. The high pigmentation levels of dry fall coatings also produce lower cohesive forces between molecules, contributing to cracking and splitting of the cured film.

The excessive thickness of the water-borne acrylic overcoat led to the development of longitudinal cracks in the panel coating and the spontaneous delamination of the acrylic topcoat from the aged alkyd.

Lack of Adhesive Bond Helps Lead to Coating Failure

In addition to the internal stress developed within the thick layer of the acrylic dry fall overcoat, poor adhesion between the acrylic topcoat and the pre-existing alkyd was the most significant factor that led to the coating failure. Because the acrylic topcoat is waterborne, it had no aggressive solvents to soften the surface of the pre-existing alkyd and create chemical adhesion. Further, with dry fall products, any co-solvents would have very limited contact time with the alkyd film.

Alkyds dry by solvent evaporation followed by curing through slow oxidation from a reaction with oxygen in the air. The air oxidation is continuous over the life of the coating. Over its 10–15 years of service life, the alkyd coating became hard through aging and air oxidation. The aged, smooth alkyd surface thus did not present any significant surface texture to enhance the mechanical bond of the acrylic topcoat. White alkyd films also turn yellow when exposed to the air without weathering forces (e.g., sunlight), and the yellowing chalks the surface. It seems reasonable to speculate that the yellowed surface layer oxidation would be more advanced, further reducing the chance of intercoat adhesion. This surface layer oxidation then would make some form of mechanical surface preparation important before applying the dry fall coating to enhance mechanical bonding of the dry fall coating to the aged alkyd.

Before applying the overcoat, the contractor did not attempt to either roughen the surface of the pre-existing alkyd or apply a bonding primer. The contractor prepared the alkyd only by power washing with water to remove surface dirt and grime.

Other Factors Affecting Coating Adhesion

The roof was assembled by fastening the interlocking steel panels to a supporting system of metal trusses. Adjacent roof panels were overlapped and fastened to each other with self-tapping screws. Layers of board insulation and a waterproof membrane were added to the exterior panel surface to complete the roof. The built-up roof was somewhat rigid but was affected by external forces from snow, ice, and wind loading. Foot traffic can also transmit loads to a corrugated roof system. Although the exterior of the roof was insulated, seasonal temperature changes could also cause expansion and contraction of the corrugated metal panels. The movement could contribute to intercoat delamination of the acrylic dry fall coating from the aged alkyd.

Effect of Hailstone Damage

So was Mother Nature to blame? No evidence suggested that the impact of hailstones on the exterior of the roof caused the delamination of the waterborne acrylic topcoat from the interior of the roof. Rather, the delamination of the coating was the result of weak adhesion between the preexisting aged alkyd and the dry fall coating. The lack of an adhesion bond resulted from the following factors:

1. 1. the lack of roughness on the surface of the pre-existing alkyd;
   2. the absence of any chemical reaction (that would have created a chemical bond) between the existing alkyd and the water-borne acrylic overcoat; and
   3. the excessive thickness of the acrylic overcoat (at 2.5 times greater than the recommended average thickness) that may have contributed to the problem by causing internal coating stress during the curing (drying) process.

In support of these conclusions, the same kind of delamination failure was evident on supporting trusses and girders throughout the hangar. These supporting surfaces were not exposed to hailstone impact.

Recommendations for Repair

The overall appearance of the coating and the results of the physical testing revealed a chronic and widespread problem manifested by the lack of adhesion between the preexisting alkyd layer and the waterborne acrylic dry fall overcoat on the interior roof and the supporting steel members. The problem could not be remediated by simply overcoating the existing alkyd and dry fall coatings. The added weight of a new layer of coating would only further exacerbate the delamination problem.

Rather, all loose coating needed to be located and removed before any overcoating operations were undertaken. This was accomplished by a combination of hand scraping and pressure washing using 1,000–3,000 psi pressurized water and a 0° oscillating tip. Following removal of loose coating, the surfaces containing intact coating were lightly abraded and a bonding primer was applied to the aged alkyd surfaces before overcoating with a waterborne acrylic dry fall coating. The tight dry fall was also allowed to remain. Where it was intact, it was feathered, lightly abraded, and primed before overcoating.

Authors:

E.  Bud  Senkowski 

PE, KTA-Tator, Inc.

Richard  Burgess 

KTA-Tator, Inc., Series Editor

As seen in JPCL