Degree Name

Doctor of Philosophy


Department of Materials Engineering


The investigation reported in this thesis was carried out to study the adhesion between epoxy coatings and an alloy-coated steel sheet substrate. The main aspects studied were (1) developing and evaluating a suitable test method for measuring adhesion; (2) the effects of test conditions (temperature and loading rate ) on the adhesion; (3) the effects of different coating curing techniques on the adhesion; and (4) the effects of material factors (thickness of coating and composition of coating materials ) on the adhesion.

A number of test methods were studied and compared in this investigation. The results have shown that a "pull-off" test is the best method for measuring the adhesion between the epoxy coating and metallic substrate because it can supply more information regarding the mechanisms of adhesion and its failure.

The adhesion was measured from -30 to 120 °C with loading rate from 0.001 to 60 mm/min. Test results showed that increasing the loading rate decreases the adhesion, since a higher loading rate results in higher stress concentration at the tips of micro cracks at the interface between the coating and substrate. However, the adhesion increases to a maximum as the test temperature increases since more viscoelastic flow and plastic deformation can occur at high temperature resulting in a higher toughness. The adhesion decreased at higher temperature due to a decrease in the modulus of the epoxy. The behaviour could be accurately predicted using a fracture mechanics model.

Analyses of the effect of cure conditions on the adhesion were undertaken. Samples baked at 232 °C for a various time from 3 to 63 minutes showed that the adhesion strength passes through a maximum. Crosslink density, modulus and Tg increase with curing time up to a maximum value with the result that the pull off force also increases. At longer cure times the adhesion decreases due to brittleness resulting from degradation of the coating. A second baking cycle ( as commonly used in industry ) was found to have a similar effect to extending the cure

In industrial coating processes, the coating is water quenched following baking. Chemical and physical changes occur in the coating during the cooling process. Chemical changes result in the formation of crosslinks, while physical changes result in internal stress occurring between the coating and the substrate. Increasing cooling rate results in decreasing adhesion because it produces a lower modulus and higher internal stress. The internal stress was found to dominate in the system studied in investigation.

The coating adhesion increases with increasing peak curing temperature and longer heating time due to an increase in modulus and better wetting of thesubstrate. If the peak curing temperature is too high, however, degradation of the coating occurs and the adhesion decreases.

The effects of three factors (thickness of coating, ratio of pigment to binder and ratio of resin to hardener ) on the adhesion were studied. Thicker coatings show lower adhesion due to the generation of higher internal stress. The adhesion generally increases as the amount of pigment used increases, because the modulus of coating also increases. However, the pigment may decrease the wettability of coating and form larger micro-cracks between the coating and the substrate, which are likely to reduce adhesion. Increasing the amount of hardener can increase the adhesion vi because the modulus of the coating increases. In some cases the increase in adhesion with increasing hardener was not profound, and in a few cases the adhesion actually decreased slightly. These anomalous results are attributed to the secondary effect of increasing crosslink density and hence an increased brittleness. Test results also showed that only the pigment to binder ratio combined with the thickness shows a two-factor effect on the adhesion; the effect of three-factors on the adhesion was not significant.



Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.