Abstract

Following pull-off coating adhesion tests such as ASTM D4541, visual examination of the fracture surfaces often reveals intriguing fracture patterns where similar (and seemingly predictable) features are observed across like coating systems. Nonetheless, adhesion test stubs are typically discarded after a brief visual inspection or no inspection at all—suggesting a missed opportunity to gain further insight into fracture processes. This paper examines the pull-off adhesion test from a fracture mechanics perspective. We present two microscopy-based methods for extracting analytically useful fracture data from pull-off adhesion tests. First, we develop an energy-based method for obtaining mode mixity from observed crack kink angles, providing the powerful capability to disambiguate mode I and mode II fracture. We demonstrate agreement between experimental results and theoretical predictions in a set of 1,305 crack path measurements for 111 pull-off adhesion test stubs, representing both single- and multi-layered coating systems. Second, we deploy a machine-learning based image segmentation technique (Trainable WEKA) to rapidly quantify the fractional area coverage of materials removed with the pull stub, enabling further insight into the surfaces most vulnerable to delamination in a layered coating stack. Beyond the immediate utility of these techniques for data enhancement in pull-off adhesion tests, these techniques also have long-term potential as tools for failure analysis in coated systems.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call