Speaker
Description
This paper has the aim at presenting a methodology of defect root cause analysis for hot-dip galvanized coatings. It has been developed to provide a global overview of all aspects related to the Materials Science pyramid – the processing route, the material properties, the structure and in-use performance – all linked by the material’s characterization. Its final aim is to provide not a single parameter leading to failure or poor performance, but to identify the space of parameters and their combinations leading to the undesired outcomes for a given coating system. In order to perform such an integrated analysis a minimal set of process parameters must be identified, generally through brainstorming.
Management of data acquisition and communication between production site, characterization team, quality surveillance, and engineering must be carefully carried out with full traceability of data to allow the assembly of the Materials Science pyramid database. Process parameters shall be acquired for a statistically significant population (comprising accepted and rejected coils) and each coil’s coating microstructure must be analyzed to provide the base of the dataset. A quality control parameter related to the coating being studied is required to finish the triangle of process-property-structure.
The method is composed by three main steps. First, establishing a set of parameters representing the structure-related features of the coating material, what is followed by data clean-up for ensuring stability of conditions for each coil. Finally, the application of clustering techniques, proves more reliable and efficient than classical statistical analysis, for interpreting the parametric effects over the quality issue being analyzed rather than studying the tendencies promoted by individual parameters.
Keywords
Image analysis, root cause analysis, defects, methodology, microstructure, process