Metallographic preparation is a vital part of quality control in manufacturing processes, including potentially complex structures such as additive materials and thermal spray. It has long been recognized that relatively small variations in approach can lead to significantly different results in some measurements.

Metallographic preparation tends to be applied dogmatically within a some manufacturing environments in an attempt to ensure reproducibility and consistency. Despite this, it has been demonstrated many times that this is not a reliably effective control.

The lack of comparability between organizations can lead to large uncertainties and associated inefficiencies in a production environment. Significant cost factors include both the time taken to re-prepare samples and the increased likelihood of rejecting perfectly good batches (or passing poor ones).

If the metallographic process and its effects are understood more scientifically, and then applied in such a way as to improve reproducibility and consistency, this would allow greater potential for meaningful comparison between laboratories and improvement in long-term process control. Accurate measurement is similarly important in the application of machine learning, as the quality of data may otherwise be a limiting factor.

This talk will discuss the influence of various parts of the metallographic process in measurement accuracy, demonstrating common problems and providing guidance in best practice.

Dr. Mike Keeble