maybe this paper would be helpful,

it deals with bainitic microstructures and their classification scheme;;

you need nital, SEM+EBSD, see Zajac, et al

best regards

Paper:

Characterisation and Quantification of Complex Bainitic Microstructures in High and Ultra-High Strength Linepipe Steels, November 2005, Materials Science Forum 500-501:387-394

DOI: 10.4028/www.scientific.net/MSF.500-501.387

• 📷Stanislaw Zajac (he is on researchgate)
• 📷Volker Schwinn
• 📷Karl-Hermann Tacke

What about to employ machine learning or AI techniques. They perfectly fit your task. All you need is to write down or find techniques, probably advanced neural networks, to accomplish your task.

It might be a long endeavor but it is where the research currently goes.

Try to read this paper

https://www.sciencedirect.com/science/article/abs/pii/S1359645417307383

You could use basic Nital etching and SE imaging; its relatively easy to distinguish between the two. Yet to bring certainty and be confident you would need to use TEM. There are always errors when using SE imaging to distinguish between various phases. Sometimes surface topography may provide wrong interpretations.

EBSD is another way, but you will need very good sample preparation. And again there are uncertainties and its not so easy and clear to distinguish between the two phases.

The common procedure for what you intend to know is as below:

1- SEM for microstructure and preliminary chemical analysis

2- XRD for crystallography as a complementary method

3- TEM could be used as well if you are interested in thorough study on the formation of bainite and relative movement of Carbon with respect to austenite (or retained austenite). But it also depends on your heat treatment as well.

The martensite like fine needles crowded