Metallography in the steelworks

Metallographic preparation by mirror-polishing a cut section through a sample is the predominant detailed inspection starting point.
No matter how scaley and rough looking the sample, when you'd gone through the sequential process of there is a perfect mirror. A perfect mirror on the abrupt cross-section through a usually scaley rough piece of mill steel.
That surface remains a perfect mirror at hundreds of times magnification under an metallurgical microscope...

This metallographic preparation facility was by far the best I ever used. Run by grump South Yorkshire hereditary industrial folk, the equipment was simple and improvised but perfect. No wheels were flatter, no setup more cleanly separated between stages, etc..
As an example - for the final emergy grind at 1200grit, you could touch-down your sample on the spinning water-covered disk then lift it off and there would not be a single mark of the 1200grit. All the prior grind at the coarser grit was fully preserved untouched, depite being right-angled to the 1200grit wheel spin. The sample was so flat it would "aquaplane". Go back, have the sample "aquaplaning", push gently and a puff of black dust in the water for 3 seconds, lift the sample and every mark was now the 1200grit running in the new grinding direction.
Contrast - in most other grinding set-ups, nothing is quite right, you end up grinding for ages trying to remove the last of the previous stage's marks, etc..
You could "knock-out" a palm-full of small samples on this set-up "just to pass the time" while you waited for someone to finish a phone-call.

Then onwards down two different investigative branches.

For shape defects like surface "cracks", straight to the microscope with the mirror-polished sample. Usually from the "billet finishing" shop-floor department on inspection at the "wave grinder". The nature of the debris inside deeply entrant ("crack-like") defect indicated origin. Smooth multiphase glassy slag oxide fillings indicated an ingot-casting defect in the melting-shop, a lightly oxidised surface with no debris would be a genuine metallurgical crack etc..

For metallurgical issues, a few seconds with an acid restrained-back by being dissolved dilutely in alcohol "mists" the surface with a fine micro-etch.
Under the metallurgical microscope, there is a section through the microscopic structure of the steel. That's a very complex matter, whose optimisation enables all the range of properties. Much of the typical phases possible are very recognisable and I was soon immersed in the topic.

Carrying that onto the Works, my interest was completely "ignited" and my career was defined.

(R. Smith, 13Dec2017, 26Dec2017, 02Jan2018)