That topic is a great way to get off into either a few reasonable assumptions or a whole pile of B.S.. The crystals inside of steel reform from heat. You cannot shrink them, fracture or break them up smaller with a hammer, all you can do is deform them. The deformation thus introduces strain energy which will affect how the heat reforms them. But then the question comes in as to which is more efficient heat and hammer or heat alone. The hammering and reheating will indeed knock the grain size down quite quickly, but unless you use something like a rolling mill so that you can do it all very evenly it could result in a very uneven grain size and distribution. However if you use thermal treaments alone you can use the phases of the steel to also quickly reduce grain size and then all you have to do is heat it evenly. We are now in the 21st century and progress is being made, virtually every credible master smith I know now wisely distances themselves from the old silly notions of edge packing.
So much of the old misconceptions are a result of our own eyes lying to us. Any time a smith gets a results that resembles what they were wishing for they jump to conclusions and a new pile of manure is born. Blindly saying "it works for me and thats all I need to know", has shackled this business in ignorance for centuries and can lock any of us in medievalism very quickly.
I just experienced it this week, while doing experiments in low temperature cycling as it relates to segregation. I knew that high termperatures would dissolve segregated constituents and move them around but after normalizing the samples microscopy showed heavy banding in the high temp pieces and no banding in the low temp pieces
Everything in all the books was wrong! Then I followed that up with a quench on the same steels and got the exact reverse, just as the books say you should
Putting the pieces under higher magnification explained it all. The piece that was cycled at higher heats did indeed have the carbon evenly dissolved so that there were no huge crabides hanging around and its grain was a couple sizes larger. So when normalized the segregated finer grained stuff was so low in hardenability that it transformed competely into softer phases with no contrast. The stuff that was cycled hot had increased its hardenability so much that large bands of martensite had formed between sheets of very fine pearlite and bainite giving a stark contrasting white and black banding. Quenching them both, of course reversed this situation.
If I had taken the prevaling attitude and approach in bladesmithing to immediately run with the initial results that I seen with my own eyes and thus not ask any deeper questions, I would be throwing away all of my metallurgy books right now because they simply don't apply to knifemaking
.
