Fred, I know it seems quite small but on the scale of carbides the HRC test dimple is huge, too big to measure the individual carbides. What the Rockwell test measures is the overall matrix of the steel. This topic touches well on the analogy I use often in my classes of a doughy ball of clay filled with small chunks of glass. If you skate a file over the ball of clay the tops of the chunks of glass will give you the impression that the entire ball is as hard as glass. If you take the handle off the file and plunge the tang into the ball, the glass chunks will just push aside and you will see that the ball is as soft as clay. This is the difference between Rockwell tests and skating a file. Hardness is actually not a property of steel as much is it is a way of us conceptualizing or quantifying other real properties. Scratch tests like the Mohs scale and files measure the property of abrasion resistance; Rockwell measures the property of strength.
So let’s start out with1080 steel. At room temperature it is pearlitic, a phase consisting of alternating bands (lamellae) of ferrite (soft iron) and cementite (very hard iron carbide). A Rockwell test will simply bury the needle into the ferrite, pushing the cementite aside as it goes. The reading will probably be HRC 32 or less. The matrix itself is not strong because the carbon is trapped in the carbide and not reinforcing the iron. But if we heat the steel up and dissolve the carbide to mix its carbon into the ferrite, and then quench it, the overall matrix will then resist the penetrator to give us a reading of 65 HRC.
Now let’s do some magic forging techniques that has been fantasized about over the years and add carbon to our 1080 until it is 1095. Pretty much all the same behavior will apply until we quench it, then we will get the same HRC readings there as well, but there will be all these extra iron carbides that are even finer than the pearlite lamellae and push aside all the easier. They contribute even less to the overall Rockwell reading, but push the abrasion resistance via scratch hardness much higher.
Adding chrome, tungsten or vanadium to the mix will increase the number of carbides formed because these elements like the carbon even more than iron does. So you must be careful not to put so many in that they use up all the free carbon and limit your maximum Rockwell hardness by robbing the matrix of carbon. These carbides will be more complex and much harder so even more abrasion resistant, but they still are too small to have an effect on the HRC value.
If you take your piece of steel, and put it under a Vickers or Knoop type micro-hardness tester, which is more like a microscope with a microscopic penetrator built in, your Rockwell dimple will look like Meteor Crater in Arizona. But if you polished a section next to it to reveal carbides and the carbides were huge, you may be able to drop the micro-hardness penetrator on one and get a reading of equivalent to 70, 80 or even 90 HRC, even though the steel itself is giving an overall Rockwell reading of only 65.
While tempering decreases the HRC reading of the steel by taking carbon out of the matrix, it actually increase the carbide numbers as it uses the freed carbon to make ultrafine tempering carbides, which you would never measure the hardness of, even with a micro-hardness tester, because they are so fine to see with anything short of an electron microscope.
So in the end you have the two concepts of hardness discussed here- scratch hardness, or abrasion resistance, and penetrative hardness or overall strength. Carbon that is free to reinforce the steel matrix increases overall strength as read by the Rockwell tester, while carbon locked in the form of carbide which contributes little to overall strength, and thus is not measured by the Rockwell test, but increases abrasion resistance, so it will eat up files, mills and even challenge abrasives (ask anybody who has hand polished Cru-forgeV how pleasant it was).