I suspect it is a bit of both.
I use a Lee Precision tool to size the cases View attachment 162202 and whilst the case is chucked up in the drill, I also offer up the RCBS chamfer tool to deburr View attachment 162203
What I discovered is that if you offer the deburrer up to the case mouth of an un-annealed piece of brass, you can get cutting edge chatter which renders a rough edge to the chamfer. Even at higher rpm.
Later trials have born out my theory: if I anneal first, the same set up gives me a smooth chamfer. And almost nil bullet swarf.
I trim and chamfer in the drill in the same way as you.
The chamfer tools themselves throw up a bit of burr, which you can feel if you run your thumbnail up the outside of a trimmed and chamfered case. That little burr is going to be the hardest part of the case mouth having been work hardened through forming by the chamfer tool pressure.
What I was trying to get at is that even were the materials of case and bullet of equal hardness, the sharp edge burr thrown up by the chamfer tools is a point loading (stiletto heel) and is therefore going to scrape or shave the bullet rather than the other way round.
But if the case mouth is flared the bullet can only make contact with the soft radiused form. The action is then one of burnishing rather than scraping.
The sketches may help explain what I am trying to describe...the chamfer burr and flare obviously exaggerated...but I found with Barnes Bullets it not only catches on the base of the bullet but also on each of the driving bands and produces those shavings seen in your seating die. You can feel the notchy seating action as the case mouth springs back into each of the cannelures and then is expanded again...
If your chamfer tools are sharp and cutting clean...you might find just a heavier internal chamfer will improve things, if the face of the case mouth is hitting the bullet rather than the chamfer, i.e. the chamfer OD is not greater than the interference with the calibre.
The expander / flare tool removes the risk of scraping completely.
Alan

Last edited:











