I saw the results of a rock chimney explosion, people thought it was a bomb but no explosive residue or parts could be found. Blew the mantle off the fireplace across the room and a hole in the wall big enough to step through.
EOD called in to look for other 'Devices' & clear the area for crime scene techs.
Lesson, don't let an 'Artistic' type build your fireplace!
I've seen small fire ring rocks 'Pop' but never explode. Don't want to be around if they do!
Virtually all sand is washed, so dry it carefully!
I don't need pieces of rock sharper than any kind flying at my eyes!
I've had refractory brick in the forges/furnaces pop/crack/blow apart. The manufacturer said it was most likely stored outside and rain allowed to soak it.
Some refractory cement is supposed to dry for 6 months before it's fired, and firing temps are supposed to increase gradually before the cement is fully 'Cured' for full temp use.
If I were to use 'Glass' media, I would heat it away from anything that could be damaged before I tried to anneal around it. Simply safety just in case anything decided to 'Pop' and throw hot/sharp stuff at me.
Glass has 'Thermal Shock' issues.
I'm connected to Mr. Murphy at the hip, if it can go wrong, it will go wrong, and it will go wrong when I'm in the line of fire!
Physics: Nothing happens until something MOVES, and that means heat energy too!
I'm thinking smaller round 'Shot' media since there is less air space between smaller balls, and smaller balls will get into the necks better.
As for what a brass fixture would look like, for the little round pot (to keep hands out of the burn zone),
A round clam shell with hinge.
Drop brass in lower with holes for brass, close clamshell top.
This would be super simple for rimmed rounds, drop them in, close the clamshell, annealing, open clamshell & dump out.
For rimless, there would need to be some sort of stop. A sheet metal 'Finger' to catch the extraction groove, that's why I suggested a stripper clip which would hold the brass by the groove.
There is also the round hole with slots to make 'Finger', push the brass in until the 'Fingers' catch the groove, then when done just push ahead through. (Push Nut)
This is as easy as an undersized hole, snip 3 or 4 place from inside hole into sheet metal for 'Fingers', and the clamshell could have 'Pins' or 'Pegs' to push brass ahead through when done.
Holes spaced on a paddle would keep spacing between brass, sheet metal on paddle.
When I say 'Sheet Metal', I mean aluminum can. I would darn sure start with something cheap to prove concept.
This is assuming you wanted to use a handle to keep fingers out of danger range...
Loading/unloading a fixture/handle would slow things down, so is your hide worth the extra time?
This will have to be an Up/Down operation.
Dragging brass through a heated media will allow the media to build up on the forward side, have a groove that doesn't make contact as high on the back side.
Dragging would also make a groove in the media making for mixed results.
An up/down rail would be the same idea for a longer media bed for higher production rates.
We have all seen loading blocks with holes for brass, with rimmed rounds, simply a cover to keep brass from backing out (clamshell).
With rimless rounds, some sort of spring to catch the groove and some sort of peg/pin board to punch the brass past the spring, or a stripper clip arrangement that is loose enough to simply allow shaking the brass out of the rails.
We all inspect brass out of the cleaner, so when sorting you simply put the brass in the fixture for annealing before sizing.
I haven't figured out how to get a common case feeder to turn brass over for automated loading of fixtures, and even if I did, Mr. Morris would figure out a way to do it 10 times more efficiently!