There's also centripetal acceleration but there *might* be a way round this.
the particles being fired around the LHC are so small relative to the pipe they're in that from their POV it is a straight pipe.
The problem is the side force or centripetal acceleration. Using a = r omega^2 with a circle 10Km in radius I got roughly a 16g side force (when the package is hitting orbital velocity of 7950 m^-1) with a 3g front to back acceleration. Not too serious for electronics or anything less complex than say an insect but bad news for humans. Humans don't seem to do 2 axis accelerations well. BTW an F1 driver at 200mph is roughly 89ms^1 or 1/90 of low earth orbit speed. The possible solution follows.
The ring is approximated by an n sided polygon with straight sides and cylindrical chambers at each chamber. The structure is evacuated to reduce the losses of the payload package pushing all that air ahead of itself.
The wall segments act as linear accelerators giving 3g kicks to the package. As the package free flies between segments (although still under levitation) its yawed about its centre of mass to line up its front end with the next segment for more acceleration. Anything at or near its CM *should* only experience pulses of linear front to back acceleration.
The structure has 2 exits, one aligned to the latitude and one for a polar orbit. Once the package has reached its release velocity it is diverted to an exit tunnel leading to an exit valve while the rest of the system is sealed to preserve vacuum.
*if* it sidesteps the centripetal acceleration then it just becomes a problem if weather a human can survive repeated acceleration pulses without their sense of balance being permanently scrambled. Using 9144ms^-1/30 000 fps and a 3g acceleration gives roughly 3mins 11 secs
At least 2 methods exist to handle the attitude problem. Control moment gyroscopes at each end of the package or push/pull electromagnets at the entry and exit of each straight section can do the job if their big enough. Passive magnets have a pull/weight ratio of c50:1 (better than any known jet engine)
People who've looked at this have talked about "Shutter" valves of metal foils moving on *very* fast carriers. The differential pressure is still only 1 atm.
Note that while the *rough* layout is simple to describe there are multiple *key* trade-offs which would radically affect how much it would cost to build and run it. The obvious ones are the number of sides in the polygon and the angle of the exit tunnels (45deg gives the shortest length to any given exit height but maximum force needed in the diversion system. Shallower angles give lower forces but much longer tunnels (but with longer to seal the main system). More subtle one would be where to apply the yaw forces. If the package is being turned while still partly *in* a straight section it will need clearance between itself and the wall. Less clearance -> more force -> less power needed.
This still leaves you with a substantial object moving at orbital velocity at near ground level. The thermal protection system is *not* trivial.
Mine would have a copy of the last even years February IEEE Trans. Magnetics in. It's the Railgun Issue, packed with all sorts of pulsed power goodness (or alternatively elaborate ways to make yourself the guest of honor at a closed casket funeral).