what to do with a free scrap dumpster find ?

the way i did my draw bar is pretty close to zero contact other than where the air cylinder will make contact to eject the tool. i went with a high strength die spring rather than spring washers. seemed like a more compact holder tension. anything beyond a #30 tapper would not work because the die springs would be hard to find or fit. all the clamping happens when the fingers get drawn into the smaller diameter on the ID of the spindle the ejection happens when the stoke of the draw bar is if forced down the ID into the large section of the ID bore. the preload on the draw bar is done by compressing the spring with a nut on the topside, most likely two with the second acting as a locknut to hold the spring tension. i wanted to do something adjustable so i could get the most out of a 100mm bore air cylinder. the one i have will do a little over 1200lbs at 100psi and can go up to 145 psi max. i don't know exactly how much i can get out of my compressor beyond 100psi the shut off is 120psi. it's not a small compressor but it's only single stage with a 6.5hp motor and a 60 gallon tank.

I made a tandem cylinder in search of more ejection force with my drawbar:

Oh,and I use a silent compressor rated at about 350W / 0.5hp. There is no airflow requirement, so no need for a large compressor.
Pretty much identical to the big picture here: bambi-air.co.uk/products/md-range-silent-air-compressors MD 35/20

i watched a couple of video's on tool changers and i have seen a few where they go into a great deal design to keep the compression force away from the bearings? i don't get the concept there? the draw bar always has force on it to keep the holder in place. the only gain is the extra force for compression to eject your tool but the stroke is so small the the force would not be much more than the clamping force. the big thing i would say is to make sure the bearings in the spindle can take the load of a spring loaded draw bar to begin with and build off that.

the bearings i went with are 45 x75 x 16 mm on the angular contact and rated at close to 24,000 N on a dynamic load each. then i have two 40 x 68 x 15mm deep groove on the drive end and those are rated around 14,000 N each. pretty big bearings i would say for the size of the machine. seems like they should handle the force ok.

The clamp springs typically rotate with the spindle, so they do not add any load to the bearings.

yeah mine is the same way, just something i seen some talk about but made absolutely no sense to me. i am no expert on the matter but humble enough to have someone explain it to me in case i am missing something the way mine is build the eject stoke is taking any load off the bearings if it were transferred up the shaft anywhere. only place it can be is in the upward direction, bearing load ratings was just to be safe if i was wrong i'm just a chip slinger