i was not replying to you.
edit: I do not think adding a sprag will be beneficial. This is something that was discussed before, the principles are different. You aren’t trying to spin this thing backwards. For it to free wheel and drive in the same direction would be impossible.
In order to use a sprag for this it would have to be held to the shaft. That would add weight.
I think you are looking at a sprag incorrectly. The whole idea here is to mechanically move a shaft to speed, then have the shaft increase speed on its own, “freewheeling” in the sprag- SAME direction.
That is exactly what a 1-way Sprag does, and it would not have to be held to the shaft during any movement other than mechanical speeds. During freewheeling, it is independant, and they would certainly need to be balanced independantly as such.
I know what a sprag is and how it operates…but how do you propose to move the shaft if it is not coupled to it? And if it is coupled to it it would add mass to something that needs to be moved. Whether its a gram or a ton, its mass
A 1-way sprag type roller bearing would be able to “grab” the bare shaft itself, making the shaft surface itself the bearing surface. This would make the free-wheeling rotating assembly otherwise identical to one without this setup in all instances whee it would matter most.
Pretty simple answer I thought. :gotme:
Most sprags are contained in a bearing…which means there is an inner race…which would have to be pressed onto the shaft.
exactly- MOST. But they do exist in other forms, and in this particular example, there is no inner race.
ok i suppose.
However at that diatmeter then you are probably looking at disaster. Most turbine shafts are about 1/4"-3/8" approx diameter
Exactly. All these ideas are nice but you guys are not thinking in terms of scale. Applying automatic transmission or even bike parts to something as tiny as a turbo is not very practical.
A sprag clutch for example:
Needs a good deal of running surface area at the rollers to transfer the torque from one shaft to another… IE its needs to be of a suffucient diameter. Try applying that to a 3/8" diameter turbo shaft (without a significant mass increase at the shaft)
Second: To start and spool an impeller assembly from a standstill or slow speed to boost speeds will take a decent amount of torque. …hence the crazy planetary gear sets in Centrifugal superchargers.
Now the smaller the diameter shaft(or gear) you are working on, the more torque is required to get it to the same speed. Trying to apply that to a 3/8 diameter shaft will be tough if not practical unless you find more surface area to transfer the torque or effective shaft diameter to reduce the torque required to accelerate the impeller assembly mass.
CVT setup would use a pulley on the shaft.
mind you this is all theoretical and impractical but with enough thought and experiementation (read Money) I’m sure something could come of it.
However, if it were practical, I’m sure Garrett would have made it already 
It wouldn’t be practical in most situations, but for those with a deep pocket, I could see it working well. It’d be great for those big turbo 4 cylinders.
Actually my idea was in terms of scale, if anything, it would need to be scaled UP, not down. The particular Sprag bearing I am refering to links to a 1/4" shaft, transmits about 5-10 lb/ft of torque easily, and has a normal differential speed of 35,000 RPM. These would not need to transmit the full load of turbo spooling, only support a median-level turbine floor speed.
Think RC Car.
Now think harder.
And what is the rotational inertia of said shaft/assembly compared to 2 turbo impellers on a steel shaft?
RC cars use lightweight plastic parts and gearboxes/gears… neither of which apply to our freewheeling turbo idea
LOL, wrong application entirely man, they are used in the crank starters of some of the higher-compression 1/8 scale gas cars- which require 5-10 lb/ft to crank due to compression, not so much rotational inertia, and freewheel at 35,000+ RPM when running. They are TOTALLY different applications, but I am looking at a single component only.
Trust me, I am on the same page you are. If you read the specs I listed, the capability of the bearing I suggested it right on par with this type of application, albeit slightly scaled down. The mass moment of inertia of the shaft is obviously in question, though the torque required to twist said shaft of course depends on both its MMOI and the rate at which you are trying to increase its speed.
Before any analysis can be done, someone would need to find out the MMOI of an average turbine/compressor/shaft assembly, and guestimate the max rate of increase in shaft speed before the assembly maxes out mechanically and begins to freewheel from the turbine generating its torque.
show pics. i like pics
Only pic of one I could find…
This one is roughly 12mm x 12mm, and fits on a 6mm shaft.
this is a fun thread. I wanna get a centrifugal supercharger and see if i can make a CVT style gearbox (modified snowblower belt drive maybe) to drive it. 
i already have homemade centrifugal supercharger project that i started in HS (11 years ago lol) and never finished. I found some bearings (ceramic) that could take the loads and speeds im looking at but i could never find the compressor map for the turbo that i was using. guess ill have to guess.
Brian
I have been wanting to build a CVT system for some time.
all you guys in here talking like you think you’re engineers and know about bearing speeds, yet everyone is talking about RPMs and not once has anyone mentioned linear (tangential) speed. haha.
I agree.
However,
I do not claim to be or have ever been an engineer, but I do not think that you need a piece of paper to use your brain. Some things just come with the job territory (advanced knowledge of things available to you and how to properly apply them).
I can understand your POV though, I feel the same way sometimes on car forums.