Centrifugal supercharger, planetary roller traction drive type - annulus material?

[BrianGar]

Hello all,

Im sure many are familiar with this type of charger. It uses 3 traction rollers instead of gears in a planet arrangement. The main annulus ring is heated, and fitted over traction rollers, it shrinks and provides the drive to rollers and onto center shaft. The annulus ring deforms ever so slightly as it rotates around the rollers. It is to be of a ground 'high tolerance' internal finish.

Any Ideas on what steels to start looking at first for the initial prototype/round of testing. Perhaps similar to what a harmonic drive flexspline is made from would be a good choice. Anyone have a call out for that?

I do realize, that the post machine treatment(s) will have an effect on the parts 'spring' behavior depending on treatment processes/times.


Thanks,

Brian,

 

[MikeHalloran]

Rollers, no. McCulloch superchargers I have seen in photos seem to use ball bearing parts, with axial preload by springs. I'd guess all parts except the springs to be made of 52100 bearing steel.

There's another centrifugal street supercharger that's been on the market for a few years that uses a timing belt and pulleys to do the overdriving. Sealed ball bearings on the shafts, no lubrication concerns for the speed increaser.

Speaking of traction drives, if you are determined to go that way, I'd suggest Dexron ATF for a lubricant, rather than engine oil. Much better for traction drives.



Mike Halloran
Pembroke Pines, FL, USA

 

[BrianGar]

Mike, Im familiar with the systems you mentioned. Im semi hell bent on the traction roller type as its pretty basic internally when you see it. Ive recently met a great guy in the Uk that chromemoly sprays parts for me. I use him lots when building custom clutch lsds, as I spray the 4340 clutch stars after I surface grind them flat to give good wear properties, and also to hold the lsd oil better - they hold up extremely well.

Do you think this would be a good idea on the traction rollers to hold oil also or would a ground finish suffice?

See link below for traction charger internal parts,

I can hear readers saying just buy one, and save all this r+d time, but I need 7, all going on the same base engines for all close friends/members - they are coming in at around 14k dollars for the 7. I have all the machinery/techniques/time/cncs, and Im sure I can make the 7 for half that. Im buying Garrett compressor wheels and snails(170dollars per set), and there isn't a whole pile left if you have those bar oil pump which can be got off the shelf also.
Have heat treaters on hand too.

Scroll down a bit,

http://passionford.com/forum/genera...78-not-good-for-supercharged-fiestas-sts.html

Thanks,

Brian,

 

[MikeHalloran]

I'd expect a sprayed surface on a traction roller to spall pretty quickly because of the Hertzian stresses.



Mike Halloran
Pembroke Pines, FL, USA

 

[BrianGar]

That is a good point now that you mention it. My diff plates see no such rolling stresses.

Brian,

 

[TVP]

I agree with Mike on both counts-- 52100 is probably a good place to start for roller contact conditions, and any kind of coating will have insufficient durability. You can purchase 52100 in the annealed condition, so it can be initially machined, but then it will need to be hardened and finish machined. Do you have access to grinding and polishing equipment?

 

[Tmoose]

What size are the parts you are looking at?

I'd be looking at buying commercial roller bearings to harvest the components to use as-is or rework, expecting them to be cheaper, available quicker and maybe even more reliable.
Harley Davidson rod and main rollers and races are available individually. 3/16 and 1/4 inch diameter of various lengths are available in "standard" and even oversize in 0.0002" increments. The races are undersize ID requiring re-finishing.

Harley "big-twin" rods run on a 1.249" crankpin, 0.1875 (std) rollers in 2 different widths, and the races have plenty of meat to finish at ~1.625.

http://www.debrix.com/v/vspfiles/photos/12-2556-VT-2.jpg

 

[BrianGar]

Tmoose, I had exactly the same thoughts when I awoke this morning. For the rollers, I could use needle bearing inner rings, and for the annulus, I could possibility gut a roller bearing for its outer shell as you mention. Infact, Im wondering if I could use complete heavy duty roller bearings for the rollers - just mount on shafts(Look at ford link above for a better understanding)
Failing that, the inner ring races alone would provide great hardness headway mounted on live 4340(or something)rollers sitting on ballraces - as in link.

http://simplybearings.co.uk/shop/ad...463_22&23_230=-1&x=0&y=0&extra_field_filter=1

For the record, the annulus over all diameter is around 140mm, but I yet have to work out surface speeds, and over all ratios.

Thanks again all,

Brian,

ps, no way connected with the ford forum, it just came up in an image search.

 

[BrianGar]

''Infact, Im wondering if I could use complete heavy duty roller bearings for the rollers''

Should read complete heavy duty needle roller bearings for rollers - as space is semi tight.

Brian,