You've got yourself quite a challenge there... Seals that will contain oil will create drag. What's the actual setup? A labyrinth seal and clever scavenging system might be the way to go. If not, would an air bearing suit your application?
The actual setup is a horizontal shaft, the housing is a tube closed on both ends with bearings. There is nothing else in there between the bearings other than the shaft.
Thank you for suggesting a labyrinth seal. I haven't considered it before. Google brings up lots of generic info and some products for very specific applications, but no DIY labyrinth seals. I could perhaps 3d print it from TPU(rubber).
If you or anyone else has some links to DIY labyrinth seals I would love to see them for inspiration.
Air bearings should present about the Lo9oo9w) west possible drag. There was a You Tube video a few years back that was about building an ultra precision machine and as I recall, he used air bearings.
Air bearings are a great idea, but this is an unpowered, portable device So I'd rather not have an air line. That's why I'm using ball bearings.
You can use shielded bearings to reduce drag, or make a labyrinth as mentioned above. To have the least amount of drag, you want a low viscosity oil or grease, and an NLGI of 1 or 0 will have a more fluid base soap as well that will help reduce drag until the grease gets pushed out of the way. The oil in the grease does the actual lubrication, the base soap just holds it in place until it is needed. Something like Kluber NBU15 is a low viscosity grease and would probably work well, but very expensive. Ceramic ball bearings also help reduce drag and allow you to use a lower viscosity lubricant. You need to be careful not to go too low with the viscosity, otherwise you will get metal to metal contact and the bearing life will be very short.
Those greases (isoflex topas etc) are expensive, but are they sufficiently low drag? Have you (or anyone else) physically spun a bearing lubricated with one in your hand? If yes, how long did it turn (and how big of a bearing was it)?
I thought shields don't really prevent stuff like dust or liquids getting in or out the bearings?
Is there a way to tell I'm using too low viscosity oil? (a different way than having the bearings fail)
Another method to reduce drag is to use an oil mist or drip system. If you can slowly drip a low viscosity oil into the bearings, you might get what you need. The best way to get long spinning duration is to have a high inertia of your rotating shaft, ie. the biggest mass as far from the center of rotation as possible like with a flywheel. If you can incorporate that into your design, you will increase your spinning duration.
Thank you for those options. Ideally I'd prefer to find a solution that doesn't require modifications( such as increasing the size, adding gravity fed oilers etc.)
DIt seems to me that a loaded bearing that is lightly oiled to reduce friction needs to be continuously oiled to avoid starvation. I would think of a way to continuously supply oil. Hopefully the shaft is running horizontally, in which case oil cups and a collection tray might be enough. But I suspect a better solution than that would be an oil bath covering the bottom portion of the bearing. You'll still get leakage past the seals, but that leakage will tell you it's working.
Rick "grease is just oil emulsified in soap--the soap is there to keep the oil there" Denney
This is what I'm thinking of doing next. I can add an oil port to the housing. I haven't done it yet, because there are parts that shouldn't get oil on them past the bearings. However if I can make a good labirynt seal that could work.
FreeGoogle7bottom bracket is low drag .
How low drag is it? What's your experience with using it
Regarding ceramic bearings. 100% ceramics are quite expensive so I looked into hybrids (ceramic balls, metal races). Unfortunately they too need lubrication so I'm back to square one with them.
Oil, then stuff the seals back in?
That's what I sort of did. Well with just one seal per bearing leaving the middle space open (one seal per bearing is sort of acceptable - drag wise, two cause too much drag). The shaaft is housed in a tube with bearings on both ends and nothing else in the mbiddble. If those seals manage to keep gthe oil in I might add an oil port to the tube and keep some very light oil there. I'm just not very hopeful those seals will hold. We'll see. I really just need enough oil there so when the shaft is not rotating the lowest ball is at least a little bit submerged to pick up the oil when rotating.
