New Greaseless Bearings Spin with 10 Times Less Friction | Hacker News Reader

New Greaseless Bearings Spin with 10 Times Less Friction

popularmechanics.com

65 pointsSwellJoe24 comments

ArduinoVsEvil made a video about this, explaining how the demonstrations gifs are pretty meaningless:

https://www.youtube.com/watch?v=uD7Lzv5fWhs

In general, the concept doesn't make sense, or at least it's being inadequately explained. It is probably a hoax.

I was thinking something similar. I immediately thought, "wait, doesn't a speed change imply a force on the bearing, and wouldn't that necessarily imply friction between the bearing and track?" Under load (and I've no idea how much), I just imagine this applying a subtle little ba-dump-ba-dump-ba-dump to the bearings as they spin around.

Their innovation seems to suggest that this is such a small change that it has less of a bump and more of a very slight shape change effect. So there may actually be something there - I've certainly seen stranger topological results. But until I see one after a few thousand hours in a high speed motor, I'll be on the fence.

> But until I see one after a few thousand hours in a high speed motor

Add some load to that as well. It seems pretty unlikely any indentations will last very long under stress.

I've replaced bearings on trailers and vehicles and they often have wear marks deeper than the indentations they showed in that video.

> doesn't a speed change imply a force on the bearing .... imply friction

Not necessarily. A speed change could also be because the track is longer. That's how trains stay on the tracks, the wheels are slanted, and if they turn off the track the length of the wheel changes, which changes the speed (relative to the wheel on the other side) and steers it back onto the track.

Feynman explains it: https://www.youtube.com/watch?v=y7h4OtFDnYE

> and if they turn off the track the length of the wheel changes

But is doesn't actually change. A different part of the wheel comes into contact with the rail, but that different part always had a higher linear speed. It just wasn't in contact with the rail until the turn.

But with these bearings, it appears that the balls themselves actually do speed up or slow down to maintain separation from one another. In that case, there must be a force coming from somewhere to effect those changes.

They explain the balls spreading apart as they accelerate rolling down the bump. What about the balls compressing as they go back up the bump? Wouldn't the balls constantly collide and separate?

To be charitable, I think the actual amount of the bump and spacing is very carefully calibrated to make sure that doesn't happen.

Really, it's almost like replacing the cage with a subtle bumpy bit in the race. And that seems less-crazy.

But without the grease, wouldn't it suffer the same as greaseless caged bearings? Albeit slower wear? Perhaps the wear is reduced enough that you'd be willing to replace them more often in exchange for the lower losses through friction. I dunno; I really think this just needs a (perhaps literal) ton of testing.

I'm always suspicious of claims of extraordinary gains from very old technology, and this is no different, but I don't understand the physics well enough to make a serious case for why. I figured someone on HN would debunk it, though I don't think that's entirely happened yet in this thread.

Then again, the people making extraordinary claims are the ones with the responsibility to provide extraordinary evidence, which they clearly haven't done. But, Popular Mechanics is excited about it, so it must be awesome. Those guys are all about bearings.

Anyway, this should be perfect for my perpetual motion machine. Less friction means it'll run for more forevers before it stops.

> Bearings work to help one surface slide along another by minimizing the points of friction between the two. Imagine sliding a book across a desk. There's friction at all points where the book is touching the surface. But if you put that book on top of some pencils, there's only friction where the book touches the pencils and where the pencils touch the desk. Suddenly that book is practically racing across the table if you so much as nudge it.

This description of why bearings work is terrible. If the advantage is just reducing surface area, then that would be a runner: pushing the book along the pencils long-ways. This actually does not reduce friction, since the equation for the force of friction does not include a term for surface area. (There are some real-world exceptions--runners on ice for example.)

Bearings work because they replace sliding friction with static friction. Static friction is actually higher than sliding friction (otherwise things would never stop sliding), but it never comes into play because the load-bearing surfaces don't slide relative to one another. This is also why antilock brakes tend to reduce stopping distances for cars.

Bearing cages do experience sliding friction, but not on load-bearing surfaces. Since the normal force is therefore much lower, so is amount of sliding friction.

It's true that in theory, eliminating the cage would reduce overall friction in a bearing. However bearings have to live in the real world, where materials are ductile or friable. Small indentations in the load-bearing race will only be durable if the load at that point is well under the yield strength of the materials. Otherwise the balls or the race will just deform and lose the advantage.

The friction under zero load is rarely a factor in a real design. You wouldn't use a bearing as big as the one in the video unless you had thousands of pounds of radial force, in which case it's the rolling friction that dominates. (Rolling friction is due to the balls compressing as they roll).

If you don't have large forces, you could just use a small bearing with correspondingly low friction.

Possibly there's some application that needs large bearings to handle high peak forces, but where it's otherwise spinning at high speed with near-zero forces and friction matters, but I can't think what it is.

Yo-Yos maybe?

Something like Aircraft jet turbines might qualify. They are designed to handel a mid sized bird stike or the loss of a blade which is way outside the normal force range. Though, I suspect they are heavy enough that rolling friction still dominates.

A lube free jet bearing could be a major plus for preventing Aerotoxic syndrome.

Or have cabin air intakes that are separate from the engines.

They're part of the engines because the air has to be compressed and heated before going into the cabin (in flight, it's 1/3 of normal pressure and -40C outside) which, it turns out, the compressor turbine is perfectly suited for.

The Boeing 787 Dreamliner does this. https://en.wikipedia.org/wiki/Bleed_air#Bleedless_aircraft

>> The friction under zero load is rarely a factor in a real design.

I remember as a kid I took my rollerblades apart, and 'cleaned' the bearings with WD40. The wheels spun so much faster when I simply spun the wheels on the skates. But I slowed down much faster when actually had the skates on and went down the street.

Turns out re-packing the bearings with new grease fixed the problems, even though the wheels didn't free spin nearly as long.

My understanding was that a retainer or cage for bearings simply allows you to use less bearings -- otherwise typically the bearings fill the cup and are all touching.

In other words I'm not sure I even buy the premise...

Also the grease used also has a huge impact on the speed of the bearings (for example a really light grease is fast while the heavy stuff is reserved for conditions that demand it)

I don't understand how this works, but the description made me think of shapes of constant width (example at http://mathsgear.co.uk/products/shapes-of-constant-width-in-...)

I think one could get the effect by slightly undulating the in- and outside tracks of a ball bearing and use balls that are slightly non-spherical. I expect making such ever-so slightly non-spherical balls to great precision would be extremely expensive.

This may (corrections welcome) be something similar, but with spherical balls, getting a similar effect by varying the point of contact of the balls with the outer track of the ball bearing.

Does it work for ball bearings? I wouldn't know. I do suspect that there will be more wear and tear on those varying points of contact, though.

Nice, but:

- If there is bump, does it not rub down over time?

- balls might jump differently when pressure changes (vehicle weight changes).

- How about vertical orientation when gravity is pushing balls in one direction?

wouldn't that graphene + micro diamond thing be great for a bearing?

edit: this http://arstechnica.com/science/2015/05/tiny-diamonds-wrapped...

Taking from tlb's currently-top-rated comment:

> Rolling friction is due to the balls compressing as they roll

So I'd assume these have extremely low rolling friction?

Ceramic balls and races are currently used in some high end bicycle parts. They have exceptionally low friction. See chris king bike hubs for an example.

As kids we used to make carts with ball bearings, here are some examples:

https://www.google.com/search?q=лагерна+количка

I'm not sure whether kids in other countries did them, but being in possession of 4 ball bearings was essential (they were not sold at the stores, or what did I know - I was just a kid, most of them we had to ... ahem steal them).

Reply on news.ycombinator.com