Ceramic Bearings vs Steel

As a few of you will know, I've finally opened my HS3540 after suffering the cut phase wire issue.

My 6203Z steel bearings are quite clearly notchy in rotation after only 2000km. Steel replacments £1.90 ea vs ABEC 5 Hybrid Ceramics at £16.99 ea.

Who uses Ceramics?
Are you having a trouble free life? How many km usage?
Have they worn out? If so what Wh abuse and after how long?
Are you seeing increased efficiency?
Are you seeing increased performance?

Way back when, Ceramics were the rage with inline street skating and boarding too. Alot of people including myself switched to ceramics and the difference was night and day. But that was under human power at barely 100W of effort if that. With HS3540 motors kicking out 1500W+ continuously, are we going to see such drastic effects or are they more a 'gimmick' or for people trying down the cost effective route?

Well after a bit of searching and general bearing knowledge from my previous motorbike history (several 100,000+ miles riding and maintenance):

Worst to Best
Standard unbranded Steel Bearings
Challenge/F&D
SKF/Koyo
SKF E2
Steel/Ceramic Hybrid
Full Ceramic skip to 2:45

From a price point perspective the SKF 6203ZZ E2 is £4.00, whilst the standard SKF 6203ZZ £3.99. The youtube shows an unloaded improvment using 9.5% less amps than the standard SKF. Whilst the Ceramic vs Standard Steel (not the greatest or most accurate clip) shows a 41% increase in rotational movement at a constant force.

Given that the difference in power consumption between a superb ball bearing and a terrible ball bearing is a fraction of a watt, why anyone would pay silly money to upgrade from a good bearing to a slightly better good bearing is beyond me.

The difference in modulus between ceramic balls and steel races means the race will fail earlier than in a normal steel ball bearing, due to higher stresses.

If you want to throw money at bearings, just get nice normal ones from a reputable Japanese or German manufacturer. Gimmicks don't help.

I agree and disagree...

However, reading this you can calculate what a percentage reduction in frictional drag will equate to in Watts.

Also, bearing failure rate is quite dramatic from standard bearings to high quality ones and even to ceramics. My '10 plate Blade needed the stock front wheel bearing replaced at 26,000 miles. I didn't know about bearings and just bought some direct replacments of the 'Challenge' brand. They lasted just over a month or 1300 miles. Put in Koyo's and they were still good when I sold her on at 38,000 miles. Anyway the other bike KTM 690E/R in sumo outfit would eat rear wheel bearings (and the axle but thats another story). Now the SKF bearings I replaced like for like lasted another 12,000 miles, so I got some hybrids not they are still going strong after winter (salted roads) and 10,000 miles.

Your failures were not due to the bearing material.
Maybe poor quality, but most likely incorrect installation, contamination(crap seals) , or insufficient lubrication.
Wheel bearings are a very low tech application for a bearing.

Hillhater said:

Your failures were not due to the bearing material.
Maybe poor quality, but most likely incorrect installation, contamination(crap seals) , or insufficient lubrication.
Wheel bearings are a very low tech application for a bearing.

Click to expand...

Speaking of crap seals, that is how bicycle parts manufacturers who were hyping ceramic bearings a few years back were getting favorable test results when comparing against steel bearings. The steel bearings had (effective) draggy seals, and the ceramic bearings had lower drag, less effective seals.

Unless the seating body isn't flush or there is excess wear on the axle, the bearing press I use is certainly capable. Normally on incorrect fitting the seating body starts to misshapen and repeated erroneous install and extracting esculating the problem.

The bearings used on any of my bikes hit either lack of lubrication or FOD ingress i.e. KTM BOAT use generally in rain or snow (it's more fun ). No need to scare potential DIYer from an extremely simple job! 2RS enclosed bearings are generally used for motorcycle use and ZZ seems to be the enclosure of choice for bicycle bearings, the former being removable for added lubrication/maintenance.

Also, to note, ceramics may not have a 'ZZ' proper seal, but due to the nature of the harder compounds you find any potential FOD is minimized to a fine dust or in wet weather, a paste. They are generally 1/3 lighter than steel, they are corrosion resistant (salty winter roads), smoother running, much better thermal properties and minimal lubrication/maintenance.

Now factor in heat as a degrader you can see here and here show that the grease in our steel bearings is only suitable for up to 98C and a normal operation temperature of 70C is expected. Now read further and it says 'If the outer ring is inaccessible, take the measurement at the housing and add 15° to 20°F to estimate the correct bearing temperature'. So that I assume would mean, anyone running a hub motor at 64+ C without the correct bearing could indeed become much more likely to have a bearing failure, if your not regularly regreasing. Oh wait ZZ bearings arent resealable...

So to sum up, if your bearing housing is operatiing above 64C and you want a trouble free life with no bearing failures, you would need to:
- Replace your ZZ with 2RS or similar replacable seals and regrease frequently with an appropriately like for like viscosity and heat tolerant lubricant.
- Replace your ZZ with Hybrid ceramics and run at a 6C higher operating temp, before egrease frequently with an appropriately like for like viscosity and heat tolerant lubricant.
- Replace with ZZ with sealed full Ceramics and forget.

In addition to the above, you should also use a bicycle frame made of bottle glass, which is surprisingly strong and stiff, and lighter than metal. Why wouldn't you?

Normal materials are normal for good reasons.

The real solution is to use any decent quality steel, 2RS industrial bearing. It's a hubmotor doing <500rpm, not a turbine...

In addition to the other issues with them, I don't see how ceramic bearings are corrosion resistant when the races are still steel?

There are some bearing cartridges made with both ceramic balls and ceramic races.

Where a steel bearing would become indented from a transient overload, an all-ceramic one simply shatters.

Bottle glass is a bit of an extreme comparasion in using a compound incorrectly, but yes fibreglass/carbonfibre are used for performance bicycles quite frequently. Similar structure, brittle on impact, but manufactured correctly can reap many benefits.

The ceramics used in bearing are much harder than your day to day pottery, however basic pottery excavated after hundreds of years stays in much better condition than quality forged steels etc...

Normal materials are normal for good reasons at 'Normal' operational parameters. Go beyond them and failure rates seem to creep up considerably.

Anyone pushing their motors harder than design spec risk bearing failure, which could result in a collapse which will in turn damage the stator, magnets and most likely fatigue the axle.

Buy cheap, buy twice. Or in this case possibly right off a $$$ motor .

I agree on a cost comparison High quality 2RS are much more favourable to ZZ, it does depend purely on what the user wants. I'm trying to get as much information on bearings and their differences as possible .i.e:

If your a weekend ebike plodder, bog standard bearings would be 'enough' to do the job.
If your a commuter HQ Steel 2RS would be 'enough' to do the job allowing maintenance.
If your a fast commuter HQ Steel 2RS E2 would be 'enough' to do the job just keep a good standard of maintenance.
If your a weekend wizz wanting a slight performance boost maybe a hybrid ceramic. You've spent decent money on the bike why not spend a few more $.
If your trying for speed trials 0-60 etc... why wouldn't you get full ceramics?

This is the kind of information that I'm getting at. Proper testing would be needed, but this is the information gleened from hovering the web so.

What significant advantages do you anticipate for ceramic bearings in the scenarios you describe?

AFAIK the main (only?) benefit of ceramic bearings in generally is their high-speed capability, which is an effect of reduced mass.

Chalo, my mistake, I thought the only ceramic ball-race bearings available in the common sizes would be hybrid ones. I see now that full-ceramics are available, but the prices are around 50-100 times that of a steel one. Ouch.

ccmdr has come up with the thing we really do need to discuss though, which is temperature. Unlike motorcycle wheels, our hubmotor axles have a direct thermal connection to the motor's primary heat source, so the bearing will see higher temps. In normal use they will commonly see higher than the 64°C, and much higher for those running moderate to high performance. Plus common ebike hubbies don't even have axle seals.

Is the 64°C limit a good number? I recommend against opening hubmotors any more than absolutely necessary, so can we re-grease from the outside? Is there a grease that will raise the temp limit significantly?

The scenarios are similar to how most I imagine would think, bearing in mind the cost involved.

This goes back to the 2nd post,

If the 0.2 Amp difference is comparable from steel to steel bearings with less frictional drag (not rotational mass), wouldn't it make sense that a ceramic bearing with the least amount of frictional drag will give a bigger power saving (cost out of the question, just performance). I'm not talking MW, but when you consistantly see a 30-40% increase in freewheeling speed from a rotating, surely that force (amps) to get you there must be lower, maybe not 30-40% less but even 10% is decent.

It's a bit far fetched until proven, but say you have a 14kw Adaptto at full power getting you to 50mph with OVS on standard steel bearings. Could you get 10% of that power back with full ceramics so effectively using 12.6kw to get you to the same 50mph. (Air resistance etc.. would have a part to play). So a minimum of 10% less force for the rotation of an entire hub/wheel assembly would mean either a faster or a longer ride.

That youtube is an unloaded comparision, so going on the theory of more load = more drag, surely having a bearing with the least friction would be an improvement?

Don't forget aswell we are on bicycles so when we actually using those pedals for something other than footrests or activating PAS, a ceramic bearing would make your natural cadence that bit easier.

Right, whilst thinking how to explain the difference again try using the simulator. You can adjust the rolling resistance. Also, I found this gem. They say you lose 6% rolling resistance with ceramics.

Now i think this completes the ceramic vs steel.
Is there an improvement over steel? Yes. 6%.
Is it much of one? Not really. 10W
Will I notice it just pedalling? 10W should be noticable
Cost effective? Not unless your doing time trials
Are they long life? Yes, varying in quality from 5-10 times longer than steel
Are standard HQ bearings suitable? Yes, aslong as you can maintain them and you use the hub at the manufactures rating. ZZ can not be maintained.
What bearings can be maintained? 2RS or any unsealed type
What is the 'Normal' operating temp of standard HQ bearing? 70C
When does the lubricant become less effective? 98C measured from bearing outer ring. Add 7C if measured from bearing housing.
Will running a bearing at 98C constantly will it degrade quicker? Yes, lubricant will need replacing. Race/bearing life will shorten.
Are ceramics more suited to a high heat enviroment? Yes, but you can get high temperature steel bearings.

Anymore guys, must be some other sources out there? Our hub motor are literally resting on these things!

*edit* @ John, sorry I mixed the normal operation temperature and the max effective temperature the lubricant operates. Also, reading the SKF page, they say to lube with high temperature grease, however an over lubrication can in fact create more heat. Also, using a grade of grease too thick can cause higher temperatures. Checking the High and Low service temperature helps.

If the bearings were dissipating ANY percents of a high powered motor's wattage, they'd get so hot they'd smoke the grease out. There is no performance advantage to be had with ceramics or anything else, because there is no significant energy loss in even a nasty bearing.

Tests at no load only tell you what the relative seal drags are, and those tiny losses don't increase with load.

You choose a bearing for its endurance, not performance. Anyone who tells you it makes a difference in performance is deluded, or trying to sell you something.

I'm told ceramic benefits on road bikes is very noticable for pedal power. Times that by maybe 10 for motor power, and that benefit obviously becomes less to even negligable. Imo unloaded friction tests have little correlation to actual loaded performance, but like I said I believe ceramic def lessens resistance there. There must be better tests for loaded bearings?

If they need less service and are more heat resistant, I would say bonuses there too. If they are sealed, chalo's argument of unbalanced wear to races wouldn't really apply as the bearing is considered 1 unit, and I'm under the impression they would last at least as long as any of steel.

But that said, are ceramic only in cartridge, and what motor hubs take them?

The losses in even the worst bearing are so negligible as to be unmeasurable in a complete system with a rider.

A fraction of an unmeasurably small value is unmeasurably small, however large that fraction might be.

Racers, gearheads, and other dorks can convince themselves that they "notice" a lot of things that can't be measured using instruments. If that makes them ride harder, well good for them.

Chalo you make me smile , your counters are very valid and bring up further questions. Placebo is a much better feeling than being left behind, you've only yourself to blame. No placebo, you can blame everything under the sun, but yourself!

We've already discussed that hub motors exceeding 98C are indeed loosing grease due to the excess heat and the viscosity becoming such that the 'dust' seals do little to retain it (which isn't their primary purpose). We aren't liberally smoking the grease from excess heat, but it's happening none the less.

I would highly recommend spending £30 of a set of Ceramic bearings for your skates and then say Racers, gearheads, and other dorks are convincing themselves. Pretty sure if your used to 6201RS then replace with 'Bones' bearings or full Ceramics you'd end up quite abruptly on your posterior :wink:. It was plenty enough shock going from ABEC 7 to Ceramics, quite literally amazing !

Your right though dust seals do also add to the drag, but the actual surface area of the bearing and the grease inside produce the most rotational friction. The heat perspective is a little off though, imagine the frictional drag of a 1 Tonne Weight of Asphalt on Asphalt compared to moving 1 Tonne of Glass on Glass.



More on the unloaded rolling resistance. I believe it has become common amongst the EV populus to ride and drive with a pulse throttle, for economic riding/driving i.e. getting to 30mph, coast until 25-27mph then throttle to 30 mph again. If you accept and take into account that most of the tests seen are unloaded and the Ceramic bearing do in fact roll considerably longer wouldn't this effect your EV range fairly considerably aswell?

Load on axle bearings includes bike and rider weight, plus chain tension from pedals, plus dynamic loads like bumps and cornering. Motor torque doesn't get applied as a force to the axle bearings.

Motors get hot because of electrical losses, not mechanical friction.

All three pulleys are fixed to 5Nm of torque at the axle bolt. Get 5Nm extra from your hub?

[youtube]r9EiX1RZDik[/youtube]

Don't bearing attach your rotating hub and the axle? Then the weight of the entire vehicle and passenger are transferred through the tiny surface area of the bearing, which are the only part of the system that allow any rotation or forward momentum. Torque from the hub is allowed movement because of the bearing surely?

These guys did a load (Bike and rider) Terminal speed test with avg 9kph gain (max 12kph) from just gravitational pull.

Why can't they do that test on even a 100lb loaded bearing though? Personally I think ceramic would still shine, but comparing losses unloaded is likely night and day to loaded, and imo some sales 'creativity'

Nobody answerd my ? - most hubs setup to take cartridge bearings then?

Both the Cromotor and HS3540 have cartridge bearings, although I'd imagine most hub motors would due to the axle width limiting the minimum I.D. for a bearing. Best way to know is to check, its the only way to know your bearing size as very little information is published.

Load on axle bearings includes bike and rider weight, plus chain tension from pedals, plus dynamic loads like bumps and cornering. Motor torque doesn't get applied as a force to the axle bearings.

Motors get hot because of electrical losses, not mechanical friction.

Is this a time vortex? Noone disagrees with this, I don't think.

I do think ceramics likely superior, but already voiced imo what needs to be seen for confirmation.

Motor torque certainly loads the motor bearings. What else transfers the drive force the motor generates? It is only connected to the frame through the bearings.

This talk of ceramics is funny. I also thought they were a joke until I started playing with them on high speed spindles. They have their place for certain. If you have plastic ball separators (cages) in your quality steel bearings, you could swap out the balls with much higher quality ceramics easily and play with them for cheap. Many common sizes on ebay http://www.ebay.com/itm/220966954701?_trksid=p2060353.m1438.l2649&ssPageName=STRK:MEBIDX:IT Grade 5 is many times better than the standard abec or the industrial rated stuff. Worth the few bucks just to play with a few.

For slow speeds such as hubs it is over kill for certain. Once you play with them, it is hard going back on anything that spins fast however.