Large bearing with grease or oil? 1

[corrosionman]

A client has a 6 ton rotor running at 500 rpm and supported by an continuously oil lubricated spherical roller block at each end with bearing type 23072 in the block. The Drive is a direct drive pulley with 20 Vee belts linking it directly to a 350 Kw motor so there is a huge tension load on the bearing.
there has been repeated failures of the bearing due to oil supply interuptions so the Engineer is intent on changing to a system where the bearing is simply greased from time to time. Can grease really be as effective as a permament oil flow for heavy loads at slow speeds ?
Any comments much appreciated please. CM

 

[diamondjim]

500 rpm is not slow.
At 90 rpm greased tapered roller
bearings work well in automobiles.
I do not know the upper limit.
Grease does work well at slow
speeds and heavy loads.
I assume that racing cars use
greased bearings.

 

[electricpete]

Some rambling statistics for p/n 23072

SKF identifies max rpm for grease lubrication of this bearing is 670 rpm and max rpm for oil lubrication of this bearing is 850 rpm.

350kw
T = 350,000 N *m / sec / (2*Pi*500/60 radians/sec) = 6700 N*m

The dynamic load rating is
C = 2,750,000 N

d = 360 mm
D = 540 mm
Dm = 450 mm

Dm * N = 225,000
Dm*N=225,000 @ 500 rpm => SKF guidelines require minimum viscosity at operating temperature ~ 17 cSt

The D*N and viscosity are well within the range of most general purpose greases assuming there is no abnormal heating.

Based on the above I think grease is probably an option.

What is the sequence which occurs when oil is lost? Possbily other measures can be put in place to mitigate this scenario - temperature alarm to alert operators....larger oil reservoir to give larger response time?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[diamondjim]

Thanks electripete.
I have not seen that formula
before. Interesting.

 

[cloche]

I know of a large class of very high-performance lubrification "fluids" that would probably fulfill your requirements: they are called "Extreme-Pressure Lubricants" and have the consistency of "something" between grease and oil. They are applied just like grease (can not be pumped in, for example), and must be tailored for the surfaces' materials because most of their action is due to molecular interactions. These interactions ensure that the lubricant film can not be "stripped away" under extremely high loads. They are not suitable if the normal load (radial load or tension load as you may name it) is below a certain threshold.

 

[corrosionman]

Electricpete
The causes of oil failure are numerous but over last 5 years there has been about 8 occasions when it happened and the bearing goes red hot and causes $30,000 of damage.
It is a very hostile enviroment with lots of dirt and corrosive chemicals. Pressure switches in the oil lines should switch off the drive, but they sieze up. Temperature switches likewise. The latest idea was a dipstick which was displaced by oilflow. It also failed to work.
I,m close to this job but not on the "decision making team"
however as a Proffessional Engineer this latest development of "grease not oil" does intrigue me.
I am appreciating the comments from everyone. CM

 

[unclesyd]

We had several very large(heavy) fans recirculating hot gas in process which we ended up lubricating the bearings with oil mist after trying different approaches. The system was redundant and was linked to the process control by both temperature of the bearings and oil pressure of the system. If there was an alarm from either sensor there was a standby "Manzel" lubricator with an added hand pump.

On other high loaded bearings we stayed with oil and used a water cooled pillow block bearing housings.

 

[electricpete]

Some more rambling thoughts. Take it fwiw

From above:
Torque = 6700 N*m
Assume sheave is 0.1 m radius (smaller assumed sheave gives higher force)
The force associated with the torque is
F = Torque / radius =67000 newtons

For V-belts, the force on the shaft is 1.5 - 2x the torque-producing force.
Fshaft = 2x67000 = 134000 N

The inboard bearing may see up to 2*Fshaft assuming equal distance from sheave center to inboard bearing to outboard bearing (outboard bearing would see 1x Fshaft in opposite direction from inboard bearing).
Finboard = 2x134000 ~ 270,000 N

Fshaft / C = 270,000 / 2,750,000 ~ 10%
Radial loading is ~ 10% of rating.

From a radial loading standpoint, the bearing appears to be lightly loaded. However the overhang distance from pulleys to bearing creates a moment. If the shaft is rigid there is no moment loading on the bearing. If the shaft is flexible, moment loading can affect the bearing. I don't know how to evaluate that or any ratings for that.

Friction coefficient for spherical roller bearings 0.002 - 0.0025
FrictionMoment = Finboard * mu * r = 270,000N * 0.0025 * 0.180m = 120 N*m

Friction heating = 2*Pi * N * FrictionMoment = 2*Pi* (500/60sec) * 120 N*m = 6300 watts ~ 6kw.

I don't have any formula handy to evaluate that but it seems like that much heat in a grease lubricated bearing would cause overheating. Maybe SKF or someone else on this board can help work through that question.

Oil mist is often ideal from equipment life standpoint, but I would think you would have even less time to react upon loss of lubricant supply. From that standpoint, if they have the same trouble keeping oil mist going that they have keeping oil flow going, they would be in worse shape with oil mist.

The dirty environment you mention creates one more possible drawback for grease - more potential for introducting contamination during periodic manual regreasing operations although better for sealing the bearing once you safely get the grease in. The concern might be addressed with a continuous regreaser system. Double-shielded might be an option? I don't see double-shielded applied on many large bearings and I'm not sure why.... maybe just the assumption that they have to be replace periodically which would get too expensive?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[unclesyd]

Grease lubrication has it drawbacks the same as oil based lubrication. Aside from the above mentioned contamination problems, over lubrication causes a tremendous number of bearing failures. The first device help alleviate some if not all of the over lubrication while a device like the Bearing Isolator will help with the contamination. The isolator can act as the double seal mention by
electricpete

http://www.lubricationventblock.com/index.htm

http://www.inpro-seal.com/

 

[sreid]

As others point out, grease has it's own set of problems. It seem to me the the oiling system should be completely repaced. Safety device that fail to work are the same as having no safety circuits at all. For critical applications multiple pumps are used, multiple sensors that "Vote", uninterruptabe power supplies, huge oil filters that never plug up in a lifetime. etc.

 

[ccw]

Hi corrosionman,
I would go with sreid's suggestion. Electricpete's 6KW per bearing.. does it include preload? Also, the 20 vee belts seems severe. The preload to keep 20 vee belts tamed down has got to be substantial. Look at toothed flat belt and sprockets, for higher efficiency, reduced preload required, eg. reduced static bearing loads.

 

[electricpete]

My 6kw is based on using shaft radial force calculated as 1.5 - 2 times the torque-producing force, which is a number intended to incorporate typical belt tensions for V-belts. I see this same number in a variety of places including the NTN catalog and Richard Nailen's "Managing Motors". They do not consider the number of belts so I assume it remains relatively true for total force regardless of number of belts. The radial force was then multiplied by friction factor to give heating. Note the 6kw only applies to inboard bearing, outboard would be roughly 1/2 of that. 6kw does seem extraordinalrily high to me. Interested to see if anyone sees any errors in the calc.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[rmw]

Corrosionman,

With grease lubrication, the type and grade of grease and the precaution of preventing the mixing of dis-similar greases are your two most important considerations.

The bearing mfg's recommendation for the grease type and grade should be followed closely, if the grease option is chosen.

Your biggest concern is that the grease can be mixed during the re-greasing process. If 'Bubba' is out in this nasty plant environment re-greasing devices that take non-bearing type grease, (which is designed to stick on surfaces, often surfaces at elevated temperatures, while bearing grade grease is designed to 'flow' into, around through, and out of bearing surfaces,) and he decides that a few pumps into the bearing would be good, he will cause bearing failure due to improper mixing of the greases. "Been there, done that" and it is not pretty.

Another concern is that some slick grease salesman comes into the shop and gives his demo where he takes some bearing grade grease, and whacks it on a flat metal surface with a hammer, and it naturally 'flows' all over the shop, because it is designed to flow, and then he takes some of his good high temperature 'chassis' grade grease which is not designed to flow and whacks it, and when it doesn't splatter all over the shop, convinces all the mechanics that it is the best thing since sliced bread, and they go to greasing everything with this stuff. "Been there, done that" too.

I have seen blued (failed) bearing roller and race surfaces with lots of high temperature grease still in them in close proximity to the blue metal parts because the grease could not 'flow' into the bearing where it was needed.

The last concern is proper regreasing intervals. As bearing greases work, the oil separates from the soaps that carry the oil. Regreasing forces the spent oils and soaps out of the bearing and replaces them with new grease. Over greasing has been adequataely discussed previously in this thread.

I am not afraid of grease in the application you mention, if the proper program is set up but like Unclesyd said, if you don't take the proper precautions, you will still have problems. If maintaining proper lubrication to these bearings (especially at 30K per failure) has been a problem with oil lubrication, I suspect that grease is not going to be your 'silver bullet' in and of itself.

rmw

 

[electricpete]

rmw's comments about the high cost of failure make me think....

Isn't it reasonable for a critical piece of equipment that we should not make design changes without assistance of a competent OEM or similar expert. Therefore it seems reasonable to insist that you have the original OEM or a bearing OEM evaluate the situation before you institute grease lubrication. One known fact is that grease bearings are not as good at dissipating heat as oiled bearings. I think that is a focus that needs to be evaluated.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[Hippo41]

The ndm (speed in rpm x average diameter) works out at 225 000. This is at the upper level of normal lithium or lithium calcium EP2 bearng greases.

I would suggest you look at something a little bit "above average". I have used FINA CERAN HV in sugar mills which have similar operating conditions. I think (from memory) that it had a limiting ndm of about 400 000 which gives you a bit of leeway.

Lester Milton
Telford, Shropshire, UK

 

[electricpete]

First I wanted to correct my comment about loading.

I forgot to include rotor weight.
12 tons = 24,000 pounds = 12000 pounds per bearing.
Weight of 12,000 pounds is weight of 6,000 kg is 60,000 Newtons.

Load = reaction force + weight = 270,000 +60,000 = 330,000N

Load/Rating = 330,000 / 2,750,000 = 12%

NTN catalogue 2200E table of shaft fits (page A-57) calls light loading P<6%*C, Normal loading P=6%*C and Heavy load P>12%C. So this application would be considered heavily loaded.

Not that the word "heavy loading" means a lot, but it implies that this is a higher fraction of loading than most bearings would see.

Now, looking at the D*N ratings of grease, they do not consdier load. Likewise the maximum speed rating of the bearing with grease lubrication criteria does not consider loading. Maybe they consider worst-case loading? I don't know. It seems like one should be caution in relying on D*N or speed rating for this situation of heavy loading.

If anyone can explain to me why grease D*N ratings and bearing speed ratings do not consider loading, please explain it.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

One more thing on the load calc - I assume your slow-speed sheave is probably much larger than 0.1 m which would reduce the loading. Also not sure if rotor weight and sheave loading should act in the same direction.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.