Centrifugal fan thrust load 2

[Abualnassr]

Hi.

I have a over-hung centrifugal fan with :
Volumetric flow rate = 4400 Nm^3/hr
Dp = 0.07 Bar
Speed = 3273 rpm
Power = 25 KW

This fan is supported with 2 double-row roller bearings.
How can I calculate the reactions on those 2 bearings and make sure the bearings are selected properly ?


Regards,

 

[dinjin]

Submit your application to the Bearing Manufacturer.

 

[Abualnassr]

Actually, i posted this thread to get help on the failure of the outboard roller bearing. I don't know if this bearing can resist the thrust force.

 

[Tmoose]

One approximation might be discharge pressure X area of back face of wheel. Some fans have mini vanes on the back side to reduce/balance that pressure.

Are they spherical roller bearings?

Either the inboard or outboard bearing should be "fixed" to serve as the thrust bearing. But not both.

What was the bearing "failure?"

 

[Abualnassr]

both bearing are self-aligned spherical ball bearings.
The failure was in the bearing in the drive side of the shaft.
What happened is that the bearing balls fall down and stuck up together producing so much heat that actually made the bearing casing become black.
After replacing the bearing and sent the fan to the end user, it ran for 3 days and some problem occurred again!!

Any help will be appreciated.

Regards,

 

[dinjin]

Are you saying that there was not a cage separating the balls in this bearing? If it had a cage, did the cage break allowing the balls to bunch up together?

 

[Abualnassr]

yes dinjin, there was a cage and it did break and let the balls to bunch together.

 

[dinjin]

Was it a full circumferential cage with just a break in it or was is a 3 or 4 piece cage made out of segments or was it a full 360 degree cage with no break in it? What was the cage material?

 

[Abualnassr]

The bearing is SKF 22213 EK.

I don't know the material of the cage, but it was 1 piece as i remember.

Now, i know this kind of bearing can take some of the thrust load. But what i want to know is how to make sure that the thrust load limit of the bearing is more than or equal the actual thrust load of the fan ?

 

[dinjin]

That part number indicates it is a double row spherical roller bearing with a tapered bore and not a ball bearing.
SKF must have tons of application engineers to solve your
questions.

 

[Abualnassr]

Well, i'm sure SKF have tons of engineers that can solve my problem. But, i want to solve it myself. This is why i'm spending so much time to figure out how to calculate the radial & axial loads on the bearings.

So, i'll be so grateful if any one can explain to me how to calculate those loads.


Regards,

 

[dinjin]

For a starter, I would recommend T.A. Harris book
Rolling Bearing Analysis. From the references in his
book, you can go on to other pioneers in this field.
Combined bearing loading section is a must.
You should find this exciting and a new adventure.
The different application catalogues from the manufacturers
will be some help as well but limited in scope.

 

[Tmoose]

skf has an online calculator. enter bearing, radial and axial load and it cranks out the "equivalent" load and L10 life.

http://www.skf.com/skf/productcatal...02213&Cr=193&C2=313&C0r=216&C02=432&Fr=4&Fa=4

If this is a tapered bore bearing, installation includes reducing internal clearance a prescribed amount. If the clearance is reduce too much, heat generation will be excessive, initiating thermal runaway, lubricant failure (generally indicated by cage failure) and could well result in the failure and short life you encountered.

 

[Abualnassr]

Thanks Tmoose for your reply.
The problem is that i don't know how to calculate the actual thrust and radial forces in the fan.
I know the suction & the discharge pressure and i have all the dimensions, but don't know how to calculate!

Any help?

 

[Gaylord01]

Hi Abualnassr,

I work for a fan manufacturer, and I work almost exclusively in Imperial units of measure.

I want to convert your fan performance data into Imperial units, but have no idea what the "N" prefix represents in your volumetric flow rate (4400 Nm^3/hr).
I disregarded the "N", and converted 4400 m^3/hr into 2590 ft^3/min. This seems to be a very small value; not correct.

I converted 0.07 bar into 2.07 inches of water (in wg).

I converted 25 kW into 33.52 HP.

If these values are correct, it appears to me that the bearings are very lightly loaded. It is likely that there isn't enough load to properly engage the rollers. The rollers then skid between the inner and outer races instead of rolling, leading to pre-mature failure. (It's more common than one might guess).

Calculating the loads:
Axial thrust = (Total fan static pressure) x (cross sectional area of fan inlet) This is only applicable if the fan wheel has a backplate. If no backplate, the thrust load is negligible, and can be disregarded.
Radial load: The radial load on each bearing is determined using a simple beam diagram. There are two supports and two un-equal loads, one beyond each bearing. One load is the weight of the wheel and the other load is the belt-pull and weight of the sheave. If you have an Arrangement 8 fan (direct drive) there will be virtually no load on the fan outboard bearing.
The simple beam diagram to use is the "Beam Overhanging One Support - Concentrated Load At End of Overhang". It will show a downward force due to the load adjacent to a bearing, and an upward force on the other bearing due to the pivot across the first bearing. The calc's will need to be run twice; once for each load at opposite ends of the fan shaft, and the results added together. Keeping track of positive and negative values is critical!

Best wishes!

 

[dinjin]

4400Nm is Newton Meters is 38943.3 pound-force inch

 

[stgrme]

Nm^3/hr is Normal cubic meters per hour, not Newton meters. This measure is not used much anymore. However, one (1) Nm^3/hr is equal to 0.625 SCFM (Standard Cubic Foot per Minute).

Best of luck!

 

[mayt4u]

Are these bearings in a pillow block? If so, has a proper alignment been preformed? Double row bearings can handle 30% thrust load of the bearing load. Self alignment does not mean that anyone can change and run these bearings. How was it changed? The old one cut off? Were measurements taken to ensure proper fit? Was the pillow block returned to the proper match mark condition? Was the new bearing properly packed with the right grease? Was the grease level proper when running? Did you take temperture readings every 5 mins for an hour as the new bearing was run up? How clean was the environment when the bearing was changed? Was the bearing in proper storage before it was installed? How long was the bearing in storage? Was the bearing subject to vibration in storage?

Kind of get my drift. You cant blame the fan for a poor fit or a bad install. If you follow the proper proceedures to install a new bearing I don't think you should have any further issues with it.

Robert

 

[diskullman]

Did anyone look at the limiting speed for this bearing? I saw one manufactrer's rating at 2900 RPM limit for grease lube, 3800 RPM for oil lube. High speed fan applications are not a bearing's best friend. Fans are difficult to keep in perfect balance.

Russell Giuliano

 

[FeX32]

What is the area of contact on the bearings?
A fundamental relation I would start with is: Power=force*velocity
which then can turn into:
P=deltaP*A*V
V*A=Q (flow rate which you have)
So, Power=deltaP*Q
Figure out your deltaP and this can be an approximate pressure on the bearings. This is just a CV analysis.
(note that this is assuming the maximum power out of the motor and no losses)
Hope this helps,

Fe