Theoretical bearing life is based on the bearings running in ideal conditions.

Life isn't like that as contaminate usually manages to bypass the seals and destroy the bearings. ■

Hello Endri ..

I have clients who insist on an L10 idler bearing life of 75 to 100 000hrs..

The trouble is they completely ignore the fact that the life of the seals is generally about half these values. Once the seal is shot, the bearing will fail prematurely.

Also, the shells don't last if the idlers aren't properly aligned

Cheers

LSL Tekpro ■

Hi

Lubrication quality is the key. If you look at the formula for L10 life, there is a lube factor.

With other running gear, you can get better L10 life if you have very clean lubes....

Note that this is all statistical.

Have you pulled apart the brgs - even before failure. At failure, it can be hard to asecertian the causes.

For rollers, you can go to better seals at higher cost etc.

The trouble with making changes is that it takes a long time to get data and analsise them and you have to keep good records of what went in where/when, run time hours.

On our site, I found that our rollers failures were higher on a conveyor on the downwind side of an open stockpile - all make sense!

A life of 3 years does not sound bad but depends on run time pa. Look at your maintenance costs per conveyor, see if there are trends etc.

Thanks ■

I thought the subject had been discussed on prior postings.

Read SKF or Precismeca Design Manuals, or Shigley Mechanical Design textbook, any edition to understand the correction required to bearing race differential.

You also need to correct for material and belt vibration, if it exists, to correct for material lump factors and belt surface anomolies that induce force fluctuations. Lubrication must include correction for cavitation at: 1. higher RPM, 2. Colder temperatures and 3. Lubricant viscosity. You also need to correct for bearing loading histogram between empty time of runnning, nominal time of running and peak times of running. All of these are included in BELTSTAT.

Lubricant viscosity and lube behavior is a complex subject. It is treated with technical discussion most significantly in the modern SKF literature, also to a lesser extent in FAG, and further reduced effort by other bearing manufacturers. There are many lubricants available. You need to become knowledgeable about the additives and temperature regimes, and how they impact the lubricant performance. This is not an easy subject and is beyond what can be discussed in the forum.

Read Professore Craig Wheelers discertation and test work on the subject. He is a resident Professor at the University of Newcastle Australia. His work notes the bearing manufacturers methods. He also conducts idler and bearing performance testing with his speed, temperature-environmental, and force loading chamber. ■

Other significant notations:

Idler run-out or Total Indicator Reading (TIR) must be evaluated. Most manufacturers do not adequately control the errors in roundness, eccentricity, concentricity, barrel run-out, or end cap tolerances and end cap attachment that affect the rotational dynamic geometry. These anomalies together with idler location in the idler frame can have profound impact on idler life.

We strive to control the roll TIR to: 1) less than 0.5 mm for more than 95% of the supply for rolls less than or equal to 152 mm, 0.6 mm for rolls <= 178 mm, and 0.65 mm for rolls <= 200 mm. This is also dependent on belt width. Roll RPM must be less than 700 RPM otherwise either balancing or reduction factors must apply. Absolute tolerances must also apply to the above. Manufacturers should be held accountable for the QA procedures that prove the production line accuracy.

Idler roll alignment in their frame should be controlled to <= +/-1.00 mm roll geometry position for high speed conveyors with > 700 RPM. This requires special roll location indexing in jigs that align all roll assemblies. Side-to-side errors, yaw, roll and skew errors in all six axes further reduce bearing life and are ignored by the designers.

Belt misalignment can also further reduce bearing life and is ignored by the designers.

Vertical curve pressures and roll position tolerance is given modest notation which does not quantify the differential failure rates.

Material loading anomalies and material crossectional coalesces lead to further reduction in an equivalent bearing loading histogram.

I am sure more can be added to this list. ■

More notes:

Idler L-10 life is highly misused as a design guide. By example if the conveyor runs for 8000 hours per year and can be designed with one idler at 8,000 L-10 Life hours versus 100 idlers at 64,000 L-10 life hours which is more beneficial to the conveyors performance? Obviously, the one roller has fewer failures per year than the 100 roller configuration. So, both capital, operational and maintenance costs should be given due credit for the superior design of the one roll versus the 100.

Yes, the example is not realistic, but, does highlight the error in strictly designing for a fix L-10 life hours.

Furthermore, the one roller will provide significantly less noise impact.

Today, we see some (many?) designers adding absurd quantities of idler rolls to conveyors where prudent design calls for 20-25% of the quantities. This is because the client is not aware of the large maintenance penalty as well as capital cost penalty, believing the designer to be credit for superior design. This is even more evident on return roll configurations where a 3-roll system can be shown to be vastly superior to a 2-roll configuration. In part, this stems from monkey-see-monkey do and 30 years of this practice. Designers often do not question the status-quo with modern engineering know-how. Very sad.

Another factor that is not well understood: small idler spacing leads to greater failure rates and reduced belt alignment. This is axiomatic. Reduced idler spacing means there is more likelihood of local errors in vertical alignment. Thus, the protruding roller will fail at a much higher rate than the nominal. Since the failure is dependent on (force)^3 power, the obvious impact to local failure should be considered. The smaller roll spacing may not allow the belt to adequately contact all three trough rollers and therefore lead to belt misalignment irregularities. Larger spacing can fix the error. This is most significant on some return idler installations. ■

Hi Endri,

We use rollers in 1600 and 1800 mm wide belts, with 5000 t/h iron ore on them.

We use rollers and get 5 years garantee from the factory. It is not unusual thet our rollers last for 10 years.

When mounting them take good notice that the rollers are not forced in the stands with a hammer. you damage the bearing.!!!!

If you want to know more let me know

Harry ■

Dear Rolland,

The L-10 life hours refers to the point where the bearing pending failure is audible, not at the point of failure.

The bearing will typically last much longer when the race has spaulled to cause sufficient heat to vaporize the lubricant and seize. ■

Dear Rolland,

Many years ago (+20), Stephen-Adamson worked on a project to quantify the potential of a greater life factor called B-Useful Life. They developed, from many tests, the failure curves and guideline design criteria for B-Useful Life Hours.

You might find references to the study work from someone who is willing to share it.

I do not know why the practice was abandon. ■

any caculation format of such issue? ■

Endri,

Yes, we have said this many times. It's the lubricant that determines life and power to all things.

Some bearing manufacturers will aid in selecting the right lubricant for your specific bearing. It is their job to maximize bearing life. It's your job to select the lubricant that maximizes your investment when looking at Total Life Cyle Cost of Ownership.

However, this is not practiced, although it should be. It takes a bit more work, plus you must understand temperature and speed issues with cavitation of the lubricant in the nominated bearing size. You do not want cavitation, but for some, this cannot be avoided. ■