|Sometimes when trying to minimize equipment downtime, an engineer trades one design problem for yet another design problem. It comes down to choosing the lesser of the two evils. The better engineer will determine which design minimizes equipment downtime AND affords the greatest amount of reliability.
Anyone who has had to replace the fan bearings on an overhung fan has experienced the hassle of removing the coupling hub and then trying to pull the failed bearings off of the fan shaft. Worse yet, if the coupling does not have a spacer it means that the motor has to be moved in order to make clearance so that the coupling hub can be removed off of the shaft. Imagine a coupling that is stuck on the shaft and needs to be heated up with a "rosebud" torch, and a 1200 HP motor that needs to be removed by a 70 ton rig each time a fan bearing needs to be changed. These thoughts would make any shell shocked maintenance foreman squirm.
Hence, enter the split bearing (picture left). This design is a life saver when it comes to minimizing repair turnaround time. Instead of having to pull the coupling hub off of the shaft and maybe moving the motor, one only has to pop the bearing cover off of the bad bearing and slip the failed components out of the bearing, much like popping the sleeves out of a pillow block sleeve bearing. Everything is split into two, the inner and outer races, the ball cage, and the liner inserts. The inner raceway is held together onto the shaft by means of collars that are bolted together. The collars and the inner raceways are shorter in circumference than the shaft. There is a very small hairline distance between the two races that allow the collars to snug the races onto the shaft when the collars are drawn together by bolts. The roller cages are held together with a simple clip. Click on the circled split ball cage for a close up look at the design.
|This design is a life saver. However, it is not without it's own set of problems. The most aggravating problem with this design is that it will eventually come loose. The bolts eventually creep loose. The picture in the upper right shows the beginning stages of fretting damage on the inner race; fretting damage is caused by looseness. The other side of that same race is shown in the picture to the upper left. Fretting has not started there yet, those bolts were still tight.
Now before anybody suggests that it wasn't properly installed, you're wrong. If anyone suggests that maybe we should check these periodically, you're correct. How long did it take for the bolts to creep loose? This bearing survived 1.95 billion revolutions. Click on the locking collars in the picture to the left to see fretting damage on the inside of the collars.
In this plant, that many revolutions came out to 2 1/2 years. If that length of service appears to be small, then consider the industry. I always chuckle at the "service factor" column in the vendor catalogs when sizing various types of components. Those service factors are rather tame for the trona mining industry. For instance, take the Samsonite luggage commercial with the gorilla in the cage. For those of you who don't remember that commercial there was a gorilla in a cage that was given a piece of Samsonite luggage. The gorilla proceeded to beat the hell out of that piece of luggage. The luggage survived, implying that it was sturdy. However, if the trona mining industry had made that same commercial, we would have given the bag to King Kong to establish the "service factor."
The other problem with these bearings is in condition monitoring. They are not your normal bearing, and their vibration signatures follow suit. That problem is highlighted in the article "Vibration - Case Study No. 72: Diagnosing Split Roller Bearing Problems." In addition, this bearing tends to run hot, and the reason is showcased in the article "Lubrication - Case Study No. 69: Eliminating Pressure Lubricated Systems and Living with High Heat."
All Pictures and Text Copyright © 2000 - 2016 Contact Mr. Adler Adler Engineering LLC of Wyoming USA
Great care has been taken in the compilation of this article. However, no warranty, expressed or implied, including without limitation, warranties of merchantability or fitness for a particular purpose, are given in connection with this article or any article archived on this website. Although this information is believed to be accurate by the author, the author cannot guarantee favorable results will be obtained from the use of this article alone. This article is intended for use by persons at their sole discretion and risk. Since the conditions of product or material use are outside of the author's control, the author assumes no liability or obligation in connection with any use of this information. The author is not liable for special, indirect or consequential damages resulting from the use of this material.
No part of this article or any article archived in this website, or any part thereof, may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise, without the prior written permission of the copyright holder R. H. Adler. Nothing contained in this article or any article archived in this website shall be construed as a grant of any right of manufacture, sale, use, or reproduction, in connection with any method, process, apparatus, product, composition, or system, whether or not covered by letters of patent, copyright, or trademark, and nothing contained in this article or any article archived in this website, shall be construed as a defense against any alleged infringement of letters of patent, copyright, or trademark, or as a defense against any liability for such infringement.