Spherical roller bearings can accept a certain amount of misalignment. However, they should not be used for high thrust loads. Couple those two attributes with the chance for human error, and you can count on a failure. In the case of kiln trunnions, too much misalignment, combined with a high thrust load, killed this trunnion bearing within 48 hours.
Spherical roller bearings have been used in rotary kiln trunnions because they're reliable. However, care must be exercised in knowing the limits of exactly how much misalignment can be safely handled before a problem begins. The bearing manufacturers do not define a misalignment limit. If some of them do, then it would be prudent to take heed of that limit. This is because the design is especially susceptible to human error. Yes, everything is susceptible to human error; but when someone mentions the words "self-aligning" it can mean different things to different people. To some people it means that care in seeking proper alignment is not necessary. This couldn't be further from the truth when installing replacement trunnions.
The picture to the left shows the outer race of a spherical roller bearing that was misaligned passed its limit. It failed within 48 hours of installation. The dark straw brown color, along what use to be the roller ball path, indicates that this bearing experienced a surface temperature between 4000 to 4500 F. What prompted people to shut the equipment down was the loud cracking sounds coming from the bearing housing. No wonder, those cracking sounds were pieces of outer race breaking off. Two of those pieces were recovered and put back in their original locations, as shown in the picture to the left.
Misaligning the inner race with respect to the outer race causes the rollers to run off the edge of the outer race. This puts tremendous shearing loads along the edges of the raceway. The picture shows numerous cracks still in progress. The close up shots of the two pieces in the picture to the left are shown below. The top piece (lower left) and the bottom piece (lower right) reveal how the pieces were pushed laterally outwards (in the axial direction) as the cracks progressed radially outwards. This is evidenced by the worn abraded edges (circled in red) showing where the rollers contacted the pieces while they were being torn off the outer race. To see what the outer raceway looked like about 900 away, click on the picture to the left. In that picture you can notice how the rollers were well within the raceway.
The color banding that parallels the crack surfaces indicates high localized temperatures.
The overhung rollers created an edge loading that initiated multiple shear cracks along the entire edge of the outer race. Once the edge started to deteriorate, the loading became uneven. A typical series of cracks can be seen in the picture to the left. In the picture, the red dot outline indicates the location of the piece shown in the picture to the upper left. The chevron markings indicate the direction of multiple parallel cracks running simultaneously in the same direction. The beach marks indicate crack arrest locations and are usually oriented perpendicular to the direction of the stress tensor.
The moral of the story is that self-aligning spherical roller bearings are tolerant to misalignment to a certain degree, but tell that to the mechanic and all that he may hear are the words "self-aligning" and leave it at that. These bearings do not have a mind of their own and cannot take care of themselves.
Inattention to alignment was the ultimate root cause of this failure. The mechanics put the new trunnion in the exact location as the old trunnion. However, the old trunnion had been moved over the years to account for wear. The mechanics did not realize that the rolling element of the trunnion that was removed had a taper on it. Thus, the new trunnion went in on an angle that created a tremendous thrust load, besides the misalignment. Bearing manufacturers indicate that spherical roller bearings are not that good in high thrust load conditions. This experiment in failure proved them correct.
All Pictures and Text Copyright © 2001 - 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.