Reliability Engineering Snapshot TM

Illustrated Case Studies in the Industrial World of Failure Analysis, Non Destructive Evaluation, and Predictive Maintenance


Vibration - Case Study No. 83: Incorrectly Installed Pillow Block Thrust Plate


The vibration analyzer never lies. You just have to know how to understand its language. When a waveform changes its shape, its trying to say something. For an entire year the waveform was different on a machine, yet nothing was done; mainly because we didn't know what it was trying to say.

Over that year the overall vibration level had been inconspicuously trending upwards. It was only because of other problems that the reason for the different pattern was discovered. The benefit now though is that we will be able to recognize this problem in the future. Vibration analysis can't be learned from books alone. A lot of it comes from field experience.

Vibration Trend Display


Fig. 1 Overall Vibration Level - Trend Display

This is the overall vibration trend level for the inboard thrust bearing on a center hung induced draft fan. The vibration had been trending upwards over the past year. Worse yet, over the past couple of months the oil supply to the bearings had been interrupted several times. After the last oil interruption, the vibration level was measured (shown as day 30 on the trend display). That signature and waveform are shown in Fig. 2.


Vibration signature - Loose Bearing Insert


Fig. 2 November 25, 2000 12:48 P.M. (Noon)

The trend level was getting too high for comfort. There were two many harmonics close to running speed, a sure sign of looseness. The waveform had been continually deteriorating. That is, its shape had been continually changing since when the fan wheel was new. Therefore, the decision was made to replace the inboard fan bearing.

For comparative purposes, a baseline vibration spectrum and waveform that were taken when the fan wheel and bearings were new, is shown in Fig. 3.

New Fan Wheel  & Bearings - Vibrntn Sig. and Waveform

Fig. 3 August 26, 1999

A brand new fan wheel and bearings. The vibration signature is clean, with only one spike at blade pass frequency. The waveform shows 12 sinusoidal waves within one revolution of the fan wheel (measured from the first vertical black line to the second black line). This corresponds to the 12 blades on the fan wheel.

New Bearing - Same Problem

Fig. 4 November 25, 2000 10:25 P.M.

The new bearing was installed and vibration readings taken. The vibration got worse with the new bearing (see red dot on vibration trend chart in Fig. 1). A laser alignment was done on the fan. The mechanics did not use the laser correctly and therefore the alignment was essentially ineffective.

Vibration After 2nd Alignmnt and Tightening Plunger

Fig. 5 November 26, 2000 12:17 P.M.

The vibration level dropped a small amount with the second alignment (reading to the right of the red dot in Fig. 1). The next step was to second guess the mechanics who repaired the bearing in the first place. There is a "plunger" on this style bearing that holds the bearing insert snug within the bearing housing. If it is not torqued correctly then the bearing insert will be loose within the housing and will vibrate a lot. Therefore, the plunger was retorqued. The vibration dropped significantly with this small adjustment (bottom of curve on Nov. 26 in Fig. 1).

However, the waveform was still markedly different than when the fan wheel was new back in August 1999. Even though the vibration level was better than it had been in a long time, something still seemed wrong. The feeling that something was still wrong was based simply upon experience with waveform patterns.

Vibration with Thrust Collar & Thrust Plates Loose

Fig. 6 December 5, 2000 11:30 A. M.

The vibration trend level was going back up. The old failed bearing was inspected closely and it was discovered that it had not been installed correctly (see LUBRICATION article). It was feared that the new bearing might not have been installed correctly either.

The vibration was measured prior to shutting the fan down (Fig. 6). The fan was then shut down for an inspection and balancing. It was discovered that both thrust plates were installed correctly with the babbitt facing the thrust collar. However, the anti-rotation tabs were not properly installed. As it turns out, the thrust plates rotated and cut off the oil to the bearing. There may have been intermittent oil supply as the thrust plates rotated and came in and out of correct alignment. In addition, the thrust collar was loose on the shaft, freely rotating with the flip of a finger. The groove that the collar fit in was oversized, having been banged out by the collar itself.

Vibration After Balancing and Fixing Thrust Collar & Plates.  Tightened Plunger

Fig. 7 December 7, 2000 8:53 A.M.

The fan wheel was balanced and the thrust plates were installed correctly. The thrust collar was locked down, although the fix for the thrust collar was temporary at best. With these last measures the vibration level dropped significantly. The axial displacement before reinstalling the thrust plates correctly was 23 mils. Afterwards, the displacement dropped to 2.8 mils. Although the waveform is not as nice as when the fan wheel was new, the waveform looks better than it had in an entire year. The 12 sinusoidal waves (denoting blade pass) are not as clean or pronounced as before, but they're still discernable, which is more than can be said for any of the other figures (i.e. 3->6). The poor fit between the thrust collar and the groove in the shaft (that the collar sits in) is probably the reason that the waveform has not returned to its original shape.

Stay tuned for the permanent repair of the thrust collar groove.

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.