What is an unstable oil film? If you could see it, what would it look like? To understand an unstable oil film, one must first know what a stable oil film does. A stable oil film is the layer of oil between the components of a bearing. It accomplishes more than just keeping two surfaces from contacting each other and welding themselves together. The oil pressure wedge helps cushion vibration energy that would otherwise be directly translated into the stationary portion of the machine. An unstable oil film cannot establish a continuous pressurized wedge of oil to separate the stationary and rotating surfaces. Instead, the oil wedge builds and collapses in an erratic manner. This fluid instability creates many mechanical problems. This case study offers a rare glimpse of oil instability in action. But first, it is important to understand how the bearing is constructed.

Figure 1

The bearing in Figure 1 is from a large 1,000 hp center-hung induced draft fan and is the bottom half of a fixed pillow block babbitt bearing assembly. The thrust collar is inserted into the bearing assembly (the shaft has been removed for illustrative purposes). The thrust collar fits into a machined groove in the shaft where it is clamped together with two machine screws located on either side of the collar split line. The fit in the shaft is precise and quite tight. Between the tight fit of the shaft and the two setscrews that keep the collar from rotating, the thrust collar is essentially a rigid fixture. It is equivalent to a collar that has been machined into the shaft from one solid piece of bar stock. The shaft and thrust collar turn in unison. The thrust collar handles the axial portion of the loading on the shaft, and none of the radial portion of the load. Focusing for a moment on the radial portion of the bearing, it is easy to visualize the oil wedge that is created between the shaft and the bearing. The eccentricity of the shaft with respect to the bearing creates the oil wedge, and the shaft is supported vertically. Any changes in the position of the shaft perpendicular to its axis will change the location of the oil pressure wedge. The wedge reacts instantly by supporting the shaft and maintaining stability.

Figure 2

Two elements called thrust plates carry the axial load from the thrust collar (Figure 2). There is no surface eccentricity of the thrust plate with respect to the thrust collar. The surfaces are parallel to each other and cannot create an oil wedge. So then, how is the oil wedge created? Figure 3 shows the thrust side of a thrust plate. There are eight regions that are depressed into the plate. The oil is uniformly distributed to all eight regions. The number, size and profile of the depressions are based upon the calculated thrust loads of the design. Notice how each depression is beveled; this is where the oil pressure wedge develops as the thrust collar rotates. It is a simple and reliable design.

Figure 3

When the thrust assembly (the thrust collar and thrust plates) is working properly, the vibration is minimal. The eight pressure wedges cushion any fluctuations in load. There is an additional thrust plate on the side of the thrust collar that is not normally loaded. It is identical to the other thrust plate in design and construction. Both thrust plates quickly absorb abnormal and unstable operating conditions that would cause the shaft to shift in the axial direction. When the thrust assembly is functioning properly, the vibration is low.

The axial vibration waveform for this bearing assembly is shown in Figure 4. A normal (top) and an unstable (bottom) operating condition are represented. Both of the waveforms are directly related to the ability of the oil film wedge to form between the thrust collar and plate. The vertical axis denotes the amount of acceleration of the bearing housing assembly. For a given mass, this translates directly into force. The force can change direction. Anytime the waveform is above the zero, or neutral axis, the force is coming toward the vibration transducer. Anytime the waveform is below the neutral axis, the force is going away from the vibration transducer. Notice how the top waveform is almost symmetrical about the neutral axis. This is a fundamental law of physics in action. For every action, there is an equal and opposite reaction.

The blades of the fan are the driving force and there are 12 blades on this fan. The distance of one revolution of the shaft is shown between 35 and 68 milliseconds on the graph (noted by the solid vertical lines). Each time a blade passes the cutwater of the fan, a pressure wave of compressed air is pushed out of the fan. This force causes the shaft to move in an axial direction, and the thrust plate absorbs and counters this force. The thrust plate does this because of the eight oil pressure wedges that develop between the thrust plate and thrust collar. As the blade passes the cutwater, the pressure is relieved and the pressure in the oil wedges push the shaft in the other direction. This action-reaction sequence repeats itself for each blade. That is why there are 12 sinusoidal waves for each revolution of the shaft.

What would happen if an oil wedge could not develop? The bottom waveform in Figure 4 is an example of this. It is not symmetrical like the top waveform. Instead, the waveform is quite erratic and nonrepeatable. This illustrates oil instability - an oil wedge that cannot form properly, and collapses upon itself. At best, the oil pressure wedge is marginal. Notice that the level of vibration has almost doubled.

The increased level translates into more wear and tear on the rest of the equipment and fatigue becomes a limiting factor. The erratic waveform looked like this for several months. Upon inspection of the bearing, the thrust plates were found to be installed backward.

Figure 5

The silhouette left by the thrust side (depressions) of the thrust plate can be seen as a footprint on the bearing housing (Figure 5). This is the wrong direction for the thrust plates. The backside of the thrust plates was essentially flat, except for two machined grooves that allowed oil to flow to the radial portion of the bearing; the thrust collar and the backward thrust plate. This illustrates how a stable oil wedge could not develop.

For those of you who are familiar with this style bearing and are rolling your eyes in disbelief, the next paragraph is for you.

Going back to Figure 2, antirotation tabs keep the thrust plates from turning. When the tabs are installed, there is only one way to fit the thrust plates into the housing. Both sets of thrust plates are installed correctly in the figure for illustrative purposes. It is impossible to put them in wrong, or so one would think. Look closely at Figure 2 again. The antirotation tab in the upper right is installed correctly. The antirotation tab in the lower left is missing. Why? When the tab is installed and the thrust plates are installed backward, the plates don’t fit properly. A good mechanic would normally catch this and simply reverse the thrust plates, causing the plates to then fit correctly. However, this mechanic figured that the tab was the problem keeping the thrust plates from fitting, and simply removed the tab so that the plates would fit. The plates fit, but they were backward.

Author Richard Adler has more than 25 years experience within the fields of maintenance and maintenance engineering. He has worked for several companies in the petrochemical, oil refining, specialty chemical and pharmaceutical industries. Photos and articles about actual failure analysis events can be viewed on his Web site: www.RESnapshot.com.

Photos © 2001 R.H. Adler