Kingsbury Engineering Staff

This paper discusses a newly found high speed, light load hydrodynamic tilting-pad thrust bearing phenomenon. The phenomenon was first witnessed on centrifugal compressors operating above 10,000 rpm. Testing results have shown that in high speed, light load applications, thrust bearing babbitt temperatures are much higher than expected at low load conditions, and then drop to expected levels as the thrust load is further increased. This behavior has been observed in flooded and directed lube type bearings as well as center- and offset-pivot designs, occurring at sliding velocities above 300 feet per second (fps) at the mean pad diameter and at thrust unit loads between zero and 100 pounds per square inch (psi).

Additionally, the authors discuss the initial anomalies encountered at an original equipment manufacturer's testing facility; extensive testing at the bearing manufacturer's facility to first reproduce the bearing behavior and then to introduce geometry modification to address the phenomenon. Finally, successful low load testing as well as high load testing is presented with the enhanced thrust bearing design.

For the complete paper with associated illustrations, 

Articles

Evaluation Of a High Speed, Light Load Phenomenon In Tilting-Pad Thrust Bearings

Choosing the right bearing is easy, once you know how. To help, this article outlines the steps to take. It includes guidelines for bearing selection and curves for estimating power loss and oil flow requirements.

The following guidelines for assigning load capacities to tilt-pad fluid film thrust bearings take into consideration theoretical analyses, field experience, and actual test results.

Load capacity is based on film thickness and white metal temperature

Larger thrust bearings for low or moderate speed applications are rated for higher specific loads (load per unit of area) than smaller thrust bearings

Frictional power loss increases with higher rotational speed, greater oil viscosity, and larger thrust bearings

Special Design Features To Increase Rated Load Or Reduce White Metal Temperatures

Journal bearing pad temperatures, oil flow requirements, and power losses can impose limitations on the design and operation of high-speed turbomachinery. Over the past few years, the authors have conducted extensive tests and studies of parameters that affect pivoted shoe journal bearing performance. A special, high-speed test rig is described, which was designed and built for the purpose of measuring bearing steady-state performance characteristics under light to moderately heavy loads, and to very high operating speeds that are being approached in new turbine and compressor designs. Instrumentation includes a detailed array of pad temperature detectors and the direct measurement of frictional torque.

Data are presented that compare the effects of pivot offset, oil flow, load orientation, method of lubrication, and oil discharge configuration on 6 inch diameter pivoted shoe journal bearing performance. The parameters are shown to significantly influence bearing pad temperature and power loss, particularly at high loads and speeds. Pad temperature profiles, isotherms, torque, and oil outlet temperatures are compared and evaluated. Discussions address the prediction and application of these parameters, and how they may be used to improve the capacity and performance of high-speed turbomachinery. The data and discussions are intended to provide useful information to engineers, programmers, and personnel involved with the study or operation of pivoted shoe journal bearings.

Contributed by: Scan DeCamillo and Keith Brockwell

A Study Of Parameters That Affect Pivoted Shoe Journal Bearing Performance In High-Speed Turbomachinery

Babbitt temperatures in pivoted shoe bearings can be significantly reduced by offsetting the pad pivots in the direction of rotation. However, reverse rotation can occur in certain types of machinery under temporary or adverse conditions. The offset is then in the wrong direction and it becomes important to know if the bearing can withstand the reverse rotation operating conditions without damage. This paper compares pad temperature data from tests of a 152.4-mm, 60 percent offset bearing in forward and reverse rotation. The data indicate that the offset pivot bearing can run in reverse without damage for the operating conditions tested. Reverse rotation pad temperatures are hotter. Discussions towards the end of the paper address assessment of the data.

Contributes By: Scan DeCamillo, Keith Brockwell, and Waldemar Dmochowski

Test Results Comparing the Effect of Reverse Rotation on Offset Pivot Journal Bearing Pad Temperatures

Peculiar, low-frequency, radial vibrations have been observed in various turbomachinery using tilt-pad journal bearings. Unlike discrete subsynchronous spikes that often indicate a serious problem, the vibrations are indiscrete and of low frequency and amplitude. The low level shaft indications have raised concern in witness tests of critical machinery, even in cases that comply with American Petroleum Institute (API) limits, owing to uncertainty regarding the cause and nature of the vibrations.

This paper presents shaft and pad vibration data from various tilt-pad journal bearing tests that were undertaken to investigate and better understand these subsynchronous indications. The vibration characteristics are defined and compared under the influence of speed, load, oil, flow, and bearing orientation. Results are presented for conventional and direct lube tilt-pad bearing designs, along with discussions of parameters and methods that were successful in reducing and eliminating these low level vibrations.

The test results indicate that the low-frequency shaft indications are caused by pad vibration. Hypotheses and analyses are presented and discussed in relation to the test observations.

Contributed By: Scan DeCamillo, C. Hunter Cloud, Minhui He, and James M. Byrne

Journal Bearing Vibration and SSV Hash

In mid 2011, technicians at Kingsbury's Repair & Service Division in Hatboro, PA were consulted by the owner-operator of a 40-year-old, 114.3 MW vertical pumped storage turbine-generator located on a major river in the southwestern United States. The combination thrust and guide bearing assembly was installed beneath the generator, with a separate guide bearing assembly installed on the turbine. The 76" diameter twelve-shoe thrust bearing was designed and produced by the original equipment manufacturer and featured a Kingsbury style, pin type, mechanical load equalizing system. By design, this thrust bearing support system automatically self-aligned the bearing system and equalized the load on the individual thrust shoes. During operation, the unit had been experiencing vibration levels generally characterized as "rough", with levels occasionally so high that they led to random shut-downs. Maintenance records indicated that there had been difficulties aligning this machine within the specification guidelines for quite some time. The operator suspected the bearing system was at fault, so during a scheduled outage at the plant, he contacted Kingsbury for field service assistance in order to troubleshoot and provide corrective action. Disassembly and inspection of the thrust bearing showed evidence of binding in the mechanical equalizing system.

The mechanical equalizing system consists of twelve lower leveling plates, which contact the thrust bearing base ring with a central rocking strip, allowing the lower leveling plates to pivot. There are twelve upper leveling plates installed above and between the lower leveling plates. The upper leveling plates are supported individually in a "wing-to-wing" arrangement by the adjacent lower leveling plates through the use of cylindrical pins. The wings of the upper leveling plates and the lower leveling plates are machined to form a slot for the cylindrical pin. A thrust shoe is positioned directly above each upper leveling plate and bears against it with a spherical contact. This system is ultimately intended to provide an articulated load path from the rotor to the bearing bracket.

In this particular design, square section slots were used in the leveling plate wings, with clearance between the pins and the side walls of the slot. Markings on the cylindrical pins and leveling plate slot walls indicated random positioning of the pins, skewed pin arrangements and hard contact between the cylindrical pins and the walls of the slots. As a result, binding conditions were obvious.

Further evidence supporting the theory of leveling plate/pin binding was the manner in which the bearing bracket had been adjusted in the past. The bracket was significantly out of level with "soft feet" conditions, and the grout beneath the sole plates (bearing bracket mounting) had also deteriorated.

For the balance of the article and associated illustrations, 

Contributed by: Michael J Brawley, VP Repair & Service Div, Kingsbury, Inc

Getting To The Bottom Of Vibration Issues

When you hear this question, it is natural to think that there is something wrong with the thrust bearing. This is not always the case. Bearing performance information can give you the symptoms of what is wrong with your machine or in this case the lubricant.

One example is a 7" LEG 6-shoe thrust bearing that over time, had pad temperatures increase from 100°C to as high as 140°C. Another example is a 6" LEG 6-shoe thrust bearing that after a few months in operation had a temperature difference of almost 20°C between two shoes. Note that rising pad temperatures and large differences in temperature readings over time are typical behaviors when varnish forms on thrust shoes.

Both bearings were installed in compressors and used ISO VG 32 viscosity oil. Each shoe shows signs of varnish deposits toward the trailing edge of rotation.

When reviewing the design conditions, the pad temperatures should have been much lower. A common thought is that the thrust load must be higher than expected. Some customers may even ask the bearing supplier to predict the actual thrust load based on the measured pad temperatures. This value can be misleading since the predicted thrust load with the higher pad temperatures does not take into account the effect of varnish deposits on the shoes.

These bearings were not overloaded. The shoe supports are in excellent condition and show no signs of indentation or hard contact.

Contributed by: David M. Casper, Engineering Service Coordinator, Kingsbury, Inc

High Thrust Bearing Temperatures Due to Varnish

How much static misalignment capability does a non-equalizing thrust bearing with spherical supports have?

A misalignment condition is based on the assumptions that the collar (or thrust disc) is perpendicular to the shaft, the collar surface is flat and smooth, and the runout is within an acceptable tolerance.

The geometry of the support is a significant factor in determining how much a non-equalizing thrust bearing can be misaligned. The line contact support cannot tilt in the direction of the misalignment and has minimal deflection under load. A spherical support tilts in both the tangential and radial directions and deflects under load. This deflection greatly improves the misalignment capability of the bearing.

A KingCole size 8225, direct lubricated, non-equalized thrust bearing was reviewed. Note that the "8" refers to the number of pads and the "225" is the radial length of the pad in inches when divided by 100.

The operational performance of this bearing was calculated using ISO VG 32 oil at 50°C inlet temperature at surface speeds of 30, 60, 90 and 110 m/s and thrust unit loads of 1.0, 2.0, and 3.0 MPa. When this bearing is severely misaligned, one pad will take the majority of the thrust load. The criteria used to determine the maximum limits for continuous operation of this pad are a unit loading of 6 MPa and a 125°C pad temperature in the 75/75 location. (If lower temperature limits are applied by specification, the allowable misalignment values will also be lower.)

Dealing With Bearing Misalignment Issues

There is always some degree of thrust bearing misalignment due to housing and bearing machining tolerances, or misalignment between the shaft and housing. This is considered static misalignment since the bearing is misaligned before a load is applied. This static misalignment is different from dynamic misalignment that may occur when thermal expansion and mechanical deflection causes shifts in the housing and shaft position while in operation.

For this example, the equalizing bearing has a spherical thrust shoe support and is a leveling plate or rocker type design. The static misalignment condition is based on the assumptions that the collar (or thrust disc) is perpendicular to the shaft, the collar surface is flat and smooth, and the runout is within an acceptable tolerance.

An equalizing thrust bearing distributes the load more evenly among the pads compared to a non-equalizing bearing. When a load is applied, the thrust pad pushes against the upper leveling plate. This forces the lower leveling plates on either side to tilt slightly and raise the pads on the other side. This reaction occurs on every pad and equalizes the thrust load.

Capacity is greatly improved over non-equalizing bearings although the system is not perfect. There are physical limitations such as loss of leverage due to shifts in pivot contacts. This results in a pad load differential that increases with misalignment and depends on the bearing geometry, contact radii, and the number of shoes.

Contributed by: David M. Casper, Engineering Service Coordinator, Kingsbury, Inc.

Preventing equalizing thrust bearing misalignment

Electrical pitting damage, caused by intermediate arcing between stationary and rotating machine components, erodes the babbitted surface of a bearing, causing it to fail. With the right precautions, however, this type of damage can be prevented.

In an example of a steam turbine application, the consequences of erratic thrust pad temperatures were not obvious until the bearing was removed from the machine. In this case, the normal pad temperature was 95° C for a 10½", 6 x 6 LEG thrust bearing operating at 5700 RPM. The pad temperatures had been rising steadily for several days when suddenly the temperature of one pad shot up to 118° C and the other to 103° C. A few days later, the pad temperatures dropped back to 111° C and 92° C.

When the bearing was removed from the machine, all the thrust pads showed some signs of distress.

The original babbitt is only present at the leading edge of the shoe. Evidence of a wipe can be seen toward the outer edge of the shoe. The areas at the trailing edge of the shoe and toward the inner diameter at the leading edge have clusters of small pitted marks.