Over the past five years, the use of directed lube tilting pad journal bearings has grown due to ever increasing speeds of turbomachinery. Directed lube bearings provide cool lubricating oil directly to the leading edge of the tilting pads, thus reducing power losses and oil film temperatures.1 (DeCamillo, Brockwell 2001) This technology has been used with thrust bearings for two decades; however, the use with journals is somewhat new.
New technology brings new problems. In late 1999, a customer approached Kingsbury with evidence of a new type of subsynchronous vibration (SSV) on a directed-lube journal bearing. The vibration was considered "new" because it occurred only in this directed-lube class of bearings - never in a flooded bearing. This new vibration was originally detected on a competitor's bearing but then also appeared on Kingsbury's LEG® journal bearing. Thus far, it has occurred on high-speed, lightly loaded shafts typically found on centrifugal compressors and turbines.
Subsynchronous vibrations are generally not allowed according to acceptance criteria placed on the performance test of machines prior to installation. This type of vibration is avoided because it typically has pointed to unstable modes or serious problems in the machine such as loose parts, rubs, cracked shafts, or poorly designed bearings. These vibrations tend to become larger in amplitude over time, thus the term unstable. Therefore, SSV has become a signature to avoid at all cost. For these reasons, original equipment manufacturers (OEMs) perform a battery of tests to ensure that they meet the industry-accepted vibration limit before installing the machinery.
The new SSV found on directed lube tilting pad journal bearings does not appear to be unstable and threatening as the more well-known types listed above. Unlike the classic SSV, this new SSV has a very low amplitude - on the order of 0.1 to 0.2 mils pp - and a low, random frequency that is less than 20 Hz. The classic SSV locks in a fixed frequency, whereas this new vibration's frequency has been termed "hash" because of its random and, somewhat, broadband nature. (see Figure 1)
The concern from original equipment manufacturers has been that this vibration, although small, reduces their working range when trying to meet overall API vibration limits. The limits can be less than 1.0 mils. If the SSV is 0.2, they're working with only 0.8 mils or less. Most people agree that this new SSV is more of an annoyance than a threat, but they would like to be rid of it just the same.
Since 1985 the patented Leading Edge Groove (LEG) bearing (see Figure 2) represented a significant technological improvement for Kingsbury in their market. Previously, the typical means of lubricating a bearing was to flood the entire area around the shoes and shaft. This method, while simple, exposes the rotating shaft to a larger volume of oil in and around the shaft, which adds to churning losses and heating of the oil. Instead of a flooded system, Kingsbury's LEG design supplies oil directly to the leading edge of each shoe and allows the bearing cavity to be evacuated. Its design provides customers with better machine efficiency, smaller lubricating systems, lower capital costs and increased load capacity. Needless to say, finding a solution to the subsynchronous vibration became a high priority.
Faced with this challenge, engineers from Kingsbury's Research & Development department set out to eliminate the subsynchronous vibration. They simulated the SSV anomaly by modeling their test equipment to replicate the conditions used by the OEM. Their in-house test rig was equipped with a similar LEG bearing and configuration then run under conditions to match the test performed by the OEM.
Studying these types of vibrations in their bearings was a new task for the team. They usually researched steady state factors such as temperatures, oil flows and pressure. Special equipment was installed to properly read these low amplitude, low frequency vibrations. Multiple radial vibration probes were mounted on Kingsbury's high-speed turbine test shaft. In addition, probes were placed on upper and lower shoes of the bearing to collect data on shoe movement and vibration. An in-house data acquisition program was written in LabView to handle the large volume of data points.
The SSV phenomenon proved to be especially complex. Most tests at Kingsbury's testing facility take about four weeks to complete. In this case, the R&D team developed and tested theories about this SSV for three months. They knew from early on that the vibration could be dampened to an acceptable level by simply flooding the bearing with oil (see Figure 3), but this defeated the entire purpose of the LEG design. Focusing on flow rates, they operated the bearings in many different configurations, including studying the effects of different seal ring designs, leading edge groove designs, pivot offsets, bearing clearances and preloads. They increased the supply of oil to the upper pads. Then they increased the flow to the lower pads. Nothing worked.
Kingsbury's engineers became convinced that the only element significantly affecting the vibration was the amount of oil entering the shoe. The team calculated what they termed the "critical oil flow" - the difference between the amount of oil that entered the leading edge versus the amount that exited the trailing edge. This flow was typically twice what had been supplied to the LEG. In order to sufficiently reduce the SSV, the "critical oil level" needed to be supplied. Kingsbury, though, wanted to keep the performance advantages of the LEG design by supplying less than the critical oil level.
The question among Kingsbury's R&D team became, "How can we capture oil from the side leakage and re-introduce it back into the bearing?" The engineers considered baffling schemes, shields and a number of other ideas. Eventually, they settled on putting grooves directly into the babbitted surface of the pad (see Figures 4 and 5). Located on both sides of the pad, the grooves captured and redirected side leakage back into appropriate areas of the shoe. By reintroducing oil to the pad, the grooves recycled side leakage and lowered the amount new oil needed to maintain the "critical oil level".
After 3 months and almost 35 tests, the Kingsbury team of engineers found a solution without sacrificing the distinctive design and customer benefits of the LEG bearing. They were and, currently, still are the only bearing manufacturer to eliminate the subsynchronous vibration from the directed-lube class bearing. Kingsbury was awarded Patent No. 6,361,215 for this invention.
The goal of Kingsbury's R&D program is to develop new products, improve existing products, resolve customer problems, mold data into a useful form, and investigate patentable ideas. Over the past few years, Kingsbury's R&D department has investigated topics other than the normal bearing related issues such as manufacturing costs, tinning compounds, adverse operating conditions, warranty claims, seals and baffles, shaft current damage, turning gear failure, oil contamination, investment castings, synthetic lubricants and additives.
As Scan DeCamillo, Manager of Research and Development said, "It is our experience that R&D provides an important advantage over our competition and contributes to our growth and profitability. Our goal is to best utilize our R&D resources to maintain leadership and technological advantage in the fluid-film industry".
1 DeCamillo, S., and Brockwell, K., "A Study of Parameters that Affect Pivoted Shoe Journal Bearing Performance in High-Speed Turbomachinery," Proceedings of the 30th Turbomachinery Symposium, Texas A & M University, pp. 9-22 (2001).
About the author...
Joe Wilkes is the Director of Engineering at Kingsbury, Inc. In addition to overseeing the Engineering department, he is responsible for all Research & Development activity at Kingsbury.