CompressorTechTwo - 2010 (Sample)

TRACKING THE SILENT KILLER, TORSIONAL VIBRATION, WITH HIGH-TECH TOOLS
A Key Question: What Level of Torsional Vibration is Acceptable
By Joe Urban, Raymond Christenson and Pavel Mayzus

A major oil and gas producer had compression plants throughout an older gas field. The field had six similar compressors driven by 1775 hp (1323 kW) motors. The units had been in service, problem free, for 15 years or more. Over time, the pressures in the gas field had dropped so low that one of the compressors was actually pulling a vacuum to get the gas out of the ground.

Suddenly, a crankshaft broke on one of the compressors. Because downtime is so costly, the gas producer hired AP Dynamics to get to the bottom of the failure and to determine whether there was a hidden problem lurking that might produce additional failures among the other units. AP Dynamics’ investigation, involving system analysis and including rotor telemetry physical measurements, determined that torsional vibration was the culprit.

Torsional vibration usually exhibits no overt symptoms. You can’t feel it resting your hand on the machine, and you won’t see the compressor trying to rip itself out of its mountings, but it can be there, nevertheless, quietly setting the stage for a catastrophic failure.

A simple definition of torsional vibration is dynamic variations in the torque that is applied to a shaft. In an ideal universe, compressors (once they are up to speed) would always operate at the same constant level of torque; ditto for the engines or motors that drive them. But this is not an ideal world. When, for example, you have a six-cylinder engine driving a four-cylinder compressor, the engine is introducing torque peaks into the shaft and so is the compressor. Working together, the compressor and engine can produce unexpected torsional vibration. Virtually any compressor package is going to have some torsional vibration, particularly when you have variations in load. The key

Joe Urban, BSEE, joined Accumetrics Associates in 1999 as a project engineer. He holds a BSEE degree from Binghamton University, Thomas J. Watson School of Engineering and Applied Science. He functions as product manager for Accumetrics single-channel digital rotor telemetry systems. He also is lead engineer for custom system design for single-channel telemetry products. Raymond Christenson, P.Eng., is field manager for AP Dynamics. He joined the company in 2004 after graduating from the University of Alberta. He developed the AP Dynamics multichannel high-speed data acquisition and logging system, improving analytical modeling through the addition of practical field verification. He also performs FEA, acoustical, mechanical and pipe stress analysis and provides leadership to the field work. Pavel Mayzus, P.Eng., MSc, is a rotordynamics and torsional vibration specialist with AP Dynamics, Calgary, Alberta, Canada. He has seven years of experience in stress, vibration and acoustical pulsation design and analysis. As part of his MSc studies, he assisted with modeling and experimental verification of a pulse tube miniature cryogenic cooler. Mayzus contributed to the development of the torsional analysis methodology at AP Dynamics and provides leadership and expertise in this area to the team.

question is: what level of torsional vibration is acceptable for the operation of this package?

That’s where a company like AP Dynamics comes in. Frequently, gas producers who are about to install a compressor set will hire a consulting firm to conduct a torsional vibration study. Detailed information on compressor and engine performance is supplied by the manufacturers and then fed into a software system that allows the opera-

(Photo courtesy of Vishay Micro-Measurements)

Figure 1. Typical strain gauge variations (Vishay, 350 Ohm strain gauges).

tion of the compressor package to be simulated, along with a number of check points such as stress on shaft, torque on coupling, etc., to ensure that the equipment will not break under various operating conditions and loads.

AP Dynamics used its simulation system (see sidebar) to determine the cause of the broken crankshaft and found that torsional vibration was, indeed, the likely cause of the failure. With this in mind, AP Dynamics