Spectral analysis is an essential tool for evaluating the dynamic behavior of rotating components under extreme conditions. In spin testing, where rotors are subjected to high speeds and extreme temperatures, subtle changes in vibration signatures can reveal emerging mechanical faults before they lead to failure.
At Barbour Stockwell, Inc., we incorporate Fast Fourier Transform (FFT)-based spectral analysis into our spin testing services to extract meaningful frequency-domain insights from complex vibration data. Our spin testing systems, which achieve rotational speeds up to 250,000 RPM and temperatures ranging from cryogenic to 1100 °C, are equipped with spectral monitoring tools that allow us to detect faults early on.
What FFT Reveals About Vibration
Fast Fourier Transform (FFT) is a mathematical algorithm used to convert vibration signals from the time domain to the frequency domain. Rather than viewing how a signal changes over time, FFT shows what frequencies are present and with what intensity, which is crucial information when assessing rotor dynamics.
Named after French mathematician Joseph Fourier, this technique allows engineers to identify specific vibration patterns that correlate with faults like imbalance, misalignment, gear mesh issues, and bearing defects.
Spectral data can be gathered from various sensors integrated into our spin test systems, such as:
Accelerometers
Velocimeters
Non-contact proximity sensors
Displacement Sensors
Strain gauges
Microphones
Barbour Stockwell’s Spin IV spectral analysis module expands our testing capabilities by offering real-time monitoring, display, and recording of spectral data across four analog input channels (-10V to +10V). This tool allows operators to zoom in on specific frequencies to isolate fault frequencies, and capture snapshots for later review.
Spectral analysis allows us to set fault thresholds as a function of frequency. In spin testing, higher trip levels can be set for low-energy sub-synchronous vibration, thereby eliminating unnecessary shutdowns. Beyond that, the employment of Spectral analysis is also very useful for monitoring the health of rotating machinery, whether components of BSi’s drive systems or other high-speed machinery.
When to Employ Spectral Analysis
While traditional vibration tools measure overall amplitude, they often lack the detail required to fully diagnose complex mechanical issues. Spectral analysis tools provide both frequency and amplitude, offering insights into the specific components and behaviors causing the vibration.
Spectral tools are ideal for:
Variable Speed Machinery. Traditional tools may miss changes that occur with speed or RPM variation. Spectral tools track frequency shifts and correlate them with speed.
Diagnosing Harmonics, Sidebands, and Resonance. Patterns in the frequency spectrum often indicate issues with bearings or gears that you can’t see in time-domain data.
Early Fault Detection. Small issues often appear first in the frequency domain long before failure occurs.
Rotating Equipment with Multiple Components. When there are multiple vibration sources (i.e., shafts, motors, gears), spectral tools can isolate and identify each source based on its frequency.
Spectral tools also offer the added benefit of being able to produce waterfall plots, which are graphs that stack frequency spectra over time or speed. This provides the insight required to determine whether the resonance is that of the rotor or the stator of the machine.
Spectral Analysis Solutions from Barbour Stockwell
Spectral analysis plays an important role in rotor testing. It offers engineers a stronger understanding of how rotating components perform under demanding conditions and allows for earlier identification of faults. Our expertise in spin testing, combined with advanced spectral analysis tools, allows us to gain a better understanding of the source of high vibration levels and facilitate measures to mitigate the problem.
For more information about our spectral analysis capabilities, or to get started on your next project, request a quote today.
Metal components of a machine can fail when subjected to loads generating stresses well below the ultimate strength of the material; even below the yield strength. The mechanism for this phenomenon is the application of the load a very large number of times. Hence, such a failure is called a fatigue failure.
The design of a reliable machine demands a firm grasp of the number of times the load can be applied before failure occurs so that this number of cycles will not be exceeded in service. The consequences for such failures, which generally occur without warning, can be catastrophic. Manufacturers of machinery that demands the utmost in reliability, such as aircraft engines, spend considerable time and money into the consideration of fatigue in the design and subsequent testing of critical components.
The importance of fatigue in the design of machinery is reflected in the effort that has gone into the development of guidelines for the fatigue life that can be expected for various metals as a function of the magnitude of the fluctuating stresses applied. This data is generally presented as a S-N curve which is merely a plot of cycles to failure against load. An example of a S-N curve is presented here.
Low Cycle Fatigue failure is caused by loads that result in stresses in excess of the yield strength of the material with failure generally occurring in the range of 104 to 105 cycles or fewer. If the loads do not result in yield, i.e., elastic deformation only, then the cycle counts to failure will be considerably higher; in the regime of High Cycle Fatigue. As can be seen in the S-N Curve, some materials, steel in particular, exhibit a threshold, known as the Endurance Limit, below which failure will never occur.
Now it is important to note that the actual life of a machined part may be significantly less than that presented in a S-N chart for the material due to parameters which, in addition to the strength of the material, influence the fatigue life. Thus, modifying factors must be applied for surface finish, stress concentrations, size and temperature among other characteristics of the application which affect the life. The best way to mitigate the uncertainty of the life that will be realized in service of a critical part is actually testing it under conditions which duplicate those which will be seen in service.
For rotating parts, the best way to perform such testing is in a spin pit. Spin Pit LCF testing allows the manufacturer of a critical component to control the important test parameters to an extent generally not possible in the finished machine. Moreover, the testing can be more economically performed in a spin pit; particularly if the destruction of the product can be avoided. Components for which LCF testing is particularly important include:
• Jet Engine Rotors
• Electric Vehicle Motor Rotors
• Energy Storage Flywheels
Barbour Stockwell provides manufacturers of high-speed rotating machinery with the facilities to perform LCF testing with unmatched precision in speed control; as precise as +/- 1 rpm. Temperatures of the test articles can be precisely controlled as well; isothermally or with gradients. To simulate the mission of an engine, cyclic tests can be programmed with a number of target speeds and associated ramp rates and dwell times. Moreover, a multitude of cyclic tests so defined can be “chained” together to be repeated as many times and in whatever sequence desired.
The manifestation of fatigue is presented in three phases; crack initiation, crack propagation and rupture. Barbour Stockwell has developed a crack detect system with proven technology that has shown to be the most effective means for detecting cracks while Spin Testing – boasting a greater than 90% detection rate. During such testing, the ability to detect the initiation of a crack is immensely valuable. You do not want the first manifestation of a crack to be a burst of the rotor. A burst will generally result in damage to the fracture surfaces adding an extra degree of difficulty to any forensic analyses. On top of that, a rotor burst will necessitate a somewhat costly and time consuming repair to the test facility.
Spin IV, BSi’s powerful test control software package, has been supplemented with a module to monitor, display, and record the spectral content of analog data channels. Spectral content is typically of interest for vibration signals from accelerometers, velocimeters, and non-contact proximity sensors.
The TC-4 has already offered an option to add up to 32 analog channels to monitor test instruments such as temperatures, pressures, flow, and vacuum level. These 32 channels are monitored at 10 Hz for fault monitoring and recorded at 2 Hz; useful for some parameters but not nearly fast enough for determining the spectral content of a vibration signal.
The new module uses the same BSi High-Speed Data Acquisition device employed by the company’s In-situ Balancing Software. Spectral content is calculated using the Fast Fourier Transform (FFT) method. It transforms a vibration signal from the time domain to the frequency domain. It essentially will provide the RMS voltage level at a selected frequency. Frequencies of up to 50 kHz are calculated.
The Spin IV FFT Option utilizes a 4 channel module that allows signals between -10 to +10V. The display is flexible allowing the operator to zoom into a frequency or RMS level. The software also has the ability to collect snapshots of a display and review the snapshots at a later point. Displaying this information can be very helpful in analyzing vibrations in rotational test systems.
A Spectrum display is presented for a single channel. However, the operator can open multiple windows to display all four channels simultaneously. Under a STATUS menu tab, the operator can select to add a Spectrum Display after the Data Acquisition Module has been activated.
As shown below the voltage and frequency parameters can be varied to provide a zooming function to display signals with greater resolution.
The Save Snapshot button will collect a data set of all four channels for the full 50 kHz spectrum irrespective of the real-time display parameters. The FFT Data Review allows for the same zooming functions as the real-time displays. A new record is created each time the Save Snapshot button is pressed.
For more information about BSi’s most powerful test control software package, be sure to visit our contact us page today to get in touch with one of our experts.
A Research and Development program or the Qualification of a new rotor design often incorporates Low Cycle Fatigue testing as part of the process. These LCF tests involve cycling the test rotor between two or more speeds many thousands of times; frequently at temperatures to match the in-service environment.
During such testing, the ability to detect the initiation of a crack is immensely valuable. You do not want the first manifestation of a crack to be a burst of the rotor. A burst will generally result in damage to the fracture surfaces adding an extra degree of difficulty to any forensic analyses. On top of that, a rotor burst will necessitate a somewhat costly and time consuming to repair to the test facility.
In the past, frequent inspections such as fluorescent penetrant or magnetic particle were scheduled during the test program to minimize the possibility of a burst. This, of course, added considerably to the duration of the test. So much better to be able to identify a crack at its inception and monitor its progress in real-time during the test.
Barbour Stockwell has developed a crack detect system with proven technology that has shown to be the most effective means for detecting cracks while Spin Testing – boasting a greater than 90% detection rate.
The technology has demonstrated the ability of measuring cracks as small as 100 µM.
It can also monitor crack growth during Low Cycle Fatigue Testing.
The CD1 System has automated shutdown capabilities, superior digital filtering, and synchronous vibration signal isolation to eliminate harmonic content.
Figure 1 BSI Demonstration Test Article – Crack Initiation
We use cookies to ensure that we give you the best experience on our website. If you continue to use this site we will assume that you are happy with it.
