The latest digital vibration transmitters provide critical machine health data for both process monitoring and vibration analysis.

Traditional vibration monitoring focuses on vibration analysis, using the Fast Fourier Transform (FFT) process for specific fault identification. This is a much more complex machine health parameter to understand than others, such as pressure, flow, or temperature, where a single value gives a full indication of the current status. 

When vibration analysis is performed regularly, developing faults are identified early, and predictive maintenance can be scheduled during planned downtime without impacting production.

Vibration analysis provides complex data that must be interpreted by skilled practitioners using tools, such as handheld vibration analyzers, that perform in-depth calculations and pinpoint developing faults.

Process monitoring is a simplified version of vibration monitoring that uses 4-20 mA data to represent overall machine vibration. Process monitoring, combined with a PLC, DCS, or SCADA system, is a cost-effective way of achieving continuous vibration trending. It does not require vibration analysis expertise – simply the interpretation of 4-20 mA alarm thresholds.

Increases in vibration correlate to deteriorating machine health but do not pinpoint specific machine faults. When an alarm or shutdown threshold is exceeded, a physical inspection may indicate the cause. If necessary, vibration analysis can be performed.

This brings us to the vibration gap. Traditional vibration analysis, with its complex information, doesn’t communicate with continuous process monitoring systems. These systems, which use overall vibration levels to monitor machine health, may overlook developing fault clues within the detailed vibration spectrum.

Analog Vibration Transmitters

Analog vibration transmitters are available in two common forms:

• a DIN-rail-mounted module
• a loop-powered sensor

Both output a 4-20 mA signal proportional to the overall vibration and can be trended with a plant control system, however:

• Vibration transmitter modules require a 24 VDC power source that must also power a local dynamic accelerometer. The dynamic signal from the accelerometer can be accessed through a BNC connector for signal analysis.
• Vibration transmitter sensors measure the vibration and perform the conversion internally. While most only output the converted 4-20 mA signal representing the overall vibration, some make the dynamic accelerometer data available, too.

While analog vibration transmitters can provide a 4-20 mA signal for process monitoring and a dynamic signal for vibration analysis, they are still subject to the vibration gap.

When vibration levels trend upward, traditional vibration analysis is performed to diagnose the specific fault.

Analog vibration transmitters have another limitation – the manufacturer presets the measurement parameters. While the full-scale range is selectable, it’s important to know the exact parameters at the time the vibration transmitter is ordered.

For example, a vibration transmitter with a 1.0 inch-per-second (ips) full-scale measurement range may saturate even though an asset is operating within normal parameters.

Conversely, a unit with a 5.0 ips full-scale range may lack the resolution to identify small, meaningful vibration changes. The full-scale range for each sensor must match the unique operating conditions of each machine; a mismatch can only be corrected by purchasing a new vibration transmitter.

Finally, analog vibration transmitters are limited to one frequency band, though more modern models can futher filter the bandwidth. Traditional vibration monitoring resolves this with narrow frequency ranges. Vibration analysis instrumentation can also assist with examining specific, narrow frequency bands and diagnosing developing faults.

Narrow-banded vibration transmitters for specific fault detection exist; the vast majority, however, are general-purpose with a wide frequency band. The 4-20 mA signal represents the “overall vibration” across that range. This can mask developing faults, especially changes at low frequencies when there is little increase in mid- and high-frequency vibrations.

Digital Vibration Transmitters

Digital vibration transmitters bridge the gap by resolving the limitations of analog transmitters and traditional vibration monitoring. Unlike analog 4-20 mA vibration transmitters that measure overall vibration, digital vibration transmitters are based on accelerometer waveforms. This allows for independently configurable bands for multi-targeted fault detection based on full vibration spectrum data and simplified, digital data for process monitoring.

Traditional vibration monitoring relies on analog data that is too complex to work with basic process monitoring systems. With digital versions, this data is processed and filtered within the transmitter and outputted as a digital signal compatible with the latest generation of process monitoring systems, like those developed to support Industry 4.0 and the Industrial Internet of Things (IIoT). Being able to feed vibration data into processing systems over an existing communication protocol simplifies integration and minimizes the need for additional equipment to interpret data.

Digital communications also allow the user to configure the vibration transmitter. An analog transmitter must be replaced and manually reinstalled to change the measurement parameters, however, with configurable digital vibration transmitters, measurement parameters can be accessed and modified digitally. HART and Ethernet protocols make this as simple as accessing an online configuration tool to change the full-scale measurement range, output type, or frequency bands of interest.

Frequency bands – plural – are where the latest generation of digital vibration transmitters approach the protection capabilities of traditional vibration monitoring. The complex signals from a traditional accelerometer are transferred to instrumentation and software that can examine frequency bands known to expose specific faults. One dynamic vibration signal can be filtered many times over for different faults.

Outputting this complex data directly would be meaningless to even the most advanced process monitoring systems. Instead, digital vibration transmitters perform the processing internally and output a simple digital value that can be used to trend vibration relative to multiple, targeted frequency bands. 

Gearbox Frequency Banding Examples

Traditional vibration monitoring is a proven best practice for the early detection of common gearbox failure modes such as:

• Tooth wear
• Tooth load
• Gear eccentricity
• Broken or cracked teeth
• Gear misalignment

When caught early, the gearbox can be serviced during planned outages to reduce unplanned downtime and repair costs when compared to a gearbox that was run to failure.

Analog vibration transmitters may indicate an increase in the overall vibration, but the machine will be closer to failure when the developing fault is detected, especially those occurring at lower amplitudes.

Digital vibration transmitters approach the prognostication capabilities of traditional vibration monitoring with simplified digital data that can be incorporated into modern process monitoring systems.

Example 1 – Moderate-Speed Gearbox

For early fault detection in moderate-speed gearboxes, it is important to monitor the Gear Mesh Frequency (GMF), including side bands and harmonics. 

GMF is calculated by multiplying the shaft speed by the number of teeth. A gearbox rotating at 1500 RPM (25 Hz) with 24 teeth has a GMF of 600 Hz. Including harmonics, the 2X and 3X of the GMF, the frequencies of interest are 600, 1200, and 1800 Hz.

GMF = (shaft speed) x (# of teeth)
GMF = (1500 RPM) x(24 teeth)        = 36,000 CPM            = 600 Hz (1X)
                                                                                                = 1,200 Hz (2X)
                                                                                                = 1,800 Hz (3X)

Let’s say we have a vibration transmitter sensor with a ±3 dB frequency response of 3.0 Hz - 1.95 kHz and three configurable frequency bands. These can be banded to monitor the 1X, 2X, and 3X harmonics of the GMF and detect meaningful vibration increases that may be obscured in the “overall vibration” of the full frequency range.

Example 2 – Highspeed Gearbox

In some cases, a 4-20 mA sensor will not have a frequency response high enough to monitor the primary running speed and important harmonics of your machinery. Take, for example, a 51-tooth gearbox with a shaft operating at 3600 RPM (60 Hz). 

As calculated below, the 1X, 2X and 3X of the GMF are at 3,060 Hz, 6,120 Hz, and 9,180 Hz, all out of the range of most 4-20 mA sensors.

GMF = (shaft speed) x (# of teeth)
GMF = (3600 RPM) x(51 teeth)        = 183,600 RPM          = 3,060 Hz (1X)
                                                                                                = 6,120 Hz (2X)
                                                                                                = 9,180 Hz (3X)

In this scenario, a more suitable monitoring method would be a general-purpose, or even high-speed, accelerometer coupled with a vibration transmitter.

Conclusion

Digital vibration sensors simplify data to give meaningful process outputs without sacrificing full spectral data for machinery fault analysis. Complex data systems can be converted into easier-to-use vibration data and integrated into existing plant infrastructures for early detection through 4-20 mA signals, while the vibration analyst is provided powerful field programmable digital tools to focus on vibration frequencies of interest.

Data for process monitoring operations allows the “early alarm” capability important for balance-of-plant coverage and enhances it by using vibration data focused on measurement bands that are defined by the analyst. It’s simpler to realize the reliability gains that digital manufacturing offers and capitalize on the smart manufacturing breakthroughs that are increasingly becoming the industry standard:

• Improved safety
• Optimized performance
• Longer machine life

Data for the vibration analyst is readily accessible, and they will have full access to raw vibration spectrum for further investigation and fault diagnosis.

This digital transformation of moving data to people instead of sending people out to the source of the data (machines) is a vital component of Industry 4.0 – using data to optimize all aspects of operations/production.

The simplification of data processing and increased accessibility will inherently help bridge the vibration monitoring gap present between process monitoring and true vibration analysis.

 

Peter Eitnier is a featured speaker at the 2023 Reliable Plant & Machinery Lubrication Conference & Exhibition. To view his session and others, click here.