Plant & Works Engineering
Vibration - an age-old problem!
Published:  10 September, 2014

Chris Hansford, managing director of Hansford Sensors, discusses the need for accurate data capture as part of a predictive maintenance plan in modern manufacturing and explains the different measurements that can be taken by vibration sensors to carry out condition monitoring in different applications.

Vibration is not a new problem. In 1907, Frederick W. Taylor described machining vibrations as “the most obscure and delicate of all the problems facing the machinist”. While this is still true, technologies have improved, allowing plant operators, managers and engineers to benefit from a host of condition monitoring systems that both provide early warning of potential faults and deliver precise real time data on operating status, allowing maintenance to be proactive and planned.

Vibration can be caused by a multitude of factors; for example, mechanical imbalance, loose components, rubbing parts and bearing wear. Although small changes in vibration levels are unlikely to spell disaster, a prolonged and un-investigated instance can result in total failure or cause problems, which can in turn affect overall productivity and output.

To prevent such outcomes, vibration monitoring has become a powerful tool used by many engineers and plant managers to help them run their manufacturing operations efficiently and sustainably. Since every single motor or gearbox is likely to experience some kind of failure or fault within its lifetime, vibration data is an invaluable resource and can be used to transform a plant plagued by unplanned downtime into an operation that runs effectively and profitably.

Conducting vibration analysis

Essentially, when conducting vibration analysis, the first decision you make is choosing a point and a direction to measure. For example, on a motor the most common positions to mount accelerometers are the Drive End (DE) and non-drive end (NDE), where sensors are mounted radially to monitor the motor bearing condition.

There are three ways to measure vibration: displacement, velocity and acceleration. Why three? And which one should you use? The answer depends on the application but before looking at where these three methods are best applied let’s look at these basic concepts of vibration analysis.

Basic concepts

To appreciate the difference between measurements of displacement, velocity and acceleration, visualise a car travelling along a road. Displacement represents the distance travelled (total mileage), velocity measures the speed of travel (mph), and acceleration describes the degree of movement made by the accelerator pedal.

The monitoring of displacement measures the physical movement of a vibrating object in mils or microns. Of the three measurements above, displacement is the easiest to interpret from visual data; movements can be clearly represented in sharp peaks and troughs on a sinusoidal ‘time domain’ graph to show the frequency and distance (or amplitude) of movement.

Velocity measures the speed at which the mass is moving in inches or millimetres per second. Velocity is recorded on a graph as a wave moving above and below a line that represents where speed is greatest. The highest and lowest points on the curve represent zero velocity (the points at which the object comes to a stop before moving in the opposite direction).

Acceleration measurement records the rate of change in velocity. Here, the graph represents movement above and below a line that represents zero movement, with the highest point on the curve describing where acceleration is greatest and the lowest point tracing where the object has decelerated and stopped and is about to begin accelerating again.

Measurement methodology in practice

So how and when are these techniques used in practice? In summary, displacement is typically used for measuring low frequencies where there is a low RPM, velocity is applied over a wide frequency range and acceleration is used for measuring high frequencies. In practice it may be that a range of methods are applied within the same system. To illustrate further, let’s look at some application examples.

Take, for example, a fan in a cooling tower. At the heart of each cooling tower is a fan system, which effectively acts as an evaporative heat sink. This system typically comprises one or more rotating fans – each often several metres in diameter – plus drive motors, gearboxes and associated linkages and control mechanisms. Almost all of these will have moving parts, which must be installed and aligned correctly to ensure consistent operation. They will, however, be subject to normal wear and tear over time and, most importantly, a wide range of fluctuating environmental conditions that can significantly increase wear rates; depending on the site of the cooling towers, environmental conditions can include high or gusting wind speeds, with down and updrafts through each tower, plus varying levels of dust, sand and, in coastal locations, high concentrations of salt. Monitoring the efficiency of such critical but potentially vulnerable systems is an essential element if availability or uptime is to be optimised, and this may require a range of monitoring methods.

Maintenance engineers have several options for monitoring cooling tower fan systems, each of which will provide an accurate, efficient and reliable method of determining changes in the frequency and amplitude of vibration signals, and thus the rate of wear in rotating parts.

In each case, vibration sensors or accelerometers are fitted on, or adjacent to, critical rotating surfaces. For example, vibration sensors may be mounted vertically on the drive end of the motor to measure peak velocity in millimetres per second and determine bearing wear and imbalance. A sensor may also be mounted axially on the motor output shaft - again to measure peak velocity in millimetres per second - to determine shaft alignment and axial thrust. Other parts of the system may also require peak velocity to be measured, such as the gearbox, but a horizontal sensor located close to the fan drive may measure displacement in microns and show fan imbalance and wear.

To take a different example, the monitoring of a bearing in a paper mill reveals the advantages of measuring acceleration. The ultra-low-speed bearings in pulp and paper mills can operate at 10rpm or less; this presents a challenge because, as rotational speeds drop, so does the energy from vibration signal frequencies caused by bearing defects, making them much harder to detect. If a bearing has a defect on its outer raceway the contacts made by each of the bearing’s rolling elements generate small vibration signals with energy content in higher frequencies (i.e. several kHz). This signal often gets lost in machine “noise” but Acceleration Enveloping can detect such signals, enabling effective diagnosis of a developing bearing fault and scheduled maintenance that could prevent costly downtime.

Increases in machinery vibration indicate deteriorating operating conditions, such as wear or misalignment, and vibration sensors can identify these changes swiftly and with exceptional reliability. By understanding the basic concepts and methods of vibration monitoring you can yield the most repeatable and consistent measurement of vibration, providing machine reliability data that can be easily analysed to predict potential problems before they occur.

For further information please visit: