Much of the rotor dynamic behaviour studied in the past has been focussed on large equipment like steam turbine driven generators, where resonant frequencies related to specific mode shapes, leading to first, second or even third critical speeds are the primary concern.  However, this is not the only type of behaviour possible, and smaller equipment like many commonly occuring pumps, fans and compressors can more commonly exhibit a range of other types of behaviour. Recent mathematical studies have shown that there is a whole range of sub-harmonic behaviour that is triggered by rubbing friction, or distress in bearings, or other non-linear phenomena.  Some of this behaviour shows up particularly clearly in the torsional domain. The measurement of Torsional Vibration has traditionally been difficult to do, requiring special torque sensors or speed sensors to be placed on the rotating shaft, and has been confined only to special situations where a particular problem was thought to exist.  Most condition monitoring practitioners never have the chance to become familiar with it as a technique, nor learn what conditions can be detected by it. However, much industrial equipment has a built-in torque sensor in the form of the electric motor driving the equipment.  By clever analysis of the subtle variations in current drawn by the motor, a wide range of phenomena can be detected, that bear many similarities to what can be detected by conventional Vibration Analysis.  Model-Based Voltage and Current (MBVI) systems can allow you to monitor otherwise difficult to monitor equipment, such as borehole pumps, submerged pumps, in-tank pumps, fans mounted entirely within ductwork, etc.

This talk will share some of this new mathematical understanding of rotor dynamic behaviours that lead to sub-harmonics and show examples of the sort of behaviour in real-world equipment, that is captured and identified by MBVI systems.  It will also explain how MBVI systems work, how you use them, what the outputs look like, how to interpret them and some brief case studies.

Takeaways

1. Torsional vibration and conventional vibration provide complementary views
2. Commonly occurring equipment can show subharmonics related to a range of non-linear phenomena, particularly including rubbing friction and distressed bearings
3. MBVI systems are one way of seeing some of these signals, providing a help to diagnosis.

Geoff Walker, Operations Director of Faraday Predictive Ltd, is a Mechanical Engineer with a focus on the Optimisation of Maintenance Strategy.
He spent the first decade+ of his career working in the Chemicals industry, working with toxic, flammable and corrosive materials and many other things that all challenged the reliable performance of equipment. It was in this environment that he was first exposed to Condition Monitoring and RCM, and where he first encountered how the current drawn by an electric motor can be used to diagnose mechanical problems of inaccessible equipment.
He spent the next decade+ working as a management consultant, helping a wide range of large, asset-intensive organisations optimise their Physical Asset management and maintenance management, where the objective was always getting more for less. He ended up as Director leading PwC’s Physical Asset Management practice for EMEA. He has spent the last decade+ developing and applying smart systems for condition monitoring and fault diagnosis using the electric motor driving the equipment, or the generator being driven by equipment, as a sensor, using the subtle changes in the relationship between current and voltage to diagnose a wide range of mechanical and operational problems as well as electrical ones. Faraday Predictive works closely with Cambridge University, developing new methods for fault diagnosis from the fundamental mathematics of rotor dynamics. Geoff sits on the ISO working group developing international standards for condition monitoring, and chairs the BINDT (British Institute of Non-Destructive Testing) panel on electrical condition monitoring.