This Research Group focuses on general problems inherent to real rotating machinery and traces an ideal path that:
- starts with the mechanical design of the rotor and its related components;
- passes through the set-up of the machine and related start-up problems;
- continues with condition monitoring of the machine;
- eventually ends with the diagnosis and identification of possible faults, with special attention being paid to the early detection of faults.
This general approach is devoted to discovering methods and models that are, in any case, related to both real and industrial applications, with follow-up of the experimental validation of the theory. The use of specially designed test-rigs or actual case studies and data obtained from industrial partners usually means that validation is based on well-grounded data. Furthermore, it is also worth pointing out that the test-rigs employed and described hereinafter are of a fairly large scale and designed to simulate the behaviour of either real machines or real components.
In order to cover the topics of the research line, four main research themes are operative:
- simulation of the dynamic behaviour of industrial rotating machines;
- identification and diagnosis of industrial rotating machines and their components;
- condition monitoring of industrial rotating machines;
- dynamic behaviour of rotating machine components, such as roller and oil-film bearing, blades etc.
Thanks to the support of industrial partners producing rotating machinery and their components, great expertise has been gained with various design problems and different types of motors, turbines for power generation (hydraulic, steam and gas) or special high-speed machines, such as multishaft geared compressors, turbo-molecular pumps or atomizers. In all these cases, simulation models and specific software tools have been successfully developed and industrially tested.
Conversely, with partners involved in the use of rotating machinery, operating problems have been investigated among which effective condition monitoring is the most important in order to avoid impending failures or malfunctions (which, in some cases, could be potentially catastrophic or very dangerous) or, in the event of a failure or malfunction occurring, to quickly identify and repair it. These two requirements are a valid reason to create a service system dealing with condition monitoring, bearing diagnostics, balancing and on-site problem solving based on individual customer requirements, the aim being to define specific alarm criteria and to develop a model based method for fault identification, mainly in the frequency domain. Not only does this method have the advantage of identifying the type of fault but also its severity (such as, for example, unbalance or crack depth) as well as location along the shaft-line (e.g. which bearing suffered a failure or which sealing is too tight or badly assembled thus causing friction when the machine is subjected to critical speeds during run-ups or coast-downs).
ContactsProf. Paolo Pennacchi