Advanced search
Start date

Model based identification for Multi-Fault detection in rotating machinery

Grant number: 19/25906-9
Support type:Research Grants - Visiting Researcher Grant - International
Duration: March 01, 2020 - April 30, 2020
Field of knowledge:Engineering - Mechanical Engineering - Mechanics of Solids
Principal Investigator:Katia Lucchesi Cavalca Dedini
Grantee:Katia Lucchesi Cavalca Dedini
Visiting researcher: Nicolo Bachschmid
Visiting researcher institution: Politecnico di Milano, Italy
Home Institution: Faculdade de Engenharia Mecânica (FEM). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil
Associated research grant:15/20363-6 - Fault tolerant identification and control of rotating systems, AP.TEM


The objective of the FAPESP thematic project entitled Fault Tolerant Identification and Control of Rotating Systems (grant #2015/20363-6), as the name suggests, is the development of a system capable of identifying faults in a rotating system in order to control or compensate them. This control, or compensation, must allow the faulty machine to remain operational in its design condition until the next scheduled maintenance, or to be safely shut down, in the case of a severe fault with the need of an emergency maintenance. In this sense, such system can be divided into four steps. First, a vast acquisition of information through the sensors is accomplished and all the collected data is filtered and processed. Next, the search for faults in the machine from the processed signals is done. If any fault is detected, the third step consists in identify the fault type, its location, and severity degree. Finally, a magnetic bearing is used to compensate the effect of the fault and allow the machine continue a safe operation. Therefore, one important task of the project will be also the separation in the identification process of the effects due to different faults from bearing wear.The possibility of identifying two or more different faults on the same rotor system, in other words "multiple fault identification" is investigated. Therefore, during the visiting of Prof. Bachschmid, multiple fault identification will be applied theoretically and experimentally by combining different tools, also in conditions where the usual methods cannot be used. (AU)