On the Effects of the Overhang Rotor of the Exciter on the Twice per Revolution Vibration of Large two-pole Power Generators

Paolo Pennacchi; Steven Chatteron; Andrea Vania; Ludovico Dassi; Andrea Riva

doi:10.37965/jdmd.2026.1596

Authors

Paolo Pennacchi Politecnico di Milano, Department of Mechanical Engineering, Via Giuseppe La Masa 1, 20156 Milan, Italy https://orcid.org/0000-0001-8174-0462
Steven Chatteron Politecnico di Milano, Department of Mechanical Engineering, Via Giuseppe La Masa 1, 20156 Milan, Italy https://orcid.org/0000-0003-4170-009X
Andrea Vania Politecnico di Milano, Department of Mechanical Engineering, Via Giuseppe La Masa 1, 20156 Milan, Italy https://orcid.org/0000-0001-8277-1498
Ludovico Dassi Dept. of Mech. Eng. - Politecnico di Milano https://orcid.org/0000-0001-6217-7483
Andrea Riva Politecnico di Milano, Department of Mechanical Engineering, Via Giuseppe La Masa 1, 20156 Milan, Italy https://orcid.org/0009-0000-3426-8724

DOI:

https://doi.org/10.37965/jdmd.2026.1596

Abstract

In large two-pole power generators, the anisotropy of the active part of the rotor causes twice per revolution (2X) vibrations whose amplitude depends on the rotational speed. When the speed approaches half the first critical bending speed, the so-called 2X resonance is excited. At such low speeds, the damping provided by oil-film journal bearings is often quite low and 2X vibration levels can become considerable. In generators mounted on only two main supports, a long and slender overhang shaft of the exciter can amplify this phenomenon, as its significant bending enhances the dynamic deflection at the midspan of the rotor. This paper investigates the beneficial effects of installing an additional oil-film journal bearing at the end of the overhang shaft of the exciter. A finite element model of a 600 MVA generator, whose equations of motion have been integrated in the time domain, was used to simulate the 1X and 2X vibrations accounting for the speed-dependent properties of the journal bearings and the periodic variations of the bending stiffness of the shaft. The numerical predictions are validated against experimental vibration data collected during speed transients before and after the installation of the additional support. Both experimental and numerical results demonstrate that the additional bearing significantly reduces the peak of the 2X vibration that occur at the main supports.

Conflict of Interest Statement

The authors declare no conflicts of interest.