Uncertainty Quantification in Hydrodynamic Bearings

Duccio Griffini, Simone Salvadori, Mauro Carnevale, Francesco Montomoli

Research output: Contribution to journalConference articlepeer-review

1 Citation (Scopus)

Abstract

Although it is possible to imagine a strong sensitivity of hydrodynamic bearings performance to geometrical and fluid dynamic uncertainties, a small amount of scientific contributions have been found in literature about the use of uncertainty quantification techniques for the numerical modeling of bearings. In the present paper we aim at quantifying the effects of the aleatory uncertainty of some relevant input values on key parameters related to rotordynamic effects in turbomachinery, and in particular on the rotor thermal instability problem (e.g. the equilibrium position and the dynamic coefficients). A methodology is initially developed in order to study the propagation of the uncertainties in the numerical analysis of Tilting Pad Journal Bearings (TPJB). Due to the characteristics of the in-house finite element code TILTPAD considered for the UQ analysis, the Monte Carlo method has been selected among the possible approaches. The analysis here presented considers the effects of both manufacturing tolerances on the assembled bearing clearance and of the tolerances adopted for the characterization of the viscosity grade of the oil. The test case adopted for the analysis is the Kingsbury D-140 TPJB. Considering the individual variation of the selected parameters, it is possible to observe that the standard deviation (STD) of the the non-dimensional dynamic coefficients is up to 2.1% in case of viscosity variation and up to 9.1% in case of clearance variation. The STD of the frictional power losses is about 2.2% and 1.4% respectively. Considering the simultaneous variation of the selected parameters, it is possible to observe a STD of the non-dimensional dynamic coefficients comprised between 6.4% and 9.4%, while the STD of the frictional power losses is about 2.7%.

Original languageEnglish
Pages (from-to)884-891
Number of pages8
JournalEnergy Procedia
Volume148
Early online date29 Oct 2018
DOIs
Publication statusE-pub ahead of print - 29 Oct 2018
Event73rd Conference of the Italian Thermal Machines Engineering Association, ATI 2018 - Pisa, Italy
Duration: 12 Sep 201814 Sep 2018

Keywords

  • Manufacturing Tolerance
  • Monte Carlo
  • Oil Viscosity
  • Tilting Pad Journal Bearing
  • Uncertainty Quantification

ASJC Scopus subject areas

  • Energy(all)

Fingerprint Dive into the research topics of 'Uncertainty Quantification in Hydrodynamic Bearings'. Together they form a unique fingerprint.

Cite this