Identification of Dynamic Coefficients of a Five-Pad Tilting Pad Journal Bearing Up to Highest Surface Speeds

Abstract

Abstract Tilting-pad journal bearings are widely used in turbomachinery industry due to their positive dynamic properties at high rotor speeds. However, the exact description of this dynamic behavior is still part of current research. This paper presents the measurement results for a five-pad tilting-pad journal bearing in load between pivot configurations. The bearing is characterized by a nominal diameter of 100 mm, a length of 90 mm, and a pivot offset of 0.6. Investigations include the results for surface speeds between 25 and 120 m/s and specific bearing loads ranging from 0.0 to 3.0 MPa. The results of theoretical predictions are commonly derived from perturbation of stationary operation under static load. Therefore, experimental results for stationary operation including pad deflection under static load are presented first to characterize the investigated bearing. Measured results indicate considerable nonlaminar flow in the upper region of the investigated range of rotor speeds. Second, dynamic excitation tests are performed with excitation frequencies up to 400 Hz to evaluate dynamic coefficients of a stiffness (K) and damping (C) KC model, and additionally, a KCM model using additional virtual mass (M) coefficients. KCM coefficients are obtained by fitting frequency dependent KC-characteristics to the KCM model structure using least square approach. The wide range of rotating and excitation frequencies leads to subsynchronous as well as supersynchronous vibrations. Excitation forces are applied with multisinus and single-sinus characteristics. The latter one allows evaluation of KC coefficients at the particular frequency ratio in the time domain. Here, frequency and time domain evaluation algorithms for dynamic coefficients are used in order to assess their special properties and quality. The impact of surface speed, bearing load, and oil flow rate on measured and predicted KCM coefficients is investigated. Measured and predicted results can be well fitted to a KCM model and show a significant influence of the ratio between fluid film and pivot support stiffness on the speed dependent characteristic of bearing stiffness coefficients. However, the impact of this ratio on damping coefficients is considerably lower. Further investigations on the impact of oil flow rates indicate that a significant decrease of direct damping coefficients exists below a certain level of starvation. Above this limit, direct damping coefficients are nearly independent of oil flow rate. Results are analyzed in detail and demands on improvements for predictions are derived.