TRC-SFD-1-07 - Tribology Group - Texas A&M University

More documents

Recommendations

Info

system. The method relies on the principle that the non-linear feedback is directly linkedto the spatial nature of the non-linear system. The application of the method is illustratedon two numerical simulations of single or multiple-input/multiple-output SDOF andMDOF systems. The results show the method is effective in determining the linear andnon-linear system parameters simultaneously.Yang et al. [14] present a technique for diagnosing vibration using sensitivityfunctions that are used to correlate the influence of each system parameter over theoverall system response. For a SDOF system, these functions are simply defined as thepartial derivative of the impedance function with respect to the specific system parameter(i.e. mass, damping, stiffness). The definition of these functions can be extended toMDOF systems. The authors include a practical application of the sensitivity function tolocate and eliminate a vibration problem in a vehicle exhaust system.Yang et al. [15] extend the method in [14] for the parameter identification ofmechanical systems using sensitivity functions that are expressed in terms of the systemtransfer functions. The identification procedure involves the linearization of non-linearsources. Although this implies the a-priori knowledge of the non-linear source, theidentification procedure can be used to evaluate if the system nonlinearities assumed inthe model (i.e. cubic stiffness, quadratic damping, etc.) characterize the system response.The application of the method is illustrated numerically with a 2-DOF system, includingan element with a non-linear quadratic stiffness, which is excited with periodic (singlefrequency) load excitations. The numerical results demonstrate the method effectivenessin determining the degree of freedom in which the nonlinearity is present, and invalidating the modeling assumption (i.e. quadratic stiffness).The present identification method is an adaptation of the Rice and Fitzpatrick [9]method, and also uses definitions from a similar implementation of the method given bySan Andrés and Aguilar [16].10
III Test Rig DescriptionFigure 1 depicts a schematic view of the test rig also described in the prior TRCreport [3]. Following the same description, the test rig consist of a vertical rigid shaft,mounted on three precision ball bearings (natural frequency 400 Hz [3]), which holds asteel journal of 5” (127 mm) diameter and 3” (76.2 mm) long. The bearing assemblyincludes two steel plates clamping an acrylic bearing. The two horizontal plates areattached by two vertical steel plates, which also serve as an interface to apply externalforces onto the bearing assembly. The top plate includes a lubricant supply connection, astatic pressure gauge displaying the feed pressure into the bearing and four eddy currentsensors facing the shaft. The composite bearing housing hangs from a top structure withfour steel rods providing structural stiffness to the test bearing section. A mechanism atopof the test rig, comprising two sliding flat plates (top and bottom support plates), allowsadjusting the position of the bearing center with respect to the shaft to simulate centeredand off-centered operation conditions.Top SupportPlateMain FramesupportsSteelRodsBottom SupportPlateAccelerometerOscilloscopeXAcrylicHousingShakerSignalDischargeorificeRing carrierJournalVerticalplateLoad CellPowerSupplyPC DAQShaft1)127mm (5 in)ShakerSignal inputShaker Cooler FanTo ComputerFigure 1 Test rig for dynamic force measurements and flow visualization in a sealed endSFDThe bearing housing design integrates a SFD land and an annulus that accommodatesa metallic ring (ring carrier). A wave spring pushes the ring holder against the journal and11
Page 1 and 2: Texas A&M UniversityMechanical Engi
Page 3 and 4: Table of ContentsExecutive Summary.
Page 5 and 6: Figure 22 Dynamic pressure measurem
Page 7 and 8: I IntroductionSqueeze film dampers
Page 9: A review of publications related to
Page 13 and 14: Oil inletEddy current sensorHousing
Page 15 and 16: 300Y Load [N](90 N,82 N)(98 N,88 N)
Page 17 and 18: M sM fK sxC SFDxxC SFDyyF djournalF
Page 19 and 20: 2( ω−ω− )F = L x + F u ; L =
Page 21: configurations including open outle
Page 24 and 25: capacity of 4.1 kJ. In actual appli
Page 26 and 27: V Onset amplitude and frequency lea
Page 28 and 29: Figure 14 Excitation load and respo
Page 30 and 31: V.2 Experimental ResultsFigure 17 s
Page 32 and 33: 1.5Pk-Pk Pressure (bar)10.532 μm50
Page 34 and 35: Absolute Pressure [bar]24001.75Pres
Page 36 and 37: VI Conclusions and RecommendationsT
Page 38 and 39: VII References[1] Delgado, A., and
Page 40 and 41: Appendix A Identification of dry fr

TRC-SFD-1-07 - Tribology Group - Texas A&M University

Create successful ePaper yourself

Delete template?

Save as template?