Studies On Dynamic Affect Factors Of Propulsion Shaft And Fault Diagnosis Method Based On EMD

The stiffness of recently designed marine propulsion shafting has been increasingremarkably, whereas hull structures have become more likely to deform as a result ofcutting-age optimized design of the scantlings and the use of high tensile steel with shipdeveloping toward largely. The offsets of the support bearings of shafting are more likely tovary under different operating conditions. On the other hand, the increased stiffness ofshafting makes shafts less adaptable to any small departure of the bearing lines from theinitial lines. This combination is thought to be the main cause of recently reported cases ofalignment related main bearing damage. The research on the dynamical characteristics ofcomplicated marine propulsion shafting is need for the fault mechanism which is still unclear.The many dynamic affected factors and their action mechanism on shafting alignment areresearched in this thesis. The earlier fault diagnosis on the ship propulsion engine and shaftingare studied. The vibration signals of a gear with pitting fault and a rolling bearing withinner-race fault have also been obtained on the rotor test-bed. The fault characters areextracted effectively by the method of the empirical mode decomposition (EMD) and themethod of the improved ensemble empirical mode decomposition (EEMD). All these madethe foundation for real propulsion shafting fault diagnosis.The diameter of equivalent circular bar representing crankshaft is determined by FEMmethod. The analysis result shows the bearing loads is affected by the position of crankshaftangle and each crank angle. The equivalent circular bar can represent crankshaftapproximately, but the bearing loads can not be exactly same. The ratio of best diameter ofequivalent circular bar is different with crankshaft, so it should be taken different valueaccording to the given crankshaft. when the number of main bearing that the reactions needs to beaccurately evaluated is known, two more main bearings must have to be incorporated in thecalculation model.The optimization algorithm of the location and the offsets of bearing is been put forward,which is based on the shafting stiffness matrix constructed by arraying the reaction influentialnumbers in a certain order. By this method, the bearing offsets can be evaluated accordingtarget reactions. The location and the offsets of intermediate bearing can be optimized when thelocation of stern tube bearing and the main bearing is determined.The main engine bearings may be damaged by unloaded due to the effects of changes in hull girder deflection. Therefore, it is important to research the influence of relativedisplacement of the hull to main bearing load when the ship is in the light draught conditionand the full draught condition. Take the hull deflection at the aftmost bulkhead of the engineroom as the reference parameter, and the calculation method was provided when the mainbearing unloaded. The possibility of the main bearing damage can be judged by comparingthe allowable limit of the reference parameter. This method was proved feasibility by takingthe176,000DWT bulk carrier as an example.The2-D plane method is usually applied in calculating hull deformation in order to verifyship construction strength. But it is not enough accurate to analyze the effect on propulsionshaft system by hull deformation. The3-D FE model of a bulk carrier is carried out in thispaper. The wave pressure is calculated and loaded automatically based on Patran commandlanguage (PCL). The deformations of ship construction in various sea waves are calculatedand the effect to bearing load is analyzed. The deflection of hull under shaft line is small inthe sea state4and under the sea state4. The deflection of hull under shaft line is increased bysea state. And it is more than4mm in the sea state8, the shaft bearings load will be affectedenormously. The deflection of hull under shaft line in vertical direction is increased when thecourse heading30°, and it is also in horizontal and longitudinal direction.The bearing’s oil film is the one of the most important factors that affected the dynamicalignment. Numerical analysis method is applied to solve the radical bearing’s Reynoldsequation in this paper. The relations of the Sommerfeld number and the attitude angle to theratio of B/D and the relative eccentricity are obtained from calculations. As an example, theshaft line alignment of the115000T cargo ship is analyzed by ANSYS and MATLABsoftware. The propeller submerge conditions are taken into account. The relations of therelative eccentricity and the attitude angle to the rotational speed of journal, and the verticaland the horizontal eccentricities of journals are also calculated. The affect of the bearing oilfilm to the bearing reaction is calculated and analyzed. The result shows that the affect isdifferent from various rotational speed. The affection is decreased by speed increased.By using a three-dimensional model of the shafting and bearing, nonlinear factorsincluding bearing stiffness and clearance, bending deformation of ship hull, ThermalDeformation of bearing, Bending moment due to hydrodynamic propeller forces, Bendingmoment that acts on thrust shaft. The integration stiffness of bearing include hull stiffness,bearing stiffness and bearing’s oil film stiffness which in same quantity level or in near quantity level.The three-dimensional crankshaft finite element model is created. The followingproperties are included: rotating and oscillating mass for each cylinder, main bearing stiffnessand thrust bearing stiffness, bearing clearance, ship construction flexibility and deformations,bearing thermal deformation, bending moment that acts on propeller shaft and thrust shaft, itis possible to achieve a more accuracy results.An improved EEMD method is proposed which can raise the SNR of the reconstructedsignal. The simulation shows the effect of the EEMD which include adding and subtractingnoise is better than the EEMD which include only adding noise. The best ratio range of thestandard deviation of the added noise and the best ensemble number is found out. The signalis decomposed into IMFs with these best parameters. The new signal is reconstructed by thoseIMFs relating mostly with the signal. The SNR of the new signal can be improved by takingthis new signal into next EEMD procedure. A new fault diagnosis method based on thisimproved EEMD and instantaneous energy density spectrum is proposed here. Thecharacteristic frequencies are found in the instantaneous energy density of Hilbert spectrum.The effectiveness of this method was demonstrated by analysis the vibration signals of a gearwith pitting fault and a rolling bearing with inner-race fault.