This thesis presents a modern method to evaluate spur gears based on the transientudelastohydrodynamic lubrication (EHL) emulation of the full meshing cycle, evaluatingudelastic stresses in the gear flanks, collecting the stress history and applying stress andudstrain-life methods to calculate fatigue parameters and cumulative fatigue damage, i.e.udpredicting the fatigue life taking measured surface roughness into account.udThe EHL model is formulated as the coupled system of the hydrodynamic Reynoldsudequation and the elastic deflection equation. These are solved simultaneouslyudincluding the transient effect by incorporating the squeeze film term of the Reynoldsudequation with a Crank-Nicolson discretization of time. The finite differenceuddiscretisation of the elastic deflection equation utilises the differential form firstudformulated at Cardiff to allow coupling of the equations. The Reynolds equation canudbe discretised either by a finite difference or by a finite element method. The coupledudsystem is solved simultaneously either by a narrow bandwidth Gaussian eliminationudor a Gauss-Seidel iterative method.udThe elastic stresses due to the superimposed discrete values of the EHL pressure andudshear stress at the EHL mesh nodes are evaluated by carrying out the necessaryudconvolution of the stresses by a Fast Fourier Transform method. The weightingudfunctions required have been calculated analytically. The stresses are obtained on theudEHL solution mesh and are interpolated to meshes fixed in the pinion and the gearudflanks. They are then sorted and stored efficiently to enable fatigue life predictionudalgorithms to be applied.udA detailed description of the EHL and the stress evaluation models are provided asudwell as a brief description of some fatigue life theories and calculations. The results ofudthe complete analysis are provided for test gears obtained from the NASA Glennudlaboratory fatigue tests and the Newcastle University Design Unit micro-pittingudinvestigation. The analyses were carried out for real operating conditions from gearudtesting under extreme conditions. The surface roughness profiles used were realudmeasured profiles taken from the test gears after initial running-in. The simulationsudreported are therefore as realistic as can be achieved and represent the true mixedudlubrication conditions occurring in heavily loaded gears. The research also shows theudimportance of precise alignment of the roughness profiles in these conditions.
展开▼
