In the presence of small-amplitude pressure oscillations, the linearized Navier{Stokes equations are solved to obtain an accurate description of the time-dependent eld in a channel having a rectangular cross-section and two equally permeable walls. The mean solution is based on Taylor's classic pro le, while the temporal solution is synthesized from irrotational and rotational elds. Using standard perturbation tools, the rotational component of the solution is derived from the linearized vorticity transport equation. In the absence of an exact solution to rely on, asymptotic formulations are compared with numerical simulations. In essence, the analytical formulation reveals rich vortical structures and discloses the main link between pressure oscillations and rotational wave formation. In the process, the explicit roles of variable injection, viscosity and oscillation frequency are examined. Using an alternative methodology, both WKB and multiple-scale techniques are applied to the linearized momentum equation. The momentum equation is of the boundary-value type and contains two small perturbation parameters. The primary and secondary parameters are, respectively, the reciprocals of the kinetic Reynolds and Strouhal numbers. The multiple-scale procedure employs two ctitious scales in space: a base and an undetermined scale. The latter is left unspeci ed during the derivation process until ®ow parameters are obtained in general form. Physical arguments are later used to de ne the arbitrary scale, which could not have been conjectured a priori. The emerging multiple-scale solution o¬ers several advantages. Its leading-order term is simpler and more accurate than other formulations. Most of all, it clearly displays the relationship between the physical parameters that control the nal motion. It thus provides the necessary means to quantify important ®ow features. These include the corresponding vortical wave amplitude, rotational depth of penetration, near-wall velocity overshoot and surfaces of constant phase. In particular, it discloses a viscous parameter that has a strong in®uence on the depth of penetration, and furnishes a closed-form expression for the maximum penetration depth in any oscillation mode. These ndings enable us to quantify the location of the shear layer and corresponding penetration depth. By way of theoretical veri cation, comparisons between asymptotic formulations and numeric predictions are reassuring. The most striking result is, perhaps, the satisfactory agreement found between asymptotic predictions and data obtained, totally independently, from numerical simulations of the nonlinear Navier{Stokes equations. In closing, a standard error analysis is used to con rm c

No abstract

In the combustion stability assessment of solid propellant rocket motors, several new destabilizing terms are introduced when rotational-flow effects are properly accounted for. Such effects must be included when the wave motion is parallel to the burning surface. A normal fluctuating velocity component then appears in a careful resolution of intrinsic fluid dynamics, including acoustico-vortical interactions that must satisfy mass and momentum conservation principles while accommodating the no-slip condition at the propellant surface. The source of this destabilizing term appears explicitly in two separate, independently derived, analytical formulations of the internal flowfield. Predictions generated by these analytical models are shown to agree with reliable computational data produced recently by a numerical code that solves the unsteady nonlinear Navier-Stokes equations. Verification of the analytical formulations by means of theoretical considerations, numerical comparisons, and global error assessments are also undertaken before examining the impact of the new time-dependent radial-velocity correction on rocket stability. The new radial-velocity fluctuations introduce a correction comparable in importance to the classical pressure coupling at the propellant surface. This effect along with several companion terms must be accounted for properly in the assessment of motor stability characteristics.

No abstract

A hybrid method of an artificial neural network (ANN) combined with a support vector machine (SVM) has been developed for the defect diagnostic system applied to the SUAV gas turbine engine. This method has been suggested to overcome the demerits of the general ANN with the local minima problem and low classification accuracy in case of many nonlinear data. This hybrid approach takes advantage of the reduction of learning data and converging time without any loss of estimation accuracy because the SVM classifies the defect location and reduces the learning data range. The results of test data have shown that the hybrid method is more reliable and suitable algorithm than the general ANN for the defect diagnosis of the gas turbine engine.