Large Eddy Simulations of a Turbocharger Radial Turbine Under Pulsating Flow Conditions

The pulsating flow conditions which a turbocharger turbine is exposed cause important deviations of the turbine aerodynamic performance when compared to steady flow conditions. Indeed, the secondary flows developing in the turbine are determined by the inflow aerodynamic conditions, which largely vary during the pulse cycle. In this paper, a high-resolved Large Eddy Simulation is performed to investigate and characterize the flow field evolution in a turbocharger radial turbine over the pulse cycle. At first, the model is validated against experimental results obtained in gas-stand flow conditions. Then, the instantaneous flow field at the rotor mid-span section is compared to the one given by the equivalent cycle-averaged steady flow conditions. The results highlight five distinct flow features. At low mass flow rates, when the relative inflow angle assumes large negative values, the flow separates at the blade pressure side, causing a secondary flow consisting in two counter-rotating vortices characterized by a diameter comparable to the blade passage. As the mass flow rate increases, the first vortex persists at the blade tip while the second one moves closer to the blade trailing edge. This corresponds to the second characteristic flow field. With increasing relative inflow angle, for the third characteristic flow feature, only the recirculation at the blade leading edge is displayed and its size gradually reduces. For the fourth characteristic flow feature, at moderate negative values of the relative inflow angle, the flow field is well aligned with the blade profile and free of secondary flows. Then, as the relative inflow angle gradually grows towards large positive values, the flow separates on the blade suction side causing the mixing of the flow with the stream flowing on the pressure side of the previous blade.

Development and Validation of a Machine Learned Turbulence Model

A stand-alone machine learned turbulence model is developed and applied for the solution of steady and unsteady boundary layer equations, and issues and constraints associated with the model are investigated. The results demonstrate that an accurately trained machine learned model can provide grid convergent, smooth solutions, work in extrapolation mode, and converge to a correct solution from ill-posed flow conditions. The accuracy of the machine learned response surface depends on the choice of flow variables, and training approach to minimize the overlap in the datasets. For the former, grouping flow variables into a problem relevant parameter for input features is desirable. For the latter, incorporation of physics-based constraints during training is helpful. Data clustering is also identified to be a useful tool as it avoids skewness of the model towards a dominant flow feature.

Efficiency improvement of English online teaching system based on bagging learning flow feature selection

In the era of artificial intelligence, the traditional English teaching model can no longer meet the needs of society, and online English teaching has become the main development direction of English teaching in the future. In order to study the efficiency of English online teaching system, based on machine learning algorithms, this paper constructs an efficiency improvement model of English online teaching system. Moreover, in view of the shortcomings of current situation estimation algorithms that cannot coexist in terms of flexibility, causal interpretability and complexity, this paper proposes a biological immune algorithm framework that uses GBDT algorithm coding, which objectively and accurately shows the spread of the situation. In addition, for the problem that redundant information between features will reduce the accuracy of the framework, this paper proposes a streaming feature selection algorithm based on bagging learning. Finally, this paper designs a control experiment to analyze the performance of the model. The research results show that the model constructed in this paper is highly reliable.

Numerical investigation of total temperature distortion problem in a multistage fan based on body force approach

This paper applies a body force model developed recently to investigate the interaction between total temperature distortion and a multistage fan. The off-design performance of the fan shows the reasonable predicting accuracy and supports the present model is applicable for high-speed multistage machines. The transfer behaviors of 90° steady-state circumferential total temperature distortion as well as combined total pressure and total temperature distortion in the multistage environment are captured successfully by the model. The mechanism of the phase shift of the high temperature sector is discussed by the model to advance the understanding of the total temperature distortion problem. The results reveal that the large-scale flow feature of total temperature distortion in the multistage environment can be capably quantified by the present body force model with the acceptable computational consumption.

Feedbacks between a non-Newtonian upper mantle, mantle viscosity structure and mantle dynamics

SUMMARY Previous studies have shown that a low viscosity upper mantle can impact the wavelength of mantle flow and the balance of plate driving to resisting forces. Those studies assumed that mantle viscosity is independent of mantle flow. We explore the potential that mantle flow is not only influenced by viscosity but can also feedback and alter mantle viscosity structure owing to a non-Newtonian upper-mantle rheology. Our results indicate that the average viscosity of the upper mantle, and viscosity variations within it, are affected by the depth to which a non-Newtonian rheology holds. Changes in the wavelength of mantle flow, that occur when upper-mantle viscosity drops below a critical value, alter flow velocities which, in turn, alter mantle viscosity. Those changes also affect flow profiles in the mantle and the degree to which mantle flow drives the motion of a plate analogue above it. Enhanced upper-mantle flow, due to an increasing degree of non-Newtonian behaviour, decreases the ratio of upper- to lower-mantle viscosity. Whole layer mantle convection is maintained but upper- and lower-mantle flow take on different dynamic forms: fast and concentrated upper-mantle flow; slow and diffuse lower-mantle flow. Collectively, mantle viscosity, mantle flow wavelengths, upper- to lower-mantle velocities and the degree to which the mantle can drive plate motions become connected to one another through coupled feedback loops. Under this view of mantle dynamics, depth-variable mantle viscosity is an emergent flow feature that both affects and is affected by the configuration of mantle and plate flow.

Full Annulus Analysis of Inlet Distortion Effects on the Performance of a Compressor

The paper presents a numerical investigation of inlet distortion effects on the performance of a whole compressor. The compressor system consists of three and a half stages: a fan stage, an axial stage, a mixed-flow stage and an outlet guide vane. There are two bypasses after the fan stator. The following two and a half stages are in the core flow. The inlet distortion is generated by blocking part of the upstream inlet duct using a plate. The inlet distortion breaks up the original axisymmetric flow feature, thus a whole annulus computational domain has to be employed throughout the numerical unsteady analyses. The commercial GPU-based software of Turbostream is used to conduct all the analyses on several GPU workstations. The purpose of the numerical analysis is to understand how the inlet distortion propagates within this compressor and how the performance of each stage and the whole compressor is affected by the inlet distortion qualitatively and quantitatively. To achieve this goal, two separate numerical analyses have been performed for the compressor with and without inlet distortion at an interested operation point. For these two simulations, we do DFT analysis for the total pressure distribution at three spans of several different axial positions, such as the inlet and outlet of each stage, to get the spatial harmonic characteristic of each stage. The finding provides the insight needed to redesign the relevant stage to increase the surge margin.