Davood
Altafi

This study compares the local losses of a radial compressor in the range from surge to choke considering shock phenomena, boundary layer separation, and mixing mechanisms. For this purpose, formulation of the local entropy generation rate (EGR) is added to the computational fluid dynamics (CFD) solver, which models the turbulent flow field of the compressor through the RANS approach. For validation, the numerical pressure rise curve of the compressor is compared with experimental data.

The results indicate that at the design point, the impeller, diffuser, and volute account for approximately 50.8%, 30.0%, and 12.3% of the EGR, respectively, with approximately 5% in the impeller backspace, 1% in the diffuser cavity, and less than 0.5% in the inlet duct. Approaching the surge condition, local losses due to mixing and shock waves decline while boundary layer losses increase.

Based on comprehensive analysis of the leading-edge boundary layer, the largest dead-air zone is found at the design point, resulting in a lower diffusion entrance loss and a higher mixing loss. Furthermore, EGR variation in the diffuser channel is investigated through mixing dynamics and classification of the channel flow into three distinct zones based on the newly introduced mixing ratio (MR). The outcomes show that the flow regime mixes more rapidly at higher MR, resulting in an EGR decline that directly benefits overall compressor performance.

Local entropy generation accounts for irreversible energy degradation through viscous shear dissipation and thermal conduction across temperature gradients:

$\dot{S}'''_{g,\text{local}} = \frac{\lambda}{T^2}\left(\frac{\partial \bar{T}}{\partial x_i}\right)^2 + \frac{\alpha_t}{\alpha}\frac{\lambda}{\bar{T}^2}\left(\frac{\partial \bar{T}}{\partial x_i}\right)^2 + \frac{2\mu \bar{S}_{ij}\bar{S}_{ij}}{\bar{T}} + \frac{\epsilon \bar{\rho}}{\bar{T}}$

To quantify the relative role of thermal dissipation versus viscous friction, the dimensionless Bejan number ($Be$) is calculated:

$Be = \frac{\dot{S}''_{\text{Thermal}}}{\dot{S}'''_{g,\text{local}}}$   ($0 \le Be \le 1$)

$Be \to 1$ indicates pure heat-transfer irreversibility, while $Be \to 0$ denotes dominance of mechanical viscous dissipation and turbulent shear stresses.

Radial flow compressors in internal combustion engine turbochargers are frequently constrained by low-flow aerodynamic instabilities such as rotating stall and surge. Casing treatment mechanisms, particularly bleed slots and ported shroud systems, provide a passive aerodynamic mechanism to bleed stalled tip leakage flows and reinject them upstream. However, conventional bleed slot designs often suffer from an isentropic efficiency penalty at design and high-flow conditions.

This study performs a comprehensive multi-objective optimization of a radial compressor bleed slot mechanism. By combining high-fidelity 3D RANS simulations, response surface surrogates, and multi-objective evolutionary algorithms (NSGA-II), the trade-off between stall margin extension (SME) and peak isentropic efficiency is systematically resolved.

Manufacturing tolerances cause physical impeller dimensions to deviate from CAD master models. This article investigates the statistical propagation of eight geometric tolerances on the pressure ratio and isentropic efficiency of a turbocharger compressor for the 1.7-liter EF7 engine.

Using Latin Hypercube Sampling (LHS) and an accurate Kriging surrogate model calibrated against 3D RANS CFD simulations, global sensitivity indices were derived using the Spearman rank correlation coefficient.

The performance of the radial flow compressor as a machine consisting of several components is always affected by uncertainties and its performance is always measured in an unpredictable range. Among the types of uncertainties, we can mention changes in environmental conditions and manufacturing tolerances that affect the performance curves. Although it is impossible to accurately estimate the final performance under the influence of such uncertainties, it is possible to calculate the range of compressor performance. The calculation of these intervals is based on statistical calculations and is known as uncertainty quantification (UQ).

In this research, first, with the help of entropy production analysis, various aerodynamic mechanisms affecting the performance and stability of the radial flow compressor will be investigated. Then, with the help of the UQ method, over two separate studies, the effect of changes in operating conditions and manufacturing tolerances on the performance of the compressor will be analyzed.