Effect of Sweep on Performance of an Axial Compressor With Casing Grooves

Abstract

The need of increased stall margin is very high for aero gas turbine engines, as they operate under varied operating conditions. A number of different options are being used to increase the stall margin of gas turbine engines. Circumferential casing groove, in the compressor section of a gas turbine engine, is one of such methods. Incorporation of the grooves on the shroud increases the stall margin of the compressor, but this generally gives rise to loss of performance, such as efficiency and pressure ratio. By employing 3D blading techniques for rotor blades as well as stator vanes, performance of a compressor can be increased. 3D blading helps in reducing secondary flow losses and hence increased performance. Sweep and lean are examples of 3D blading, which is very common in any modern gas turbine compressor. A number of literatures are available in public domain, giving detailed understanding of effect of circumferential casing grooves and 3D blade features, but the interaction effect of sweep and casing grooves are not well published in public domain literature. In this work, an effort is made to understand, numerically, the interaction effect of sweep with circumferential grooves, using Computational Fluid Dynamics (CFD). Any numerical tool needs thorough validation before the results of numerical analyses can be used for analyzing the underlying physics. NASA Rotor37 is used to validate current CFD methodology. Mesh sensitivity is carried out to get mesh independence solution. Different turbulence models are used to get the best turbulence model for the problem in hand. 1D averaged performance data as well as hub to shroud variation of various flow parameters are compared to have full confidence on the CFD methodology. A baseline axial compressor rotor, without sweep and lean is generated, as the first step of this study. This rotor is created by using hub and tip profiles of NASA Rotor37. The profiles are stacked along a radial line through their center of gravities, which has resulted in rotor geometry without any sweep and lean. Modifications are done to the tip profile of the baseline rotor, in terms of stagger angle, to get comparable performance w.r.t. NASA Rotor37. Casing of the NASA Roto37 is used as the redesigned compressor casing. Circumferential casing grooves, with five grooves between leading edge to trailing edge, are created as per industry standards. Meshing and modeling are done according to the best practices developed while validating CFD methodology. It is to be noted that the casing grooves and the main flow domain are meshed with one to one mesh connectivity, in order to avoid any numerical losses due to interface interpolations. This is considered very critical in this work, as the vortices from the tip is expected to have a strong interaction with grooves. This interaction is expected to create high gradients of flow variables in this region. Valuable flow information might be lost, if flow variables are interpolated in this region. Baseline rotor is analyzed with and without casing grooves from choke to stall at 100% corrected speed. As expected, introduction of casing grooves has resulted in increased stall margin. A number of rotor geometries are created with different amount of sweeps. In the current study, blades are swept in the direction of chord, in order to avoid introduction of any sweep induced lean. The span location, where sweep starts, is also changed to understand the localized and global effect of this blade design features. Results obtained from numerical simulations of these geometries are presented in this paper. The performance and flow features are compared with respect to baseline rotor, with and without circumferential grooves, in an attempt to understand the underlying flow physics.

Effect of part sweep on axial flow compressor performance in the presence of circumferential casing grooves

“…Numeca AutoGrid5Òversion 9.0-2 is used to generate structured mesh along with matching groove mesh. Details about numerical methodology are presented in already published work by current authors [21]. To obtain meshindependent results from the numerical analyses, mesh sensitivity study is carried out with three different mesh sizes (1.5 million, 2.5 million and 3.0 million).…”

“…Minimum skewness obtained for the final mesh is approximately equal to 20 degrees. The final mesh used for comparison of numerical results to experimental data is shown in figure 1 [21]. Ansys CFXÒversion 15.0 is used to carry out CFD analysis (single passage, steady, RANS).…”

The Influence of Backward Sweep on Aerodynamic Performance of a 1.5-Stage Highly Loaded Axial Compressor

“…The results indicated that backward sweep rotor exhibited more stall warning time than zero sweep rotor. Goswami et al (2016) found that backward sweep increased stall margin and forward sweep increased the choked mass flow rate.…”