Prediction of Airfoil Stall Based on a Modified k−v2¯−ω Turbulence Model

The accuracy of an airfoil stall prediction heavily depends on the computation of the separated shear layer. Capturing the strong non-equilibrium turbulence in the shear layer is crucial for the accuracy of a stall prediction. In this paper, different Reynolds-averaged Navier–Stokes turbulence models are adopted and compared for airfoil stall prediction. The results show that the separated shear layer fixed k−v2¯−ω (abbreviated as SPF k−v2¯−ω) turbulence model captures the non-equilibrium turbulence in the separated shear layer well and gives satisfactory predictions of both thin-airfoil stall and trailing-edge stall. At small Reynolds numbers (Re~105), the relative error between the predicted CL,max of NACA64A010 by the SPF k−v2¯−ω model and the experimental data is less than 3.5%. At high Reynolds numbers (Re~106), the CL,max of NACA64A010 and NACA64A006 predicted by the SPF k−v2¯−ω model also has an error of less than 5.5% relative to the experimental data. The stall of the NACA0012 airfoil, which features trailing-edge stall, is also computed by the SPF k−v2¯−ω model. The SPF k−v2¯−ω model is also applied to a NACA0012 airfoil, which features trailing-edge stall and an error of CL relative to the experiment at CL>1.0 is smaller than 3.5%. The SPF k−v2¯−ω model shows higher accuracy than other turbulence models.

HEAT TRANSFER IN ROTATING, TRAILING EDGE, CONVERGING CHANNELS WITH FULL AND PARTIAL HEIGHT STRIP-FINS

A wide variety of pin-fins have been used to enhance heat transfer in internal cooling channels. However, due to their large blockage in the flow direction, they result in an undesirable high pressure drop. This experimental study aims to reduce pressure drop while increasing the heat transfer surface area by utilizing strip-fins in converging internal cooling channels. The channel is designed with a trapezoidal cross-section, converges in both transverse and longitudinal directions, and is also skewed β=120° with respect to the direction of rotation in order to model a trailing edge cooling channel. Only the leading and trailing surfaces of the channel are instrumented, and each surface is divided into eighteen isolated copper plates to measure the regionally averaged heat transfer coefficient. Utilizing pressure taps at the inlet and outlet of the channel, the pressure drop is obtained. Three staggered arrays of strip-fins are investigated: one full height configuration and two partial fin height arrangements (Sz=2mm and 1mm). In all cases, the strip fins are 2mm wide (W) and 10mm long (Lf ) in the flow direction. The fins are spaced such that Sy/Lf = 1 in the streamwise direction. However, due to the convergence, the spanwise spacing, Sx/W, was varied from 8 to 6.2 along the channel. The rotation number of the channel varied up to 0.21 by ranging the inlet Reynolds number from 10,000 to 40,000 and rotation speed from 0 to 300rpm. It is found that

On the frequency selection mechanism of the low-Re flow around rectangular cylinders

In the flow past elongated rectangular cylinders at moderate Reynolds numbers, vortices shedding from the leading- and trailing-edge corners are frequency locked by the impinging leading-edge vortex instability. The present work investigates how the chord-based Strouhal number varies with the aspect ratio of the cylinder at a Reynolds number (based on the cylinder thickness and the free-stream velocity) of $Re=400$ , i.e. when locking is strong. Several two-dimensional, nonlinear simulations are run for rectangular and D-shaped cylinders, with the aspect ratio ranging from $1$ to $11$ , and a global linear stability analysis of the flow is performed. The shedding frequency observed in the nonlinear simulations is predicted fairly well by the eigenfrequency of the leading eigenmode. The inspection of the structural sensitivity confirms the central role of the trailing-edge vortex shedding in the frequency locking, as already assumed by other authors. Surprisingly, however, the stepwise increase of the Strouhal number with the aspect ratio reported by several previous works is not fully reproduced. Indeed, with increasing aspect ratio, two distinct flow behaviours are observed, associated with two flow configurations where the interaction between the leading- and trailing-edge vortices is different. These two configurations are fully characterised, and the mechanism of selection of the flow configuration is discussed. Lastly, for aspect ratios close to the jump between two consecutive shedding modes, the Strouhal number is found to present hysteresis, implying the existence of multiple stable configurations. Continuing the lower shedding-mode branch by increasing the aspect ratio, we found that the periodic configuration loses stability via a Neimark–Sacker bifurcation leading to different Arnold tongues. This hysteresis can explain, at least partially, the significant scatter of existing experimental and numerical data.

LETR: An End-to-End Detector of Reconstruction Area in Blade’s Adaptive Machining with Transformer

In the leading/trailing edge’s adaptive machining of the near-net-shaped blade, a small portion of the theoretical part is retained for securing aerodynamic performance by manual work. However, this procedure is time-consuming and depends on the human experience. In this paper, we defined retained theoretical leading/trailing edge as the reconstruction area. To accelerate the reconstruction process, an anchor-free neural network model based on Transformer was proposed, named LETR (Leading/trailing Edge Transformer). LETR extracts image features from an aspect of mixed frequency and channel domain. We also integrated LETR with the newest meta-Acon activation function. We tested our model on the self-made dataset LDEG2021 on a single GPU and got an mAP of 91.9\%, which surpassed our baseline model, Deformable DETR by 1.1\%. Furthermore, we modified LETR’s convolution layer and named the new model after GLETR (Ghost Leading/trailing Edge Transformer) as a lightweight model for real-time detection. It is proved that GLETR has fewer weight parameters and converges faster than LETR with an acceptable decrease in mAP (0.1\%) by test results.