scholarly journals Transition Modeling for Low Pressure Turbines Using Computational Fluid Dynamics Driven Machine Learning

Energies ◽  
2021 ◽  
Vol 14 (15) ◽  
pp. 4680
Author(s):  
Harshal D. Akolekar ◽  
Fabian Waschkowski ◽  
Yaomin Zhao ◽  
Roberto Pacciani ◽  
Richard D. Sandberg
Keyword(s):  
Machine Learning ◽  
Shear Stress ◽  
Wall Shear Stress ◽  
Low Pressure ◽  
Stress Profile ◽  
Transition Model ◽  
Multi Objective Optimization ◽  
Multi Objective ◽  
Transition Prediction ◽  
Wall Shear

Existing Reynolds Averaged Navier–Stokes-based transition models do not accurately predict separation induced transition for low pressure turbines. Therefore, in this paper, a novel framework based on computational fluids dynamics (CFD) driven machine learning coupled with multi-expression and multi-objective optimization is explored to develop models which can improve the transition prediction for the T106A low pressure turbine at an isentropic exit Reynolds number of Re2is=100,000. Model formulations are proposed for the transfer and laminar eddy viscosity terms of the laminar kinetic energy transition model using seven non-dimensional pi groups. The multi-objective optimization approach makes use of cost functions based on the suction-side wall-shear stress and the pressure coefficient. A family of solutions is thus developed, whose performance is assessed using Pareto analysis and in terms of physical characteristics of separated-flow transition. Two models are found which bring the wall-shear stress profile in the separated region at least two times closer to the reference high-fidelity data than the baseline transition model. As these models are able to accurately predict the flow coming off the blade trailing edge, they are also able to significantly enhance the wake-mixing prediction over the baseline model. This is the first known study which makes use of `CFD-driven’ machine learning to enhance the transition prediction for a non-canonical flow.

scholarly journals Transition Modeling for Low Pressure Turbines Using Computational Fluid Dynamics Driven Machine Learning

2021 ◽  
Author(s):  
Harshal D. Akolekar ◽  
Fabian Waschkowski ◽  
Yaomin Zhao ◽  
Roberto Pacciani ◽  
Richard D. Sandberg
Keyword(s):  
Machine Learning ◽  
Shear Stress ◽  
Wall Shear Stress ◽  
Low Pressure ◽  
Stress Profile ◽  
Transition Model ◽  
Optimization Approach ◽  
Multi Objective Optimization ◽  
Multi Objective ◽  
Wall Shear

Existing Reynolds Averaged Navier-Stokes based transition models do not accurately predict separation induced transition for low pressure turbines. Therefore, in this study, a novel framework based on computational fluids dynamics driven machine learning coupled with multi-expression and multi-objective optimization is explored to develop models which can improve the transition prediction for the T106A low pressure turbine at an isentropic exit Reynolds number of Re2is=100,000. Model formulations are proposed for the transfer and laminar eddy viscosity terms of the laminar kinetic energy transition model using seven non-dimensional pi groups. The multi-objective optimization approach makes use of cost functions based on the suction-side wall-shear stress and the pressure coefficient. A family of solutions is thus developed, whose performance is assessed using Pareto analysis and in terms of physical characteristics of separated-flow transition. Two models are found which bring the wall-shear stress profile in the separated region at least two times closer to the reference high-fidelity data than the baseline transition model. As these models are able to accurately predict the flow coming off the blade trailing edge, they are also able to significantly enhance the wake-mixing prediction over the baseline model.

Generation and Development of Klebanoff Streaks in Low-Pressure Turbine Cascade Under Upstream Wakes

2020 ◽  
Author(s):  
Shuang Sun ◽  
Xingshuang Wu ◽  
Tianrong Tan ◽  
Canlin Zuo ◽  
Sirui Pan ◽  
...  
Keyword(s):  
Boundary Layer ◽  
Shear Stress ◽  
Wall Shear Stress ◽  
Separation Bubble ◽  
Transition Process ◽  
Low Pressure ◽  
Suction Surface ◽  
High Lift ◽  
Low Pressure Turbine ◽  
Wall Shear

Abstract At low Reynolds numbers operating condition, the boundary layer of the high-lift low-pressure turbine (LPT) of aero-engines is prone to separate on the suction surface of the airfoil. The profile losses of the airfoil are largely governed by the size of the separation bubble and the transition process in the boundary layer. However, the wake-induced transition, the natural transition and the instability induced by the Klebanoff streaks complicate the transition process. The boundary layer on the suction surface of a high-lift LPT was investigated at Re = 50,000 with upstream wakes. The numerical simulation was performed with the CFX software using large eddy simulations (LES), and the experiment was performed on a linear cascade. In this study, the wake is divided into the wake center and the wake tail, the unsteady formation process of the streaks and the wall shear stress caused by the wake are analyzed. A new mechanism of generation and development of Klebanoff Streaks was presented to better understand the effect of the wake on the boundary layer. Moreover, it was found that after entering the blade passage, the wake center does not contact the blade but causes the wall shear stress of the front part on the suction surface to increase. However, it is not possible to form strong Klebanoff streaks at the leading edge of the blade by shear sheltering effect. Only the wake tail can form Klebanoff streaks when it contacts the blade.

Microscopic Velocimetry With a Scaled-Up Model for Evaluating a Flow Field Over Cultured Endothelial Cells

2002 ◽  
Vol 124 (2) ◽  
pp. 176-179 ◽  
Author(s):  
Shuichiro Fukushima ◽  
Takaaki Deguchi ◽  
Makoto Kaibara ◽  
Kotaro Oka ◽  
Kazuo Tanishita
Keyword(s):  
Endothelial Cells ◽  
Shear Stress ◽  
Wall Shear Stress ◽  
Flow Field ◽  
Cell Model ◽  
Stress Profile ◽  
Measured Velocity ◽  
Cultured Endothelial Cells ◽  
Sinusoidal Cell ◽  
Wall Shear

A microscopic velocimetry technique for evaluating the flow field over cultured endothelial cells was developed. Flow around a cell model scaled up by a factor of 100 was visualized by using an optical microscope and was quantified by using particle-tracking velocimetry. Wall shear stress on the model surface was determined from a two-dimensional velocity field interpolated from measured velocity vectors. Accuracy of the velocimetry was verified by measuring the flow over a sinusoidal cell model that had a wall shear stress profile analytically determined with linear perturbation theory. Comparison of the experimental results with the analytical solution revealed that the total error of the measured wall shear stress was 6 percent.

Cerebral aneurysm rupture status classification using statistical and machine learning methods

2021 ◽  
pp. 095441192110004
Author(s):  
Nicolás Amigo ◽  
Alvaro Valencia ◽  
Wei Wu ◽  
Sourav Patnaik ◽  
Ender Finol
Keyword(s):  
Machine Learning ◽  
Shear Stress ◽  
Random Forest ◽  
Wall Shear Stress ◽  
Confusion Matrix ◽  
Cerebral Aneurysms ◽  
Machine Learning Algorithms ◽  
Size Ratio ◽  
Patient Specific ◽  
Wall Shear

Morphological characterization and fluid dynamics simulations were carried out to classify the rupture status of 71 (36 unruptured, 35 ruptured) patient specific cerebral aneurysms using a machine learning approach together with statistical techniques. Eleven morphological and six hemodynamic parameters were evaluated individually and collectively for significance as rupture status predictors. The performance of each parameter was inspected using hypothesis testing, accuracy, confusion matrix, and the area under the receiver operating characteristic curve. Overall, the size ratio exhibited the best performance, followed by the diastolic wall shear stress, and systolic wall shear stress. The prediction capability of all 17 parameters together was evaluated using eight different machine learning algorithms. The logistic regression achieved the highest accuracy (0.75), whereas the random forest had the highest area under curve value among all the classifiers (0.82), surpassing the performance exhibited by the size ratio. Hence, we propose the random forest model as a tool that can help improve the rupture status prediction of cerebral aneurysms.

Laminar-Turbulent Transition Flows of Non-Newtonian Slurries: Models Assessment

2018 ◽  
Vol 141 (1) ◽  
Author(s):  
Kofi Freeman K. Adane ◽  
Martin Agelin-Chaab
Keyword(s):  
Shear Stress ◽  
Wall Shear Stress ◽  
Yield Stress ◽  
Stress Ratio ◽  
Rheological Parameters ◽  
Transition Model ◽  
Flow Conditions ◽  
Wide Range ◽  
Engineering Models ◽  
Wall Shear

In this study, a qualitative assessment of transitional velocity engineering models for predicting non-Newtonian slurry flows in a horizontal pipe was performed using data from a wide range of pipe diameters (25–268 mm). In addition, the gamma theta transition model was used to compute selected flow conditions. These models were used to predict transitional velocities in large pipe diameters (up to 420 mm) for slurries. In general, it was observed that most of the current engineering models predict transitional velocities conservatively. Based on the gamma theta transition model results, for large Hedström numbers (He ≳ 105), other methods should be used to predict transitional velocities if a change in the pipe diameter (scale-up) results in an order of magnitude increase in the He value. It was also found that the gamma theta transition model predicted a laminar flow condition in the fully developed region for flow conditions with a small plug region (low-yield stress-to-wall shear stress ratio), which is contrary to what has been observed in some experiments. This is attributed to the local fluid rheological parameters values, which might be different from those reported. However, the gamma theta transition model results are in good agreement with the experimental data for flow conditions that have a large plug region (high-yield stress-to-wall shear stress ratio).

Sign in / Sign up

Export Citation Format

Share Document