scholarly journals Turbine Design and Optimization for a Supercritical CO2 Cycle Using a Multifaceted Approach Based on Deep Neural Network

Energies ◽  
2021 ◽  
Vol 14 (22) ◽  
pp. 7807
Author(s):  
Muhammad Saeed ◽  
Abdallah S. Berrouk ◽  
Burhani M. Burhani ◽  
Ahmed M. Alatyar ◽  
Yasser F. Al Wahedi
Keyword(s):  
Neural Network ◽  
Deep Neural Network ◽  
Design Parameters ◽  
Speed Ratio ◽  
Inlet Flow ◽  
Flow Angle ◽  
Multifaceted Approach ◽  
Design And Optimization ◽  
Radial Turbine ◽  
Turbine Design

Turbine as a key power unit is vital to the novel supercritical carbon dioxide cycle (sCO2-BC). At the same time, the turbine design and optimization process for the sCO2-BC is complicated, and its relevant investigations are still absent in the literature due to the behavior of supercritical fluid in the vicinity of the critical point. In this regard, the current study entails a multifaceted approach for designing and optimizing a radial turbine system for an 8 MW sCO2 power cycle. Initially, a base design of the turbine is calculated utilizing an in-house radial turbine design and analysis code (RTDC), where sharp variations in the properties of CO2 are implemented by coupling the code with NIST’s Refprop. Later, 600 variants of the base geometry of the turbine are constructed by changing the selected turbine design geometric parameters, i.e., shroud ratio (rs4r3), hub ratio (rs4r3), speed ratio (νs) and inlet flow angle (α3) and are investigated numerically through 3D-RANS simulations. The generated CFD data is then used to train a deep neural network (DNN). Finally, the trained DNN model is employed as a fitting function in the multi-objective genetic algorithm (MOGA) to explore the optimized design parameters for the turbine’s rotor geometry. Moreover, the off-design performance of the optimized turbine geometry is computed and reported in the current study. Results suggest that the employed multifaceted approach reduces computational time and resources significantly and is required to completely understand the effects of various turbine design parameters on its performance and sizing. It is found that sCO2-turbine performance parameters are most sensitive to the design parameter speed ratio (νs), followed by inlet flow angle (α3), and are least receptive to shroud ratio (rs4r3). The proposed turbine design methodology based on the machine learning algorithm is effective and substantially reduces the computational cost of the design and optimization phase and can be beneficial to achieve realistic and efficient design to the turbine for sCO2-BC.

An Integrated Software Procedure to Support Teaching of Radial Inflow Turbine Design

2007 ◽  
Author(s):  
Carlo Cravero ◽  
Martino Marini
Keyword(s):  
Design Analysis ◽  
Design Parameters ◽  
Analysis Tool ◽  
Software Suite ◽  
Computational Tools ◽  
Blade Loading ◽  
Radial Turbine ◽  
Integrated Software ◽  
Turbine Design ◽  
Turbomachinery Design

The authors decided to organize their design/analysis computational tools in an integrated software suite in order to help teaching radial turbine, taking advantage of their research background and a set of codes previously developed. The software is proposed for use during class works and the student can either use a single design/analysis tool or face a complete design loop consisting of iterations between design and analysis tools. The intended users are final year students in mechanical engineering. The codes output are discussed with two practical examples in order to highlight the turbomachinery performance at design and off-design conditions. The above suite gives the student the opportunity of getting used to different concepts (choking, blade loading, performance maps, …) that are encountered in turbomachinery design and of understanding the effects of the main design parameters.

Axial Turbine Design for a Twin-Turbine Heavy-Duty Turbocharger Concept

2018 ◽  
Author(s):  
Nicholas Anton ◽  
Magnus Genrup ◽  
Carl Fredriksson ◽  
Per-Inge Larsson ◽  
Anders Christiansen-Erlandsson
Keyword(s):  
Engine Performance ◽  
Operating Conditions ◽  
Single Stage ◽  
Flow Coefficient ◽  
Speed Ratio ◽  
Heavy Duty ◽  
Turbine Stage ◽  
Axial Turbine ◽  
Radial Turbine ◽  
Turbine Design

In the process of evaluating a parallel twin-turbine pulse-turbocharged concept, the results considering the turbine operation clearly pointed towards an axial type of turbine. The radial turbine design first analyzed was seen to suffer from sub-optimum values of flow coefficient, stage loading and blade-speed-ratio. Modifying the radial turbine by both assessing the influence of “trim” and inlet tip diameter all concluded that this type of turbine is limited for the concept. Mainly, the turbine stage was experiencing high values of flow coefficient, requiring a more high flowing type of turbine. Therefore, an axial turbine stage could be feasible as this type of turbine can handle significantly higher flow rates very efficiently. Also, the design spectrum is broader as the shape of the turbine blades is not restricted by a radially fibred geometry as in the radial turbine case. In this paper, a single stage axial turbine design is presented. As most turbocharger concepts for automotive and heavy-duty applications are dominated by radial turbines, the axial turbine is an interesting option to be evaluated for pulse-charged concepts. Values of crank-angle-resolved turbine and flow parameters from engine simulations are used as input to the design and subsequent analysis. The data provides a valuable insight into the fluctuating turbine operating conditions and is a necessity for matching a pulse-turbocharged system. Starting on a 1D-basis, the design process is followed through, resulting in a fully defined 3D-geometry. The 3D-design is evaluated both with respect to FEA and CFD as to confirm high performance and durability. Turbine maps were used as input to the engine simulation in order to assess this design with respect to “on-engine” conditions and to engine performance. The axial design shows clear advantages with regards to turbine parameters, efficiency and tip speed levels compared to a reference radial design. Improvement in turbine efficiency enhanced the engine performance significantly. The study concludes that the proposed single stage axial turbine stage design is viable for a pulse-turbocharged six-cylinder heavy-duty engine. Taking into account both turbine performance and durability aspects, validation in engine simulations, a highly efficient engine with a practical and realizable turbocharger concept resulted.

On the Design and Matching of Turbocharger Single Scroll Turbines for Pass Car Gasoline Engines

2014 ◽  
Vol 136 (12) ◽  
Author(s):  
Marc Gugau ◽  
Harald Roclawski
Keyword(s):  
Fuel Consumption ◽  
Turbine Blade ◽  
Design Parameters ◽  
Speed Ratio ◽  
Element Analysis ◽  
Gasoline Engines ◽  
Turbine Efficiency ◽  
Component Design ◽  
Almost All ◽  
Turbine Design

With emission legislation becoming more stringent within the next years, almost all future internal combustion gasoline engines need to reduce specific fuel consumption, most of them by using turbochargers. Additionally, car manufactures attach high importance to a good drivability, which usually is being quantified as a target torque already available at low engine speeds—reached in transient response operation as fast as possible. These engine requirements result in a challenging turbocharger compressor and turbine design task, since for both not one single operating point needs to be aerodynamically optimized but the components have to provide for the optimum overall compromise for maximum thermodynamic performance. The component design targets are closely related and actually controlled by the matching procedure that fits turbine and compressor to the engine. Inaccuracies in matching a turbine to the engine full load are largely due to the pulsating engine flow characteristic and arise from the necessity of arbitrary turbine map extrapolation toward low turbine blade speed ratios and the deficient estimation of turbine efficiency for low engine speed operating points. This paper addresses the above described standard problems, presenting a methodology that covers almost all aspects of thermodynamic turbine design based on a comparison of radial and mixed-flow turbines. Wheel geometry definition with respect to contrary design objectives is done using computational fluid dynamics (CFD), finite element analysis (FEA), and optimization software. Parametrical turbine models, composed of wheel, volute, and standard piping allow for fast map calculation similar to steady hot gas tests but covering the complete range of engine pulsating mass flow. These extended turbine maps are then used for a particular assessment of turbine power output under unsteady flow admission resulting in an improved steady-state matching quality. Additionally, the effect of various design parameters like either volute sizing or the choice of compressor to turbine diameter ratio on turbine blade speed ratio operating range as well as well as turbine inertia effect is analyzed. Finally, this method enables the designer to comparatively evaluate the ability of a turbine design to accelerate the turbocharger speed for transient engine response while still offering a map characteristic that keeps fuel consumption low at all engine speeds.

scholarly journals Design and Optimization of a Low-Resonant-Frequency Piezoelectric MEMS Energy Harvester Based on Artificial Intelligence

Proceedings ◽  
2018 ◽  
Vol 2 (13) ◽  
pp. 930 ◽  
Author(s):  
Seyedfakhreddin Nabavi ◽  
Lihong Zhang
Keyword(s):  
Neural Network ◽  
Artificial Intelligence ◽  
Resonant Frequency ◽  
Deep Neural Network ◽  
Energy Harvester ◽  
Vibration Energy Harvester ◽  
Design And Optimization ◽  
Speed Up ◽  
Operational Frequency ◽  
Piezoelectric Mems

In this study we propose a piezoelectric MEMS vibration energy harvester with the capability of oscillating at low (i.e., less than 200 Hz) resonant frequency. The mechanical structure of the proposed harvester is comprised of a doubly clamped cantilever with a serpentine pattern associated with several discrete masses. In order to obtain the optimal physical aspects of the harvester and speed up the design process, we have utilized a deep neural network, as an artificial intelligence (AI) method. Firstly, the deep neural network was trained with 108 data samples gained from finite element modeling (FEM). Then this trained network was integrated with the genetic algorithm (GA) to optimize geometry of the harvester to enhance its performance in terms of resonant frequency and generated voltage. Our numerical results confirm that the accuracy of the network in prediction is above 90%. Consequently, by taking advantage of this efficient AI-based performance estimator, the GA is able to reduce the device operational frequency from 169 Hz to 110.5 Hz and increase its efficiency on harvested voltage from 2.5 V to 3.4 V under 0.25 g excitation.

Prediction of Non-Linear Mechanical Behavior With Deep Neural Network: Application on Low Pressure Turbine Disc

2020 ◽  
Author(s):  
Yuan Jin ◽  
Weichen Li ◽  
Zheyi Yang ◽  
Olivier Jung
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Displacement Field ◽  
Deep Neural Network ◽  
Low Pressure ◽  
Constitutive Law ◽  
Design Parameters ◽  
Low Pressure Turbine ◽  
Turbine Disc ◽  
Nonlinear Constitutive Law

Abstract Thanks to the increase of computational capacity and the diversification of computational means, deep learning techniques have shown great successes in learning representations from data in the past decade. Following this trend, efforts have been made in the literature to apply Deep Neural Network (DNN) as surrogate model. Common practice consists in utilizing a single DNN to predict a certain physical property given input design parameters, and the DNN is trained by corresponding simulation results. However, most of the complex high-fidelity simulations involve nonlinear physical laws, e.g. elasto-plasticity, which cannot be explicitly depicted by the applied single DNN model. In the present work, static mechanical simulation with nonlinear constitutive law is addressed with a novel approach in a deep learning framework. We approximate the displacement and the nonlinear constitutive law by two deep neural networks. The first DNN acts as a prior on the unknown displacement field, while the second network aims at describing the nonlinear strain-stress relationship. The dependence of the strainstress relationship on the strain level is taken into consideration by taking the first order derivative with respect to spatial coordinates of the first DNN as an input of the second network. A new loss model combining the error in displacement field prediction and constitutive law description is proposed to train the two DNNs together. We demonstrate the effectiveness of the proposed framework on a low pressure turbine disc design problem.

Design and optimization of an electromechanical actuator for the latch of a foldable vehicle seat

2020 ◽  
Vol 62 (7) ◽  
pp. 749-755
Author(s):  
Z. K. Kocabicak ◽  
U. Demir
Keyword(s):  
Neural Network ◽  
Design Parameters ◽  
Design Criteria ◽  
Model Parameters ◽  
The Finite Element Method ◽  
Electromechanical Actuator ◽  
Design And Optimization ◽  
Artificial Neural Network Ann ◽  
Critical Model ◽  
Vehicle Seat

Abstract This paper deals with the electromechanical actuator (EAct) design for a seat latch while maintaining required force and displacement according to the boundary conditions and design criteria for the finite element method (FEM) in an Ansys Maxwell environment. Before presenting the analysis studies, some EAct models are parameterized according to the Taguchi’s design of experiment (DoE) method. After that, analysis results are evaluated to define the critical model parameters of the EAct according to the DoE method. Furthermore, the DoE results and design parameters of the EAct are trained in some cases by an artificial neural network (ANN). The dynamic behavior of the models from the ANN and DoE results are analyzed and the results obtained are compared. Finally, the optimal EAct model is defined taking into account design criteria.

scholarly journals Design and Optimization of a Deep Neural Network Architecture for Traffic Light Detection

2020 ◽  
Vol 77 ◽  
pp. 01002
Author(s):  
Tomohide Fukuchi ◽  
Mark Ogbodo Ikechukwu ◽  
Abderazek Ben Abdallah
Keyword(s):  
Neural Network ◽  
Network Architecture ◽  
Deep Neural Network ◽  
Autonomous Driving ◽  
Research Trend ◽  
Traffic Light ◽  
Light Detection ◽  
Driving System ◽  
Design And Optimization ◽  
Autonomous Driving System

Autonomous Driving has recently become a research trend and efficient autonomous driving system is difficult to achieve due to safety concerns, Applying traffic light recognition to autonomous driving system is one of the factors to prevent accidents that occur as a result of traffic light violation. To realize safe autonomous driving system, we propose in this work a design and optimization of a traffic light detection system based on deep neural network. We designed a lightweight convolution neural network with parameters less than 10000 and implemented in software. We achieved 98.3% inference accuracy with 2.5 fps response time. Also we optimized the input image pixel values with normalization and optimized convolution layer with pipeline on FPGA with 5% resource consumption.

Design of a Controllable One-Meter Scale Research Wind Turbine

2017 ◽  
Author(s):  
Samuel Cole ◽  
Gavin Hess ◽  
Martin Wosnik
Keyword(s):  
Reynolds Number ◽  
Wind Turbine ◽  
Reynolds Numbers ◽  
Low Flow ◽  
Design Parameters ◽  
Speed Ratio ◽  
Flow Physics ◽  
Manual Adjustment ◽  
High Reynolds ◽  
Turbine Design

A research wind turbine of one meter diameter was designed for the UNH Flow Physics Facility (FPF), a very large flow physics quality turbulent boundary layer wind tunnel (W 6m, H 2.7m, L 72m), which provides excellent spatial and temporal resolution, low flow blockage and allows measurements of turbine wakes far downstream due its long fetch. The initial turbine design was carried out as an aero-servo model of the NREL 5MW reference turbine, with subsequent modifications to both the hub to accommodate blade mounting and pitch-adjustment, and increases in model blade chord to achieve sufficiently high Reynolds numbers. A trade-off study of turbine design parameters in scale space was conducted. Several candidate airfoil profiles were evaluated numerically with the goal to reach Reynolds-number independence in turbine performance in the target operating range. The model turbine will achieve Reynolds numbers based on blade chord, an important consideration for airfoil performance and near-wake evolution, greater than 100,000, and Reynolds numbers based on turbine diameter, important for far-wake transport, on the order of 1,000,000. The blockage ratio is less than 5% based on swept area. A motor and controller combination was implemented that allows to precisely prescribe the turbine tip-speed ratio (at maximum power coefficient for optimum blade chord), which can remain stable and absorb the generated electric power for long periods of time. The turbine nacelle was designed with a blade mounting mechanism which allows for precise manual adjustment of blade pitch angle, while allowing for future implementation of actuated pitch control. The O(1m) turbine scale is viewed as a cost-effective compromise between size, driven by the need for sufficiently high Reynolds number, and the need for detailed measurements for significant distances downstream of the turbine under controlled conditions.

Evaluation and Optimization of Trimaran Configurations Using Deep Neural Network

2019 ◽  
Author(s):  
Dongchi Yu ◽  
Lu Wang ◽  
Qian Zhong ◽  
Ronald W. Yeung
Keyword(s):  
Neural Network ◽  
Deep Neural Network ◽  
Transportation Cost ◽  
Computational Method ◽  
Design Parameters ◽  
Minimal Amount ◽  
Computational Results ◽  
Typical Situation ◽  
Water Transportation ◽  
Calm Water

Abstract To determine the optimal trimaran configuration for best calm-water transportation efficiency, a Deep Neural Network (DNN) is trained with sufficient computational results provided by, as an example, an in-house developed potential-flow code called Multi-hull Simple-source Panel Method (MSPM). Even though the computational method is extremely efficient in accurately establishing the mapping relation between the key design parameters governing the trimaran configuration problem and the resulting calm-water transportation cost, the modeling efforts are non-trivial since the number of geometric and configuration parameters in a typical situation is large. In this work, we demonstrate how the “Big Data” of computational results can be effectively utilized in training a DNN. An optimal trimaran configuration solution within a specified design space, subject to realistic range constraints, can be quickly determined in a minimal amount of time with the DNN. A demonstrative case study is provided for illustration.

scholarly journals Numerical Simulation of the Performance of a Twin Scroll Radial Turbine at Different Operating Conditions

2019 ◽  
Vol 2019 ◽  
pp. 1-13 ◽  
Author(s):  
Carlo Cravero ◽  
Davide De Domenico ◽  
Andrea Ottonello
Keyword(s):  
Operating Conditions ◽  
Inlet Section ◽  
Flow Distortion ◽  
Experimental Database ◽  
Design And Optimization ◽  
Radial Turbine ◽  
Ansys Cfx ◽  
Wide Range ◽  
Partial Admission ◽  
Turbine Design

Twin scroll radial turbines are increasingly used for turbocharging applications, to take advantage of the pulsating exhaust gases. In spite of its relevance in turbocharging techniques, scientific literature about CFD applied to twin scroll turbines is limited, especially in case of partial admission. In the present paper a CFD complete model of a twin scroll radial turbine is developed in order to give a contribution to literature in understanding the capabilities of current industrial CFD approaches applied to these difficult cases and to develop performance index that can be used for turbine design optimization purposes. The flow solution is obtained by means of ANSYS CFX ® in a wide range of operating conditions in full and partial admission cases. The total-to-static efficiency and the mass flow parameter (MFP) have been calculated and compared with the experimental database in order to validate the numerical model. The purpose of the developed procedure is also to generate a database for twin scroll turbines useful for future applications. A comparison between performances obtained in different admission conditions was performed. In particular the analysis focused on the characterization of the flow at volute outlet/rotor inlet section. A flow distortion index at rotor inlet was introduced to correlate the turbine performance and the flow nonuniformities generated by the volute. Finally the influence of the backside cavity on the performance parameters is also discussed. The introduction of these new nonuniformity indices is proposed for volute design and optimization procedures.

Sign in / Sign up

Export Citation Format

Share Document