High-fidelity Methods Research Articles

HPC-enabling technologies for high-fidelity combustion simulations

With the increase in computational power in the last decade and the forthcoming Exascale supercomputers, a new horizon in computational modelling and simulation is envisioned in combustion science. Considering the multiscale and multiphysics characteristics of turbulent reacting flows, combustion simulations are considered as one of the most computationally demanding applications running on cutting-edge supercomputers. Exascale computing opens new frontiers for the simulation of combustion systems as more realistic conditions can be achieved with high-fidelity methods. However, an efficient use of these computing architectures requires methodologies that can exploit all levels of parallelism. The efficient utilization of the next generation of supercomputers needs to be considered from a global perspective, that is, involving physical modelling and numerical methods with methodologies based on High-Performance Computing (HPC) and hardware architectures. This review introduces recent developments in numerical methods for large-eddy simulations (LES) and direct-numerical simulations (DNS) to simulate combustion systems, with focus on the computational performance and algorithmic capabilities. Due to the broad scope, a first section is devoted to describe the fundamentals of turbulent combustion, which is followed by a general description of state-of-the-art computational strategies for solving these problems. These applications require advanced HPC approaches to exploit modern supercomputers, which is addressed in the third section. The increasing complexity of new computing architectures, with tightly coupled CPUs and GPUs, as well as high levels of parallelism, requires new parallel models and algorithms exposing the required level of concurrency. Advances in terms of dynamic load balancing, vectorization, GPU acceleration and mesh adaptation have permitted to achieve highly-efficient combustion simulations with data-driven methods in HPC environments. Therefore, dedicated sections covering the use of high-order methods for reacting flows, integration of detailed chemistry and two-phase flows are addressed. Final remarks and directions of future work are given at the end.

Effectiveness of high- and low-fidelity simulation-based medical education in teaching cardiac auscultation: a systematic review and meta-analysis

Simulation-based medical education (SBME) is an evolving method of teaching cardiac examination skills to healthcare learners. It has been deliberated how effective this teaching modality is and whether high-fidelity methods are more effective than low-fidelity methods. This systematic review aimed to assess the effectiveness of high-fidelity SBME in teaching cardiac auscultation compared with no intervention or another active teaching intervention (low-fidelity SBME) using evidence from randomized controlled trials (RCTs).Literature searches were performed on Medline, Embase, PsychInfo and Cinahl. RCTs that compared the effectiveness of high-fidelity simulation against no intervention or high-fidelity simulation against low-fidelity simulation in teaching cardiac auscultation to healthcare learners were included. Outcomes were knowledge, skills and satisfaction relating to cardiac auscultation education. Data were analyzed using Review Manager 5.3 software.Seventeen RCTs (n = 1055) were included. Twelve RCTs (n = 692) compared high-fidelity simulation with no intervention. The pooled effect sizes for knowledge and skills were 1.39 (95% confidence interval [CI], 0.39–2.38;This review’s findings suggest that high-fidelity SBME is an effective teaching method for cardiac auscultation education. Interestingly, there was no significant difference in knowledge or skills among learners when comparing high-fidelity simulation with low-fidelity simulation. Further research is needed to establish the effectiveness of different forms of SBME as educational interventions.

Fast Predictions of Aircraft Aerodynamics Using Deep-Learning Techniques

The numerical analysis of aerodynamic components based on the Reynolds–averaged Navier–Stokes equations has become critical for the design of transport aircraft but still entails large computational cost. Simulating a multitude of different flow conditions with high-fidelity methods as required for loads analysis or aerodynamic shape optimization is still prohibitive. Within the past few years, the application of machine learning methods has been proposed as a potential way to overcome these shortcomings. This is leading toward a new data-driven paradigm for the modeling of physical problems. The objective of this paper is the development of a deep-learning methodology for the prediction of aircraft surface pressure distributions and the rigorous comparison with existing state-of-the-art nonintrusive reduced-order models. Bayesian optimization techniques are employed to efficiently determine optimal hyperparameters for all deep neural networks. Three data-driven methods which are Gaussian processes, proper orthogonal decomposition combined with an interpolation technique, and deep learning are investigated. The results are compared for a two-dimensional airfoil case and the NASA Common Research Model transport aircraft as a relevant three-dimensional case. Results show that all methods are able to properly predict the surface pressure distribution at subsonic conditions. In transonic flow, when shock waves and separation lead to nonlinearities, deep-learning methods outperform the others by also capturing the shock wave location and strength accurately.

Furthering knowledge on the flow pattern around high-rise buildings: LES investigation of the wind energy potential

To understand the favourable locations for maximum yield of urban wind energy, understanding the flow pattern on the roof region of high-rise buildings is crucial. However, very few studies in the literature report on the wind energy resource using high-fidelity methods, which combined with the lack of wind tunnel investigations, explains the failure of urban wind energy as a viable contributor to the renewable energy mix.This work aims at proposing a strategy for the reliable evaluation of the wind energy resource by building on validated Large Eddy Simulation and performance maps based on the wind power density and the turbulence intensity above the roof of a prismatic 1:3 square building. Analyses include the effect of flat and decked roof shapes in two wind directions. Results around the facade reveal zones of accelerated flow close to separation edges, while above the roof, a higher potential is noticeable at 45°, than at 0°. Decking of the roof shows insensitivity to the change in wind direction. The tested methodology can be extended to other configurations to predict the wind energy potential as well as used to indicate the positioning of wind turbines in the built environment.

Multi-Fidelity Combustor Design and Experimental Test for a Micro Gas Turbine System

A multi-fidelity micro combustor design approach is developed for a small-scale combined heat and power CHP system. The approach is characterised by the coupling of the developed preliminary design model using the combined method of 3D high-fidelity modelling and experimental testing. The integrated multi-physics schemes and their underlying interactions are initially provided. During the preliminary design phase, the rapid design exploration is achieved by the coupled reduced-order models, where the details of the combustion chamber layout, flow distributions, and burner geometry are defined as well as basic combustor performance. The high-fidelity modelling approach is then followed to provide insights into detailed flow and emission physics, which explores the effect of design parameters and optimises the design. The combustor is then fabricated and assembled in the MGT test bench. The experimental test is performed and indicates that the designed combustor is successfully implemented in the MGT system. The multi-physics models are then verified and validated against the test data. The details of refinement on lower-order models are given based on the insights acquired by high-fidelity methods. The shortage of conventional fossil fuels and the continued demand for energy supplies have led to the development of a micro-turbine system running renewable fuels. Numerical analysis is then carried out to assess the potential operation of biogas in terms of emission and performance. It produces less NOx emission but presents a flame stabilisation design challenge at lower methane content. The details of the strategy to address the flame stabilisation are also provided.

Objective and Subjective Stress Parameters in Response to High and Low-Fidelity Simulation Activities.

Nursing graduates are required to have both excellent theoretical and practical skills that should be used during stressful emergency interventions. Since the received knowledge should be practiced to gain skills and trained to achieve competences, simulation exercises can be beneficial to even reduce the stress that each individual may face during emergency management of patients. A total of 146 first-year nursing students participated in the study, including 124 women and 22 men aged between 19 and 50 years, with a mean age of 32 years. The objective method estimated psychophysiological parameters (serum cortisol). Objective and subjective methods were used. The subjective method assessed stress experienced by students based on the standardized Stress Appraisal Questionnaire Version B for dispositional assessment. The study was conducted in the Monoprofile Medical Simulation Centre at the University of Economics and Innovation in Lublin, Poland and was approved by the University Research Ethics Committee. Both participants under and over 25 years of age showed increased levels of stress after low and high-fidelity simulations, with statistically significantly higher stress levels found for the low fidelity method. Low-fidelity simulation methods generated a greater increase in cortisol levels, indicating a higher stress level than the high-fidelity methods. The analysis of the scores obtained in the Stress Appraisal Questionnaire (KOS-B) showed that higher cortisol levels after the low-fidelity simulation reduced the subjective perception of a threat, while higher cortisol levels before the high-fidelity simulation promoted higher intellectual activity among the students. Levels of stress in the education of nursing students using low and high-fidelity methods can limit the sense of threat and activate professional task performance. The use of low and high-fidelity simulation does not generate destructive stress levels.

Machine Learning for the Development of Data-Driven Turbulence Closures in Coolant Systems

Abstract This work shows the application of Gene Expression Programming to augment RANS turbulence closures for flows though complex geometries. Specifically, an optimized internal cooling channel of a turbine blade, designed for additive manufacturing. One of the challenges in internal cooling design is the heat transfer accuracy of the RANS formulation in comparison to higher fidelity methods, which are still not used in design on account of their computational cost. However, high fidelity data can be extremely valuable for improving current lower fidelity models and this work shows the application of data-driven approaches to develop turbulence closures for an internally ribbed duct. Different approaches are compared and the results of the improved model are illustrated; first on the same geometry, and then for an unseen predictive case. The work shows the potential of using data-driven models for accurate heat transfer predictions even in non-conventional configurations and indicates the ability of closures learnt from complex flow cases to adapt successfully to unseen test cases.

Toward Vehicle-Level Optimization of Compound Rotorcraft Aerodynamics

This paper presents high-fidelity and simplified computational fluid dynamics modeling approaches within an optimization framework for compound rotorcraft configurations with rotor/propeller aerodynamic interactions. Simulations of an axial-flight ducted propeller and a rotor/wing test case are first presented to validate the in-house Helicopter Multi-Block 3 solver and to assess the actuator disk/line models for the simplified main rotor modeling. The actuator disk/line models are then used to represent the main rotor for simulations of a generalized rotor/propeller combination. The propeller performance is analyzed in detail, and large variations are observed in the single blade loading due to the main rotor wake. A simplified model for the rotor/propeller interaction simulation is also put forward, and an inflow distortion metric is proposed to quantify the aerodynamic interactions. With the help of a kriging surrogate model and the inflow distortion metric, aerodynamic interferences through the propeller disk are quantitatively visualized with variations in the propeller position, propeller thrust, and main rotor advance ratio. Optimization of the propeller position under the main rotor for minimized interference with rolling/pitching moment constraints is also attempted using both gradient-based (adjoint) and gradient-free (efficient global optimization) approaches. The optimization results are verified using blade resolved simulations, and fluctuations of the propeller single blade loading were effectively reduced due to the optimization. The work is a first step toward high-fidelity methods for vehicle and configuration optimization.

The Effect of Using Different Wake Models on Wind Farm Layout Optimization: A Comparative Study

Abstract The accuracy of analytical wake models applied in wind farm layout optimization (WFLO) problems is of great significance as the high-fidelity methods such as large eddy simulation (LES) and Reynolds-averaged Navier–Stokes (RANS) are still not able to handle an optimization problem for large wind farms. Based on a variety of analytical wake models developed in the past decades, Flow Redirection and Induction in Steady State (FLORIS) have been published as a tool that integrated several widely used wake models and their expansions. This paper compares four wake models selected from FLORIS by applying three classical WFLO scenarios. The results illustrate that the Jensen wake model is the fastest, but the issue of underestimating the velocity deficit is obvious. The multi-zone model needs additional tuning on the parameters inside the model to fit specific wind turbines. The Gaussian-curl hybrid (GCH) wake model, as an advanced expansion of the Gaussian wake model, does not provide a significant improvement in the current study, where the yaw control is not included. The Gaussian wake model is recommended for the WFLO projects implemented under the FLORIS framework and has similar wind conditions with the present work.

A new simplified numerical approach for shape design of underwater wings

ABSTRACT This paper presents a new simplified numerical approach (SNA) for predicting hydrodynamic data of underwater wings (UWs). Instead of full three-dimensional (3D) flow field analyses, SNA combines Vortex Lattice Method (VLM) and several two-dimensional Reynolds-Average Navier-Stokes (2D RANS) simulations of wing sections to gain results. Hence, SNA shows a remarkably lower computational cost than higher-fidelity methods such as 3D RANS. Moreover, SNA shows good agreements with 3D RANS and experimental data. Thereafter, SNA is applied to shape optimizations of UWs for efficiency and accuracy. Two kinds of optimisation algorithms that include local and global finders are adopted in the optimiser. Searching behaviours of the two algorithms indicate that the problem is not proper to be solved by the local optimiser for its multimodality, while the global optimiser is more efficient. Furthermore, the final optimised wing has few wing-tip vortices and a smooth pressure distribution.

Challenges and opportunities for artificial intelligence and high-fidelity simulations in turbomachinery applications: A perspective

Turbomachines are still required for the efficient conversion of renewable, chemical or potential energy into propulsion, mechanical power, or electricity. The increasing demand for efficiency, availability, reduced footprint and cost of ownership poses fundamental challenges to design methods the accuracy of which needs to be constantly upgraded to keep the pace with the availability of new materials, type of fluids and fuels, manufacturing methods and technologies. This paper discusses recent trends in design methods that take advantage of both artificial intelligence and high-fidelity simulations techniques that guide the design process by harvesting design data from multiple sources and improve the accuracy of design verification respectively. Such approach has been successfully used in aero and thermal design as well as in the critical area of materials and fluids engineering. In the future the concerted use of machine learning and high-fidelity methods may allow researchers to conduct reliable virtual tests in the framework of design loops to cut down design time and risk.

Low-fidelity method for rapid aerostructural optimisation and design-space exploration of planar wings

ABSTRACTDuring early phases of wing design, analytic and low-fidelity methods are often used to identify promising design concepts. In many cases, solutions obtained using these methods provide intuition about the design space that is not easily obtained using higher-fidelity methods. This is especially true for aerostructural design. However, many analytic and low-fidelity aerostructural solutions are limited in application to wings with specific planforms and weight distributions. Here, a numerical method for minimising induced drag with structural constraints is presented that uses approximations that apply to unswept planar wings with arbitrary planforms and weight distributions. The method is applied to the National Aeronautics and Space Administration (NASA) Ikhana airframe to show how it can be used for rapid aerostructural optimisation and design-space exploration. The design space around the optimum solution is visualised, and the sensitivity of the optimum solution to changes in weight distribution, structural properties, wing loading and taper ratio is shown. The optimum lift distribution and wing-structure weight for the Ikhana airframe are shown to be in good agreement with analytic solutions. Whereas most modern high-fidelity solvers obtain solutions in a matter of hours, all of the solutions shown here can be obtained in a matter of seconds.

Fluid–structure interaction simulations of a wind turbine rotor in complex flows, validated through field experiments

AbstractAeroelastic simulations of a 2.3 MW wind turbine rotor operating in different complex atmospheric flows are conducted using high fidelity fluid–structure interaction (FSI) simulations. Simpler blade element momentum (BEM) theory based simulations are likewise conducted for comparison, and measurements from field experiments are used for validation of the simulations. Good agreement is seen between simulated and measured forces. It is found that for complex flows, BEM‐based simulations predict similar forces as computational fluid dynamics (CFD)‐based FSI, however with some distinct discrepancies. Firstly, stall is predicted for a large part of the blade using BEM‐based aerodynamics, which are not seen in either FSI simulations or measurements in the case of a high shear. This leads to a more dynamic structural response for BEM‐based simulations than for FSI. For a highly yawed and sheared flow case, the BEM‐based simulations overpredict outboard forces for a significant part of the rotation. This emphasizes the need of validation of BEM‐based simulations through higher fidelity methods, when considering complex flows. Including flexibility in simulations shows only little impact on the considered rotor for both FSI‐ and BEM‐based simulations. In general, the loading of the blades increases slightly, and the rotor wake is almost identical for stiff and flexible FSI simulations.

Accurate Prediction of Loss Using High Fidelity Methods

Abstract Further improvements in aero-engine efficiencies require accurate prediction of flow physics and incurred loss. Currently, the computational requirements for capturing these are not known leading to inconsistent loss predictions even for scale-resolving simulations depending on the chosen convergence criteria. This work investigates two aspects of loss generation using high-fidelity simulation. In the first case study, we look at the effect of resolution on capturing entropy generation rate by simulating a Taylor-Green vortex canonical flow. The second case study focuses on the effect of resolution on flow physics and loss generation and uses a compressor cascade subjected to freestream turbulence. The results show that both resolving local entropy generation rate and capturing the inception and growth of instabilities are critical to accuracy of loss prediction. In particular, the interaction of free-stream turbulence at the leading-edge and development of instabilities in the laminar region of the boundary layer are critical to capturing loss.

StressNet - Deep learning to predict stress with fracture propagation in brittle materials

Catastrophic failure in brittle materials is often due to the rapid growth and coalescence of cracks aided by high internal stresses. Hence, accurate prediction of maximum internal stress is critical to predicting time to failure and improving the fracture resistance and reliability of materials. Existing high-fidelity methods, such as the Finite-Discrete Element Model (FDEM), are limited by their high computational cost. Therefore, to reduce computational cost while preserving accuracy, a deep learning model, StressNet, is proposed to predict the entire sequence of maximum internal stress based on fracture propagation and the initial stress data. More specifically, the Temporal Independent Convolutional Neural Network (TI-CNN) is designed to capture the spatial features of fractures like fracture path and spall regions, and the Bidirectional Long Short-term Memory (Bi-LSTM) Network is adapted to capture the temporal features. By fusing these features, the evolution in time of the maximum internal stress can be accurately predicted. Moreover, an adaptive loss function is designed by dynamically integrating the Mean Squared Error (MSE) and the Mean Absolute Percentage Error (MAPE), to reflect the fluctuations in maximum internal stress. After training, the proposed model is able to compute accurate multi-step predictions of maximum internal stress in approximately 20 seconds, as compared to the FDEM run time of 4 h, with an average MAPE of 2% relative to test data.

A Comparison of Advanced Boiling Water Reactor Simulations between Serpent/CTF and Polaris/DYN3D: Steady State Operational Characteristics and Burnup Evolution

High fidelity modelling for nuclear power plant analysis is becoming more common due to advances in modelling software and the availability of high-performance computers. However, to design, develop and regulate new light water nuclear reactors there are, up until now, limited requirements for high fidelity methods due to the already well-established computational methods already being widely accepted. This article explores the additional detail which can be obtained when using high fidelity methods through Monte Carlo/Sub-channel analysis compared to industrial methods of cross-section/nodal analysis using the Advanced Boiling Water Reactor as a case study. This case study was chosen due to the challenges in modelling two phase flow and the high levels of heterogeneity within the fuel assembly design. The article investigates how to implement such an approach, from a bottom-up procedure, by analysing each stage of the modelling process.

Three-dimensional low-order surge model for high-speed axial compressors

Surge in modern aero-engines can lead to violent disruption of the flow, damage to the blade structures and eventually engine shutdown. Knowledge of unsteady performance and loading during surge is crucial for compressor design, however, the understanding and prediction capability for this phenomenon is still very limited. While useful for the investigation of specific cases, costly experimental tests and high-fidelity CFD simulations cannot be used routinely in the design process of compressor systems. There is therefore an interest in developing a low-order model which can predict compressor performance during surge with sufficient accuracy at significantly reduced computational cost. This paper describes the validation of an unsteady 3D through-flow code developed at Cranfield University for the low-order modelling of surge in axial compressors. The geometry investigated is an 8-stage rig representative of a modern aero-engine IP compressor. Two deep surge events are modelled at part speed and full speed respectively. The results are compared against high-fidelity, full annulus, URANS simulations conducted at Imperial College. Comparison of massflow, pressure and temperature time histories shows a close match between the low-order and the higher-fidelity methods. The low-order model is shown capable of predicting many transient flow features which were observed in the high-fidelity simulations, while reducing the computational cost by up to two orders of magnitude.

Multidisciplinary and Multi-Objective Optimization Considering Aircraft Program Cost and Airline Network

Aerial network routes and their flight frequencies are crucial for the strategic planning of airlines, which must choose optimum airplane types to improve revenue and/or reduce operating costs. In addition, aircraft manufacturers need to identify airplane configurations that better suit airline operations and establish list prices for their products as well as calculate the production and development costs of their products. To address these issues, the present study describes and applies a methodology to determine the optimal aerial transport network simultaneously with the identification of the optimum fleet for that network, namely, an integrated design. In the optimization simulations carried out in the present work, the objectives are the maximization of the airline’s network daily profit and the minimization of the airplane fleet acquisition cost. Each fleet is composed of three types of airliners, selected according to their passenger capacity. Airplanes are designed with high-fidelity methods and realistic performance calculations and must obey a set of requirements, including some related to FAR 25 certification rules. Optimization for a Brazilian network considering 21 cities was carried out with the maximization of the network daily profit and the minimization of the fleet acquisition cost. A comprehensive airplane manufacturer program cost estimation model was implemented, enabling the calculation of net present value of a transport airplane program and financial parameters.

Evaluation of low fidelity and CFD methods for the aerodynamic performance of a small propeller

The increasing use of unmanned aerial vehicles and micro air vehicles creates a strong demand for the accurate aerodynamic performance of small-diameters fixed-pitch propellers. The techniques range from low-fidelity to high-fidelity methods. Among the low-fidelity techniques, there is the blade element momentum theory (BEMT). Computational fluid dynamics (CFD) is categorized as a high-fidelity method. This study aims to employ these methods for the analysis of an APC propeller 14 × 7e. Geometric modeling was performed in a 3D scanner and simulations via BEMT and CFD. Analyses using the BEMT model were conducted with two formulations. The first one considered a linear lift coefficient up to the stall limits, without specifying the post-stall lift coefficients. The second one incorporates three-dimensional flow equilibrium effects, and best represents the post-stall conditions. In the CFD models, two approaches for the compatibility of the rotating and stationary domains were evaluated, the stationary frozen rotor (FR) and the transient arbitrary mesh interface (AMI). The kω SST turbulence model was employed. To investigate the influence of the transition laminar-turbulent boundary layer, the Gamma Theta transitional model was also considered. In the comparison of the results with the experimental tests, it was observed that the appropriate choice of analysis method had a strong relationship with the Reynolds number. For low advanced ratio velocities, the best results were obtained with the BEMT models. BEMT linear model overpredicted the power and efficiency for high advance ratio, while BEMT with tridimensional flow equilibrium presented consistent results with a low computational cost. On the other hand, the CFD-AMI model shows better results for higher advanced ratio velocities but is computationally more expensive. The CFD-FR model maintains a constant pattern of error that, in the analysis of efficiency, did not compromise the results.

Structural Sizing of a Rotorcraft Fuselage Using an Integrated Design Approach

For many years, the primary design objective of new helicopters was the design of the main rotor(s). Within the last couple of years, this approach has changed into an assessment of all helicopter components as an overall system, thus turning rotorcraft design into a highly interdisciplinary process. For instance, aerodynamics, flight mechanics, and the structural evaluation strongly affect each other, and these mutual influences are taken into account from the early phase of the conceptual design. Weight prediction in early design stages represents an essential part of the design process as it determines the basic properties of the rotorcraft. Owing to its function to carry crew and payload but also to serve as the central mounting for all components, the fuselage represents a major part of the rotorcraft. Therefore, the structural design of the fuselage airframe constitutes a significant factor of the rotorcraft design at the preliminary level.<br/> In this paper, an approach to include a higher fidelity method using finite elements for the structural analysis of rotorcraft fuselages within an integrated design environment is presented. Model generation and static analysis are conducted automatically. The helicopter is described using a common parametric data model during the complete design process, therefore providing a fast analysis of model changes. The generic finite element model presented in this paper was generated and structurally sized in about 2.5 min using a standard office computer, thus offering the integration of higher fidelity methods into early design sizing loops.