Multi-dimensional Computational Fluid Dynamics Research Articles

A multi-dimensional CFD framework for fast patient-specific fractional flow reserve prediction

Fractional flow reserve (FFR) is considered as the gold standard for diagnosing coronary myocardial ischemia. Existing 3D computational fluid dynamics (CFD) methods attempt to predict FFR noninvasively using coronary computed tomography angiography (CTA). However, the accuracy and efficiency of the 3D CFD methods in coronary arteries are considerably limited. In this work, we introduce a multi-dimensional CFD framework that improves the accuracy of FFR prediction by estimating 0D patient-specific boundary conditions, and increases the efficiency by generating 3D initial conditions. The multi-dimensional CFD models contain the 3D vascular model for coronary simulation, the 1D vascular model for iterative optimization, and the 0D vascular model for boundary conditions expression. To improve the accuracy, we utilize clinical parameters to derive 0D patient-specific boundary conditions with an optimization algorithm. To improve the efficiency, we evaluate the convergence state using the 1D vascular model and obtain the convergence parameters to generate appropriate 3D initial conditions. The 0D patient-specific boundary conditions and the 3D initial conditions are used to predict FFR (FFRC). We conducted a retrospective study involving 40 patients (61 diseased vessels) with invasive FFR and their corresponding CTA images. The results demonstrate that the FFRC and the invasive FFR have a strong linear correlation (r = 0.80, p < 0.001) and high consistency (mean difference: 0.014 ±0.071). After applying the cut-off value of FFR (0.8), the accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of FFRC were 88.5%, 93.3%, 83.9%, 84.8%, and 92.9%, respectively. Compared with the conventional zero initial conditions method, our method improves prediction efficiency by 71.3% per case. Therefore, our multi-dimensional CFD framework is capable of improving the accuracy and efficiency of FFR prediction significantly.

The methodology of decoupling fuel and thermal nitrogen oxides in multi-dimensional computational fluid dynamics combustion simulation of ammonia-hydrogen spark ignition engines

Under the paradigm of carbon neutrality, ammonia-hydrogen (NH3–H2) blended fuel presents itself as a zero-carbon alternative to petroleum-based fuels, effectively reducing carbon emissions originating from internal combustion engines. In the combustion process of conventional hydrocarbon fuels, the production of nitrogen oxides (NOX) predominantly arises from nitrogen present in the atmosphere, which occurs through the Zeldovich mechanism under high-temperature conditions. These NOX species, commonly referred to as thermal NOX, rely on inert nitrogen. However, the utilization of ammonia fuel activates the reactivity of nitrogen element, leading to the nitrogen-containing species formation, including NOX, termed as fuel NOX. Consequently, the combustion of ammonia-hydrogen fuel entails the coupling of thermally formed nitrogen oxides and fuel-derived nitrogen oxides. The generation of NOX is significantly influenced by the physicochemical environment, while the transient combustion conditions within the engine combustion chamber further complicate the process of NH3–H2 combustion and NOX formation. As a consequence, comprehensively investigating the NOX emission characteristics in NH3–H2 engines presents a considerable challenge. By decoupling fuel NOX and thermal NOX, a more profound understanding of NOX emission control strategies for ammonia-hydrogen engines can be attained. This research paper accomplishes the decoupling of fuel nitrogen elements and atmospheric nitrogen elements in three dimensional computational fluid dynamics engine combustion simulations. The results indicate that this decoupling methodology disregards the differentiation between the fuel NOX pool and the thermal NOX pool, resulting in a slight modification in NOX concentration. Nevertheless, this approach has minimal impact on the combustion process, ensuring that the NOX formation environment remains largely unchanged. Furthermore, it successfully demonstrates the spatial and temporal distribution characteristics of thermal NOX and fuel NOX, thereby furnishing an effective analytical tool for the comprehensive study of NOX emission characteristics in NH3–H2 engines.

Adaptive physics-informed neural operator for coarse-grained non-equilibrium flows

This work proposes a new machine learning (ML)-based paradigm aiming to enhance the computational efficiency of non-equilibrium reacting flow simulations while ensuring compliance with the underlying physics. The framework combines dimensionality reduction and neural operators through a hierarchical and adaptive deep learning strategy to learn the solution of multi-scale coarse-grained governing equations for chemical kinetics. The proposed surrogate’s architecture is structured as a tree, with leaf nodes representing separate neural operator blocks where physics is embedded in the form of multiple soft and hard constraints. The hierarchical attribute has two advantages: (i) It allows the simplification of the training phase via transfer learning, starting from the slowest temporal scales; (ii) It accelerates the prediction step by enabling adaptivity as the surrogate’s evaluation is limited to the necessary leaf nodes based on the local degree of non-equilibrium of the gas. The model is applied to the study of chemical kinetics relevant for application to hypersonic flight, and it is tested here on pure oxygen gas mixtures. In 0-textrm{D} scenarios, the proposed ML framework can adaptively predict the dynamics of almost thirty species with a maximum relative error of 4.5% for a wide range of initial conditions. Furthermore, when employed in 1-textrm{D} shock simulations, the approach shows accuracy ranging from 1% to 4.5% and a speedup of one order of magnitude compared to conventional implicit schemes employed in an operator-splitting integration framework. Given the results presented in the paper, this work lays the foundation for constructing an efficient ML-based surrogate coupled with reactive Navier-Stokes solvers for accurately characterizing non-equilibrium phenomena in multi-dimensional computational fluid dynamics simulations.

Numerical investigation on selecting appropriate piston bowl geometry and compression ratio for gasoline-fuelled homogeneous charge compression ignited light-duty diesel engine

The gasoline-fuelled homogeneous charge compression ignition (HCCI) combustion is a promising approach to mitigate the significant limitations of the traditional compression-ignition combustion for light-duty engine applications. In the present work, the required autoignition quality of gasoline was achieved by blending 2-ethylhexyl nitrate (EHN) with gasoline to ensure combustion stability at the achieved engine loads. A reduced chemical reaction mechanism consisting of 185 species and 219 reactions was developed in the present study for EHN-gasoline-fuelled-HCCI combustion. The mechanism was validated by conducting multi-dimensional computational fluid dynamics (CFD) simulations of HCCI combustion in a modified light-duty diesel engine. A detailed numerical study was done using the validated CFD model to select the suitable piston bowl geometry and geometric compression ratio for better engine performance and lower regulated emissions. Eleven different piston bowl geometries were explored, which also included piston bowl designs popular in traditional compression-ignition (toroid) and spark-ignition (flat) engines. The results indicated that the flat-shaped piston bowl exhibited a 17% improvement in the gross indicated thermal efficiency and a 16% and 40% reduction in UHC and CO emissions while maintaining very low NOx and smoke levels compared to the baseline hemisphere-shaped piston bowl. The engine performance remained the same for the flat-shaped piston bowl geometry even for a 17% increased compression ratio with the added benefit that the reduced EHN concentration (by 48 vol.%) was needed to achieve the same combustion phasing. Thus, the present numerical work proposes a promising approach to improve the overall performance of gasoline-fuelled HCCI combustion by piston bowl design and compression ratio optimization.

Scientific machine learning for modeling and simulating complex fluids

The formulation of rheological constitutive equations-models that relate internal stresses and deformations in complex fluids-is a critical step in the engineering of systems involving soft materials. While data-driven models provide accessible alternatives to expensive first-principles models and less accurate empirical models in many engineering disciplines, the development of similar models for complex fluids has lagged. The diversity of techniques for characterizing non-Newtonian fluid dynamics creates a challenge for classical machine learning approaches, which require uniformly structured training data. Consequently, early machine-learning based constitutive equations have not been portable between different deformation protocols or mechanical observables. Here, we present a data-driven framework that resolves such issues, allowing rheologists to construct learnable models that incorporate essential physical information, while remaining agnostic to details regarding particular experimental protocols or flow kinematics. These scientific machine learning models incorporate a universal approximator within a materially objective tensorial constitutive framework. By construction, these models respect physical constraints, such as frame-invariance and tensor symmetry, required by continuum mechanics. We demonstrate that this framework facilitates the rapid discovery of accurate constitutive equations from limited data and that the learned models may be used to describe more kinematically complex flows. This inherent flexibility admits the application of these "digital fluid twins" to a range of material systems and engineering problems. We illustrate this flexibility by deploying a trained model within a multidimensional computational fluid dynamics simulation-a task that is not achievable using any previously developed data-driven rheological equation of state.

Numerical study on the wall-impinging diesel spray soot generation and oxidation in the cylinder under cold-start conditions of a diesel engine

The combustion of wall-impinging diesel spray of heavy-duty diesel engines deteriorates combustion quality under cold-start conditions, making it difficult to control soot emissions. To investigate the causes of soot increase in the combustion of wall-impinging spray at low temperature and low speed starting conditions, the effect of the starting fuel mass on the soot formation and oxidation process was analyzed using a multidimensional computational fluid dynamics (CFD) model. The results show that the diesel spray is guided by the piston wall and the limited space, the spray impinged on the wall and the vapor-phase fuel flowed in the spray interaction zone. Thus, the soot mainly accumulates in the spray interaction zone, the region near the cylinder head and the bowl wall in the combustion chamber bowl. The soot from the vapor of deposited fuel film in the piston bowl wall and near wall region accumulates continuously in the after combustion stage, becoming the main source of soot emissions at the time of exhaust valve opening (EVO). Increasing the mass of starting fuel raises the mass of the rich mixture and wall-impinging fuel, which enhances the mismatch between fuel and air, resulting in higher soot generation, while soot is more difficult to be completely oxidized by OH radicals, and ultimately soot emissions increase significantly. It can be deduced that the engine-optimized injection strategy may mitigate the increase in soot emissions at high start-up fuel injection conditions.

Modelling of the pulverized coal plasma preparation for combustion

The main goal of this research is to show the action of pulverized coal in the plasma chamber throughcomputer simulation and numerical experiments carried out with the aid of developed well knownthermodynamic, kinetic and multi dimensional computational fluid dynamics based on mathematicalmodels. The data needed for the validation of the numerical procedure were obtained from a cylindricaldirect flow burner equipped with a plasmatron (plasma generator) with 100kW of electric power andmounted on a full-scale boiler (Gusinoozersk TPP, Eastern Siberia). The experiments were carried outusing ‘Tugnuisk’ bituminous coal. Two mathematical models were employed: the one (‘1D Plasma-Coal’) being one-dimensional, but with an emphasis on complex chemistry, the other (3D FAFNIR) beingfully three-dimensional with emphasis on the geometry and overall combustion processes. 1D Plasma-Coal numerical experiments gave the predicted temperatures and velocities of gas and solids along thechamber length; while the concentrations of the gas components (CO, CO2, H2, CH4, C6H6, N2, H2O)were reported for the chamber exit. The degree of coal gasification showed that 54% of coal carbon wasgasified within the plasma chamber. 3D numerical results for plasma jet spreading length were in goodagreement with the measured data, while the temperature profiles within the plasma chamber were overpredicted. The predictions of main species concentrations reveal that oxygen was completely consumedwith the exit product stream consisting of combustible gases, un-burnt volatiles and char particles.

A threshold soot index-based fuel surrogate formulation methodology to mimic sooting tendency of real fuels in 3D-CFD simulations

Particulate Matter emission is an increasing concern for engine manufacturers due to the strong limits imposed by worldwide regulations. Fuel composition plays a key role in determining the extent to which soot is formed during the combustion process. The availability of advanced multidimensional computational fluid-dynamics soot models incorporating soot chemistry pushed researchers to formulate fuel surrogates able to represent sooting tendency of real fuels in the numerical framework. Such studies, which provide information to target research grade fuels, are scarcely present in literature. A methodology is proposed to estimate sooting tendency of commercial gasolines, based on Threshold Soot Index and basic composition information, as well as to formulate tailored ethanol-toluene reference fuel surrogates. The technique relies on a purely mathematical approach to estimate Threshold Soot Index of individual compounds and blends using a structural group contribution-based approach, available in literature, which allows to adapt fuel surrogate palette as needed. Firstly, this approach is demonstrated by means of constant pressure reactor simulations using the Method of Moments. Secondly, a methodology is proposed to formulate fuel surrogates simultaneously targeting the main chemical and physical auto-ignition characteristics and estimated Threshold Soot Index. Several oxygenated and non-oxygenated commercial gasolines available on the market are targeted to provide a wide number of validation cases. Finally, surrogates are used to generate a selection of constant pressure-based soot libraries tested in conjunction with the Sectional Method on a 3D-CFD model of a single-cylinder optically accessible gasoline engine. Surrogates exhibit the same qualitative ranking estimated based on Threshold Soot Index and retain a quantitative scaling in terms of particulate matter formation.

Numerical Investigation of a Free-Piston Hydrogen-Gasoline Engine Linear Generator

The free-piston engine linear generator (FPELG) is being investigated by many researchers because of its high thermal efficiency and its variable compression ratio. However, all researchers focused on the FPELG characteristics with mono-fuel. Therefore, in this paper, the performance of the FPELG that has adopted gasoline with hydrogen as fuel is investigated. The method of coupling the zero-dimensional dynamics model with the multi-dimensional CFD (Computational Fluid Dynamics) combustion model was applied during the simulation process. According to the results, the piston TDC (Top Dead Center), the piston peak piston velocity, and the system operation frequency show a negative correlation with the increase of hydrogen fractions. However, the peak in-cylinder pressure was increased with the hydrogen volume fraction increase, due to the fast flame speed and short combustion duration characteristics of hydrogen. Meanwhile, the indicated efficiency of the free-piston engine was increased from 32.3% to 35.3% with the hydrogen volume fraction change from 0% to 4.5%, when the free-piston engine operates at stoichiometric conditions with fixed ignition timing. In addition, with the ignition timing advance increase, the piston TDC was decreased. The peak piston velocity and the peak in-cylinder pressure were in negative correlation with the ignition timing advance. While the engine indicated that the efficiency was increased with the equivalent degree of ignition timing from 20° to 16°. Therefore, the ignition timing of the FPELG under the spark-ignition combustion mode is supposed to be an effective and practical control variable.

Modelling internal combustion engines with dynamic staggered mesh refinement

Modelling internal combustion engines (ICEs) with multidimensional computational fluid dynamics (CFD) is challenging due to the fuel spray, combustion chemistry and moving piston involved. To accurately simulate ICEs as well as to maintain acceptable computational cost, dynamic mesh refinement techniques could be applied. Although modelling ICEs with staggered mesh is ubiquitous in the engine combustion community, the dynamic refinement of staggered mesh in the context of multidimensional engine simulations is not reported in the literatures due to the numerical difficulties in the momentum solution at the gird fine-coarse interface and to deal with moving boundaries. To address these difficulties, we propose a dynamic refinement scheme to cell-by-cell refine the staggered mesh on-the-fly based on the fuel evaporation and thermochemical state in the multidimensional dynamic staggered mesh and it is able to handle the complexity caused by the moving boundaries in engine simulations and momentum solution at the fine-coarse interface in the dynamic mesh. This mesh refinement scheme is then implemented in the widely-used ICEs simulation codes for engine simulations with fuel sprays. A wide variety of computational examples including shock wave, compression, constant volume fuel spray, constant volume fuel spray combustion are first employed to test the scheme and finally, compression ignition combustion engine simulations are conducted with the proposed methodology.

Control of combustion phasing and operating range extension of natural gas PCCI engines using ozone species

There are two major issues that restrain premixed charge compression ignition (PCCI) engines from being widely employed for industrial applications: difficulties in combustion controllability and limited operating range. These restrictions become even more severe when the engine runs with a very low reactivity fuel, such as natural gas (NG). Ozone gas, which is a species with a very high level of chemical reactivity, can be utilized to control and enhance the combustion process in the PCCI combustion strategy. In the present study, coupled multidimensional computational fluid dynamics (CFD) with proper detailed chemical kinetic mechanisms is employed to investigate the effects of ozone addition to the intake air and fuel mixture in a PCCI engine with different intake temperatures and equivalence ratios. It can be inferred from the results that combustion phasing advancement is in a direct proportion to the amount of ozone addition. It is seen that even low amounts of ozone (10 ppm) have considerable influences on the engine performance and emissions. The results show that adding 1000 ppm ozone makes it possible to decrease the needed initial temperature for PCCI combustion by 80 K. The results also reveal that ozone addition makes it possible to run engine with ultra lean fuel mixtures. It can be concluded that adding controlled amounts of ozone can assist in controlling PCCI engines and also extend operating range of the engine to lower intake air temperatures and also leaner fuel mixtures.

Multi-Dimensional CFD Analysis for the Prediction of Transient Wellbore Circulating Temperature Profile to Guide Offshore Cementing Job

This study investigates how temperature profile changes with time during fluid circulation through the wellbore and annulus by using ANSYS Fluent, to help designing offshore cementing job in the petroleum industry. The use of two-dimensional nature of transient computational-fluid-dynamics (CFD) simulations shows that both flow rheology and fluid distribution in the well affect heat transfer calculations and such an effect is more pronounced at the bottom hole where the flow direction changes from downward to upward. When compared with one-dimensional (1D) simulation, multi-dimensional simulation results predict a longer time for heat transfer consistently, implying that simple 1D simulation techniques tend to under-predict the amount of time associated with cementing job. How fluid distribution within the well changes with time and how the resulting pressure and temperature calculations are affected at different times and locations are also visualized in this study.

Numerical Investigation on Knock Combustion in a Diesel–Dimethyl Ether Dual-Fuel Engine

Numerical investigation on knock was conducted in a direct-injection diesel engine with port premixed dimethyl ether (DME) fuel. Knock combustion at various engine operating conditions with a fixed speed was simulated, and the effects of premixed DME ratio, injection timing, and exhaust gas recirculation (EGR) rate on knock suppression were studied based on the multidimensional computational fluid dynamics (CFD) analysis. A reduced diesel–DME kinetic mechanism coupled with the CFD model was used to obtain local pressure and radical concentration at various monitoring locations within the computation domain. A method with a band-pass filter was then employed to deal with pressure data, and maximum amplitude of pressure oscillation was adopted as an evaluation index of the knock intensity. The analysis of species concentration was also used to quantify knock here. The research results indicated the profiles of pressure and major intermediate species could provide useful information to quantify and predict knock for DME–diesel dual-fuel premixed compression-ignition combustion. Moreover, using EGR was an effective way to reduce the knock intensity and delay the knock onset. A lower DME premixed ratio and retarded injection timing could also mitigate the knock.

The impact of combustion chamber configuration on combustion and emissions of a single cylinder diesel engine fuelled with soybean methyl ester blends with diesel

The multi-dimensional computational fluid dynamics (CFD) simulation involving in-cylinder flow and combustion have been done to study the effect of soybean methyl ester and piston bowl configuration on performance, combustion and pollutant emissions from a single cylinder diesel engine. The baseline engine configuration consists of a hemispherical piston bowl. The investigation has been conducted for biodiesel blends with diesel and different piston bowl configurations such as Toroidal Re-entrant Combustion Chamber (TRCC), Re-entrant Combustion Chamber (RCC) and baseline Hemispherical Combustion Chamber (HCC) for same bowl volume to have constant compression ratio of 17.5. To simulate the in-cylinder flow and combustion process, AVL FIRE code was performed and experimental results of baseline hemispherical bowl have been used to validate the numerical model. The simulation results show that flow behaviour inside the combustion chamber strongly depends on the piston bowl configuration in diesel engine. The results obtained from the simulation for the fuel blends are compared with that of baseline diesel fuel. The brake specific fuel consumption is higher for biodiesel due to its lower heating value compared to baseline mineral diesel. However, significantly better results were obtained from engine having modified combustion chambers mainly due to better air movement and charge mixing.

Forced convection steam-water experimental database in a vertical annulus with local measurements

A new database of steam water flow in a vertical annulus is developed in the Multiphase Thermo-fluid Dynamics Laboratory (MTDL) at the University of Illinois at Urbana-Champaign. The focus of the dataset is on the two-phase phenomena of condensing, flashing, and saturated flow in a vertical channel. The 2 m annular channel with inner and outer diameter of 19.1 mm and 38.1 mm respectively is located downstream of an internally heated annular section. Five measurement ports in this unheated section provide measurement of pressure and local measurements of void fraction, interfacial area concentration, gas velocity, Sauter mean diameter, and liquid temperature. The dataset was designed for validation of one-dimensional system codes and multi-dimensional computational fluid dynamics codes. Detailed local measurements of three representative cases are shown to highlight the complexity of local two-phase flow phenomena observed along the test section. A parametric analysis is also conducted to show the sensitivity of system parameters on the axially evolving two-phase flow. The focus of the adiabatic steam-water flow fills an important gap in validation data between the simplified hydro-dynamic data of air-water flow and the complex thermo-fluid dynamic data of boiling flows.

Numerical Calculation Method of Model Predictive Control for Integrated Vehicle Thermal Management Based on Underhood Coupling Thermal Transmission

The nonlinear model predictive control (NMPC) controller is designed for an engine cooling system and aims to control the pump speed and fan speed according to the thermal load, vehicle speed, and ambient temperature in real time with respect to the coolant temperature and comprehensive energy consumption of the system, which serve as the targets. The system control model is connected to the underhood computational fluid dynamics (CFD) model by the coupling thermal transmission equation. For the intricate thermal management process predictive control and system control performance analysis, a coupling multi-thermodynamic system nonlinear model for integrated vehicle thermal management was established. The concept of coupling factor was proposed to provide the boundary conditions considering the thermal transmission interaction of multiple heat exchangers for the radiator module. Using the coupling factor, the thermal flow influence of the structural characteristics in the engine compartment was described with the lumped parameter method, thereby simplifying the space geometric feature numerical calculation. In this way, the coupling between the multiple thermodynamic systems mathematical model and multidimensional nonlinear CFD model was realized, thereby achieving the simulation and analysis of the integrated thermal management multilevel cooperative control process based on the underhood structure design. The research results indicated an excellent capability of the method for integrated control analysis, which contributed to solving the design, analysis, and optimization problems for vehicle thermal management. Compared to the traditional engine cooling mode, the NMPC thermal management scheme clearly behaved the better temperature controlling effects and the lower system energy consumption. The controller could further improve efficiency with reasonable coordination of the convective thermal transfer intensity between the liquid and air sides. In addition, the thermal transfer structures in the engine compartment could also be optimized.

Optimization of the injection parameters of a diesel/natural gas dual fuel engine with multi-objective evolutionary algorithms

In this study, the injection parameters of a High-Pressure Direct-Injection (HPDI) natural gas engine was optimized by coupling a multi-dimensional computational fluid dynamics (CFD) code with the genetic algorithm (GA). Moreover, the analysis of the combustion and emission characteristics of a diesel/natural gas dual fuel engine are performed in this paper with the introduction of the diesel/natural gas dual fuel injection model into the CFD code. The result reveals that the developed gas fuel injection model is capable to predicate the penetrations of the natural gas injection. For the dual fuel engine with simultaneous reduction of NOx and soot, the optimized diesel injection timing is at −15.2 °CA ATDC, the gas injection timing is around −6.0 °CA ATDC, and the circumferential offset angle between the diesel and natural gas nozzle hole is around 8.8°. The lowest ISFC condition is achieved when the injection interval between the diesel and natural gas is 1.38 °CA. For the high soot emissions cases, the gas injection timings are later compared with the other cases. In the dual fuel engine, most of the NO emissions are generated through the combustion of the direct injected natural gas. Soot formation region is mainly located in the squish volume close to the cylinder walls, where the temperature is low which leads to a lower soot oxidation rate.

Exploring the potential of machine learning in reducing the computational time/expense and improving the reliability of engine optimization studies

Past research has shown that multidimensional computational fluid dynamics modeling in combination with a genetic algorithm method is an effective approach for optimizing internal combustion engine design. However, optimization studies performed with a detailed computational fluid dynamics model are time intensive, which limits the practical application of this approach. This study addresses this issue by using a machine learning approach called Gaussian process regression in combination with computational fluid dynamics modeling to reduce the computational optimization time. An approach was proposed where the Gaussian process regression model could be used instead of the computational fluid dynamics model to predict the outputs of the genetic algorithm optimization. In this approach, for every nth generation of the genetic algorithm, the data from the previous n − 1 generations was used to train the Gaussian process regression model. The approach was tested on an engine optimization study with five input parameters. When the genetic algorithm was run solely with computational fluid dynamics, the optimization took 50 days to complete. In comparison with the computational fluid dynamics and Gaussian process regression approach, the computational time was reduced by 62%, and the optimization was completed in 19 days using the same amount of computational resources. Additional parametric studies were performed to investigate the impact of genetic algorithm + Gaussian process regression parameters. Results showed that either reducing the initial dataset size or relaxing the error criterion resulted in increased Gaussian process regression evaluations within the genetic algorithm. However, relaxing the error criterion was found to impact the model predictions negatively. The initial dataset size was found to have a negligible impact on the final optimum design. Finally, the potential of machine learning in further improving the optimization process was explored by using the Gaussian process regression model to check for the robustness of the designs to operating parameter variations during the optimization. The genetic algorithm was repeated with the modified procedure and it was shown that adding the stability check resulted in a different, more reliable and stable optimum solution.

Modelling compression ignition engines by incorporation of the flamelet generated manifolds combustion closure

Tabulated chemistry models allow to include detailed chemistry effects at low cost in numerical simulations of reactive flows. Characteristics of the reactive fluid flows are described by a reduced set of parameters that are representative of the flame structure at small scales so-called flamelets. For a specific turbulent combustion configuration, flamelet combustion closure, with proper formulation of the flame structure can be applied. In this study, flamelet generated manifolds (FGM) combustion closure with progress variable approach were incorporated with OpenFOAM® source code to model combustion within compression ignition engines. For IC engine applications, multi-dimensional flamelet look-up tables for counter flow diffusive flame configuration were generated. Source terms of non-premixed combustion configuration in flamelet domain were tabulated based on pressure, temperature of unburned mixture, mixture fraction, and progress variable. A new frozen flamelet method was introduced to link one dimensional reaction diffusion space to multi-dimensional Computational Fluid Dynamics (CFD) physical space to fulfill correct modelling of thermal state of the engine at expansion stroke when charge composition was changed after combustion and reaction rates were subsided. Predictability of the developed numerical framework were evaluated for Sandia Spray A (constant volume vessel), Spray B (light duty optical Diesel engine), and a heavy duty Diesel engine experiments under Reynolds averaged Navier Stokes turbulence formulation. Results showed that application of multi-dimensional FGM combustion closure can comprehensively predict key parameters such as: ignition delay, in-cylinder pressure, apparent heat release rate, flame lift-off , and flame structure in Diesel engines.

Petascale scramjet combustion simulation on the Tianhe-2 heterogeneous supercomputer

Abstract Combustion simulation is complex and computationally expensive as it involves integration of fundamental chemical kinetics and multidimensional Computational Fluid Dynamics (CFD) models. This paper presents our efforts porting a real-world supersonic combustion simulation application to the heterogeneous architecture consists of multi-core CPUs and Intel Many Integrated Core (MIC) coprocessors. Scalable OpenMP parallelization is added to make use of the large number of cores on CPUs and MIC coprocessors. Single thread performance optimizations are addressed to improve the computational efficiency. CPU and MIC collaborative algorithm, along with a series of techniques to improve the data transfer efficiency and load balance, are applied. Performance evaluation is performed on the Tianhe-2 supercomputer. The results show that on a single node, the optimized CPU-only version is 8.33 times faster than the baseline version, and the CPU + MIC heterogeneous version is again 3.07 times faster than the optimized CPU-only version. The resulting codes effectively scale to 5120 nodes (998,400 cores) on a mesh with 27.46 Giga cells. Given that the total number of floating-point operations is reduced by about 10 times after our optimizations, the heterogeneous version still achieves a sustained double precision floating-point performance of 0.46 Pflops on 5120 nodes. This demonstrates Petascale heterogeneous computing capabilities for real-world supersonic combustion problems.