Bulut Tekgül

Diesel-fueled engines still hold a large market share in the medium and heavy-duty transportation sector. However, the increase in fossil fuel prices and the strict emission regulations are leading engine manufacturers to seek cleaner alternatives without a compromise in performance. Alcohol-based fuels, such as ethanol, offer a promising alternative to diesel fuel in meeting regulatory demands. Ethanol provides cleaner combustion and lower levels of soot due to its chemical properties, in particular its lower level of carbon content. In addition, the stoichiometric operating conditions of alcohol fueled engines enable the mitigation of NOx emissions in aftertreatment stage. With the promise of retrofitting diesel engines to run on ethanol to reduce emissions, the thermal efficiency of these engines remains the primary optimization target. In order to find the optimal ethanol-fueled engine design that maximizes the thermal efficiency, a large design space needs to be investigated using engineering tools.

In this study, previous research by the authors on optimizing the design of a single-cylinder ethanol-fueled engine was extended to explore the design space for a heavy-duty multi-cylinder engine configuration. A heavy-duty engine setup with multiple operating conditions at different engine speeds and loads were considered. A design optimization analysis was performed to identify the potential designs that maximize the indicated thermal efficiency in an ethanol-fueled compression ignition engine. First, a computational fluid dynamics (CFD) model of the engine was validated using experimental data for four drive cycle points. Using a design of experiments (DoE) approach and a parameterized piston bowl geometry, the model was then exercised to explore the relationship among geometric features of the piston bowl and spray targeting angle and indicated thermal efficiency across all tested operating conditions. After evaluating 165 candidate designs, a piston bowl geometry was identified that yielded an increase between 1.3 to 2.2 percentage points in indicated thermal efficiency for all tested conditions, while satisfying the operational design constraints for peak pressure and maximum pressure rise rate. The increased performance was attributed to enhanced mixing that led to the formation of a more homogeneous distribution of in-cylinder temperature and equivalence ratio, higher combustion temperatures, and shorter combustion duration. Finally, a Bayesian optimization (BOpt) analysis was employed to find the optimal piston bowl geometry with a fixed spray injector angle for one of the operating conditions. Using BOpt, a piston candidate was identified that resulted in a 1.9 percentage point increase in thermal efficiency from the baseline design, yet only required 65% of the design samples investigated using the DoE approach.

In this study, we investigate the effect of different split injection strategies on ignition delay time (IDT) and heat release rate (HRR) characteristics in Reactivity Controlled Compression Ignition conditions via large-eddy simulation and finite-rate chemistry. A diesel surrogate (n-dodecane) is injected into a domain with premixed methane and oxidiser in two separate injection pulses. Three different split injection strategies are investigated by fixing the amount of total fuel mass: varying the first injection timing, varying the second injection timing, and changing the fuel mass ratio between the two injections at a fixed injection timing. A compression heating mass source term approach is utilised to take compression heating into account. The main findings of the study are as follows: (1) In general, the IDT shifts towards the top-dead centre when the first injection is advanced or the second injection is retarded. The size and spatial pattern of the ignition kernels are shown to depend on the dwell time between the injections. (2) A precisely timed first injection offered the best control over ignition and HRR characteristics. However, advancing the first injection may lead to over-dilution downstream, preventing volumetric ignition and reducing the peak HRR value. (3) Approximately 21% decrease in the maximum HRR value, as well as a factor of 2.8 increase in combus- tion duration could be achieved by advancing the first injection timing. (4) As indicated by frozen-flow chemistry analysis, in the investigated configurations, the reactivity stratification is controlled by mixture stratification rather than temperature. The find- ings indicate that the first injection controls the downstream reactivity stratification, offering ignition and HRR control.

Detailed chemistry-based computational fluid dynamics (CFD) simulations are computationally expensive due to the solution of the underlying chemical kinetics system of ordinary differential equations (ODEs). Here, we introduce a novel open-source library aiming at speeding up such reactive flow simulations using OpenFOAM, an open-source software for CFD. First, our dynamic load balancing model by Tekgül et al. [“DLBFoam: An open-source dynamic load balancing model for fast reacting flow simulations in OpenFOAM,” Comput. Phys. Commun. 267, 108073 (2021)] is utilized to mitigate the computational imbalance due to chemistry solution in multiprocessor reactive flow simulations. Then, the individual (cell-based) chemistry solutions are optimized by implementing an analytical Jacobian formulation using the open-source library pyJac, and by increasing the efficiency of the ODE solvers by utilizing the standard linear algebra package. We demonstrate the speed-up capabilities of this new library on various combustion problems. These test problems include a two-dimensional (2D) turbulent reacting shear layer and three-dimensional (3D) stratified combustion to highlight the favorable scaling aspects of the library on ignition and flame front initiation setups for dual-fuel combustion. Furthermore, two fundamental 3D demonstrations are provided on non-premixed and partially premixed flames, viz., the Engine Combustion Network Spray A and the Sandia flame D experimental configurations, which were previously considered unfeasible using OpenFOAM. The novel model offers up to two orders of magnitude speed-up for most of the investigated cases. The openly shared code along with the test case setups represent a radically new enabler for reactive flow simulations in the OpenFOAM framework.

In dual-fuel compression-ignition engines, replacing common fuels such as methane with renewable and widely available fuels such as methanol is desirable. However, a fine-grained understanding of diesel/methanol ignition compared to diesel/methane is lacking. Here, large-eddy simulation (LES) coupled with finite rate chemistry is utilized to study diesel spray-assisted ignition of methane and methanol. A diesel surrogate fuel (n-dodecane) spray is injected into ambient methane-air or methanol-air mixtures at a fixed lean equivalence ratio ϕLRF = 0.5 at various ambient temperatures (Tamb = 900, 950, 1000 K). The main objectives are to (1) compare the ignition characteristics of diesel/methanol with diesel/methane at different Tamb, (2) explore the relative importance of low-temperature chemistry (LTC) to high-temperature chemistry (HTC), and (3) identify the key differences between oxidation reactions of n-dodecane with methane or methanol. Results from homogeneous reactor calculations as well as 3 + 3 LES are reported. For both DF configurations, increasing Tamb leads to earlier first- and second-stage ignition. Methanol/n-dodecane mixture is observed to have a longer ignition delay time (IDT) compared to methane/n-dodecane, for example ≈ three times longer IDT at Tamb = 950 K. While the ignition response of methane to Tamb is systematic and robust, the Tamb window for n-dodecane/methanol ignition is very narrow and for the investigated conditions, only at 950 K robust ignition is observed. For methanol at Tamb = 1000 K, the lean ambient mixture autoignites before spray ignition while at Tamb = 900 K full ignition is not observed after 3 ms, although the first-stage ignition is reported. For methanol, LTC is considerably weaker than for methane and in fully igniting cases, heat release map analysis demonstrates the dominant contribution of HTC to total heat release rate for methanol. Reaction sensitivity analysis shows that stronger consumption of OH radicals by methanol compared to methane leads to the further delay in the spray ignition of n-dodecane/methanol. Finally, a simple and novel approach is developed to estimate IDT in reacting LES using zero-dimensional IDT calculations weighted by residence time from non-reacting LES data.

Computational load imbalance due to direct integration of chemical kinetics is a well-known performance issue in parallel reacting flow simulations. We introduce an open-source dynamic load balancing model to address this problem within OpenFOAM, an open-source C++ library for Computational Fluid Dynamics (CFD). Due to the commonly applied operator splitting practice in reactive flow solvers, chemistry can be treated as an independent stiff ordinary differential equation (ODE) system within each computational cell. As a result of highly non-linear characteristics of chemical kinetics, a large variation in convergence rates of ODE integrator may occur, leading to a high load imbalance across multiprocessor configurations. However, the independent nature of chemistry ODE systems leads to a problem that can be parallelized easily (also termed embarrassingly parallel in the literature) during the flow solution. The presented model takes advantage of this feature and balances the chemistry load across available resources. When the load balancing is utilized together with a reference mapping model also introduced in this paper, a speed-up by a factor of 10 is reported for practical reactive flow simulations. To the best of our knowledge, this model is the first open-source implementation of chemistry load balancing in the literature.

This dissertation belongs to the research field of computational physics and numerical combustion modeling. Computational fluid dynamics (CFD) and finite-rate chemistry are employed to numerically investigate reacting spray combustion at internal combustion engine (ICE) conditions. The main focus of the study is the ignition characteristics of dual-fuel (DF) sprays, where a high-reactivity liquid fuel spray (here n-dodecane) is injected into an hot ambient environment with premixed low-reactivity fuel (here methane) and oxidizer. In particular, operating conditions similar to reactivity controlled compression ignition (RCCI) combustion technology are adopted. Such an approach may provide better ignition timing control with lower emissions and higher thermal efficiency. Although DF combustion and RCCI are promising engine combustion concepts, there are still operational issues related to RCCI such as ignition delay time (IDT) and heat release rate (HRR) control, as well as finding optimal operating conditions. Numerical simulations and high-performance computing enable detailed investigations on the 3D physics and chemistry of such reacting flow problems.

The dissertation consists of three journal publications. Large-eddy simulation (LES) coupled with finite-rate chemistry approach is utilized using OpenFOAM, a C++ library for CFD simulations. In Publication I, we investigate the effect of ambient temperature on DF spray ignition characteristics, and the inhibiting effect of methane on n-dodecane oxidation chemistry at lower temperatures. In Publications II and III, the effect of spray injection timing using single and double injection strategies within RCCI context is investigated by introducing a compression heating model to account for dynamic thermophysical conditions due to piston compression. The ignition characteristics of DF sprays are investigated and the effect of injection timing is analyzed.

The main findings of this dissertation can be summarized as follows: 1) In stationary conditions without dynamic compression, DF spray simulations at varying engine-relevant ambient temperatures indicate that methane has an inhibiting effect on n-dodecane spray IDT, especially at lower temperatures. This behavior was also shown to be qualitatively insensitive to the selection of chemical mechanism. 2) Under dynamic compression, in RCCI relevant conditions, it is observed that advancing injection timing first advances, then retards the IDT, due to over-leaning of the injected spray at very advanced injection timings. The finding is noted to be consistent with experimental observations in a laboratory engine. 3) It was found that the source of reactivity stratification in DF spray combustion is mostly mixture stratification, rather than thermal stratification. 4) With advanced injection timing, the contribution of lean mixture conditions and low-temperature chemistry modes to HRR increases. 5) In a double-injection scenario, the injection timings and the dwell time between the two injections can be adjusted to control the local reactivity, IDT, and HRR characteristics.

Here, a large-eddy simulation and finite-rate chemistry solver (see Kahila et al. Combustion and Flame, 2019) is utilized to investigate diesel spray assisted ignition of a lean methane-air mixture. A compression heating model is utilized to emulate ambient temperature and pressure increase in the envisioned compression ignition (CI) system. The key parameter is the start of injection (SOI) relative to a virtual top dead center (TDC), where the peak adiabatic compression pressure/temperature would be achieved. Altogether five different cases are investigated by advancing the SOI further away from the TDC with constant injection duration. The main findings of the paper are as follows: 1) Advancing the SOI advances the ignition timing of the spray with respect to the TDC from 0.91 to 7.08 CAD. However beyond a critical point the ignition time starts retarding towards the TDC to 4.46 CAD due to the excessively diluted diesel spray. 2) Advancing the SOI increases the contribution of leaner mixtures to heat release rate (HRR). Consequently, the low-temperature combustion HRR mode becomes more pronounced (from 33.9% to 76.7%) while the total HRR is reduced by a factor of 4. 3) Ignition is observed for all the investigated SOIs. However, the numerical findings indicate that advancing the SOI decreases the ignition kernel size, resulting in weaker ignition. 4) An ignition index analysis with frozen flow assumption indicates that for the SOIs close to the TDC the HRR mode appears as spray mixing controlled, while for advanced SOI it becomes reactivity controlled, dominated by fuel stratification.

Development of marine engines could largely benefit from the broader usage of methanol and hydrogen which are both potential energy carriers. Here, numerical results are presented on tri-fuel (TF) ignition using large-eddy simulation (LES) and finite-rate chemistry. Zero-dimensional (0D) and three-dimensional (3D) simulations for n-dodecane spray ignition of methanol/hydrogen blends are performed. 0D results reveal the beneficial role of hydrogen addition in facilitating methanol ignition. Based on LES, the following findings are reported: 1) Hydrogen promotes TF ignition, significantly for molar blending ratios βX = [H2]/([H2]+[CH3OH]) ≥0.8. 2) For βX = 0, unfavorable heat generation in ambient methanol is noted. We provide evidence that excessive hydrogen enrichment (βX ≥ 0.94) potentially avoids this behavior, consistent with 0D results. 3) Ignition delay time is advanced by 23–26% with shorter spray vapor penetrations (10–15%) through hydrogen mass blending ratios 0.25/0.5/1.0. 4) Last, adding hydrogen increases shares of lower and higher temperature chemistry modes to total heat release.

Synthetic fuels are needed to replace their fossil counterparts for clean transport. Presently, their production is still inefficient and costly. To enhance the process of methanol production from CO2 and H2 and reduce its cost, a particle-resolved numerical simulation tool is presented. A global surface reaction model based on the Langmuir-Hinshelwood-Hougen-Watson kinetics is utilized. The approach is first validated against standard benchmark problems for non-reacting and reacting cases. Next, the method is applied to study the performance of methanol production in a 2D fixed-bed reactor under a range of parameters. It is found that methanol yield enhances with pressure, catalyst loading, reactant ratio, and packing density. The yield diminishes with temperature at adiabatic conditions, while it shows non-monotonic change for the studied isothermal cases. Overall, the staggered and the random catalyst configurations are found to outperform the in-line system.

We present 3D numerical results on tri-fuel (TF) combustion using large-eddy simulation (LES) and finite rate chemistry. The TF concept was recently introduced by Karimkashi et al. (Int. Journal of Hydrogen Energy, 2020) in 0D. Here, the focus is on spray assisted ignition of methane-hydrogen blends. The spray acts as a high-reactivity fuel (HRF) while the ambient premixed methane-hydrogen blend acts as a low-reactivity fuel (LRF) mixture. Better understanding on such a TF process could enable and motivate more extensive hydrogen usage in e.g. compression ignition (CI) marine engines where spray assisted dual-fuel (DF) combustion is already utilized. The studied spray setup is based on the modified ECN Spray A case, see Kahila et al. (Combustion and Flame, 2019) for dual-fuel (DF) combustion. The ambient pressure and temperature are Tamb = 900 K and pamb = 60 bar. The hydrogen content of the LRF blend is varied systematically by changing the molar fraction. The main added value of the study is that we extend the TF concept to 3D. The particular findings of the study are as follows: 1) Consistent with Karimkashi et al. 2020, hydrogen delays ignition also in 3D and the effect becomes significant for x>0.5. 2) The ratio between the first and the second stage ignition delay times are 2 for 0D and 3 for 3D. Furthermore, the ratio between 3D and 0D ignition delay times is given as 2 for all TF cases. 3) Finally, consistent with Karimkashi et al. 2020, also in 3D the high temperature combustion (HTC) heat release mode is shown to appear stronger in TF than the low temperature combustion (LTC) mode compared to DF methane-diesel combustion.

Here, a finite-rate chemistry large-eddy simulation (LES) solver is utilized to investigate dual-fuel (DF) ignition process of n-dodecane spray injection into a methane–air mixture at engine-relevant ambient temperatures. The investigated configurations correspond to single-fuel (SF) = 0 and DF = 0.5 conditions for a range of temperatures. The simulation setup is a continuation of the work by Kahila et al. (2019, Combustion and Flame) with the baseline SF spray setup corresponding to the Engine Combustion Network (ECN) Spray A configuration. First, ignition is investigated at different ambient temperatures in 0D and 1D studies in order to isolate the effect of chemistry and chemical mechanism selection to ignition delay time (IDT). Second, 3D LES of SF and DF sprays at three different ambient temperatures is carried out. Third, a reaction sensitivity analysis is performed to investigate the effect of ambient temperature on the most sensitive reactions. The main findings of the paper are as follows: (1) DF ignition characteristics depend on the choice of chemical mechanism, particularly at lower temperatures. (2) Addition of methane to the ambient mixture delays ignition, and this effect is the strongest at lower temperatures. (3) While the inhibiting effect of methane on low- and high-temperature IDT’s is evident, the time difference between these two stages is shown to be only slightly dependent on temperature. (4) Reaction sensitivity analysis indicates that reactions related to methane oxidation are more pronounced at lower temperatures. The provided quantitative results indicate the strong ambient temperature sensitivity of the DF ignition process.

In dual-fuel compression ignition engines, a high-reactivity fuel, such as diesel, is directly injected to the engine cylinder to ignite a mixture of low-reactivity fuel and air. This study targets improving the general understanding on the dual-fuel ignition phenomenon using zero-dimensional homogeneous reactor studies and three-dimensional large eddy simulation together with finite-rate chemistry. Using the large eddy simulation framework, n-dodecane liquid spray is injected into the lean ambient methane–air mixture at 𝜙=0.5. The injection conditions have a close relevance to the Engine Combustion Network Spray A setup. Here, we assess the effect of two different chemical mechanisms on ignition characteristics: a skeletal mechanism with 54 species and 269 reaction steps (Yao mechanism) and a reduced mechanism with 96 species and 993 reaction steps (Polimi mechanism). Altogether three ambient temperatures are considered: 900, 950, and 1000 K. Longer ignition delay time is observed in three-dimensional large eddy simulation spray cases compared to zero-dimensional homogeneous reactors, due to the time needed for fuel mixing in three-dimensional large eddy simulation sprays. Although ignition is advanced with the higher ambient temperature using both chemical mechanisms, the ignition process is faster with the Polimi mechanism compared to the Yao mechanism. The reasons for differences in ignition timing with the two mechanisms are discussed using the zero-dimensional and three-dimensional large eddy simulation data. Finally, heat release modes are compared in three-dimensional large eddy simulation according to low- and high-temperature chemistry in dual-fuel combustion at different ambient temperatures. It is found that Yao mechanism overpredicts the first-stage ignition compared to Polimi mechanism, which leads to the delayed second-stage ignition in Yao cases compared to Polimi cases. However, the differences in dual-fuel ignition for Polimi and Yao mechanisms are relatively smaller at higher ambient temperatures.

In this study, various mixing and evaporation modeling assumptions typically considered for large-eddy simulation (LES) of the well-established Engine Combustion Network (ECN) Spray A are explored. A coupling between LES and Lagrangian particle tracking (LPT) is employed to simulate liquid n-dodecane spray injection into hot inert gaseous environment, wherein Lagrangian droplets are introduced from a small cylindrical injection volume while larger length scales within the nozzle diameter are resolved. This LES/LPT approach involves various modeling assumptions concerning the unresolved near-nozzle region, droplet breakup, and LES subgrid scales (SGS) in which their impact on common spray metrics is usually left unexplored despite frequent utilization. Here, multi-parametric analysis is performed on the effects of (i) cylindrical injection volume dimensions, (ii) secondary breakup model, particularly Kelvin–Helmholtz Rayleigh–Taylor (KHRT) against a no-breakup model approach, and (iii) LES SGS models, particularly Smagorinsky and one-equation models against implicit LES. The analysis indicates the following findings: (i) global spray characteristics are sensitive to radial dimension of the cylindrical injection volume, (ii) the no-breakup model approach performs equally well, in terms of spray penetration and mixture formation, compared with KHRT, and (iii) the no-breakup model is generally insensitive to the chosen SGS model for the utilized grid resolution.

The present study is a continuation of the previous work by Kahila et al. (2019), in which a dual-fuel (DF) ignition process was numerically investigated by modeling liquid diesel-surrogate injection into a lean methane-air mixture in engine relevant conditions. Earlier, the injection duration (tinj) of diesel-surrogate exceeded substantially the characteristic autoignition time scale. Here, such a pilot spray ignition problem is studied at a fixed mass flow rate but with a varying tinj. The focus is on understanding the influence of pilot quantity on spray dilution process and low- and high-temperature chemistry. In total, ten cases are computed with multiple diesel pilot quantities by utilizing a newly developed large-eddy simulation/finite-rate chemistry solver. The baseline spray setup corresponds to the Engine Combustion Network (ECN) Spray A configuration, enabling an extensive validation of the present numerical models and providing a reference case for the DF computations. Additionally, experimental results from a single-cylinder laboratory engine are provided to discuss the ignition characteristics in the context of a real application. The main results of the present study are: (1) reducing tinj introduces excessive dilution of the DF mixture, (2) dilution lowers the reactivity of the DF mixture, leading to delayed high-temperature ignition and slow overall methane consumption, (3) low enough pilot quantity (tinj < 0.3 ms) may lead to very long ignition delay times, (4) cumulative heat release is dominated by low/high-temperature chemistry at low/high tinj values, (5) analysis of the underlying chemistry manifold implies that the sensitivity of ignition chemistry on mixing is time-dependent and connected to the end of injection time, and 6) long ignition delay times at very low tinj values can be decreased by decreasing injection pressure.

This thesis presents a comprehensive optimization study of In-situ Adaptive Tabulation (ISAT) control parameters in the context of enhancing the computational efficiency of reactive marine diesel engine simulations using computational fluid dynamics (CFD). Historically, despite ISAT seeing numerous successful applications in the realm of homogeneous charge combustors, the choice of ISAT parameters has largely eluded detailed studies. This is especially evident in the context of highly in-homogeneous combustion environments, such as those typically found in diesel engines. To address this, a detailed Bayesian optimization study was carried out to systematically tune ISAT parameters, with the objective of minimizing computational cost while preserving the fidelity of the resolved combustion processes. The methodology of this thesis encompasses the formulation and implementation of an end-to-end, multi-objective Bayesian optimization framework, innately integrated with the OpenFOAM™ CFD environment to efficiently explore the ISAT configuration parameter space. Standard Gaussian process surrogate models were combined with state-of-the-art acquisition functions, facilitating the exploration of the trade-offs between simulation accuracy and the computational reduction achieved through tabulation. The optimization campaign was conducted using a series of 3D CFD simulations of a marine diesel engine under reactive diesel combustion conditions, incorporating detailed chemistry, a moving computational mesh, and diesel surrogate fuel injection. This provided a challenging, constant thermochemical environment for evaluating ISAT's behavior in. The results highlight the importance of rigorously choosing ISAT's configuration parameters. Poor ISAT configurations were shown to crash and stall simulations, resulting in a negative effect on computational efficiency. Despite this, a wide range of computational reduction was shown to be possible, ranging from less than 5 % to up to 50 %, with broad effects on simulation accuracy. Overall, the results show a strong dependence between the gain in computational efficiency, and the loss in solution accuracy. Finally, guidelines for maximizing the likelihood of good parameter selection are given, together with example configuration values.

This thesis concerns computational fluid dynamics (CFD) simulation of an internal combustion engine (ICE). The aim is to perform an unsteady Reynolds averaged Navier-Stokes (RANS) simulation of an internal combustion engine under motored conditions, using newly implemented features in OpenFOAM, an open-source software for CFD simulations. Due to its simplistic design and availability of experimental data, the TCC-III engine running at 800 rpm was chosen as the engine case study. The engine was developed at the University of Michigan, USA, and the geometry and the experimental data are publicly available. The meshing of the engine was performed using two different software, a commercial (GridPro) and an open-source (snappyHexMesh). Engine valve and piston motion were achieved through Wärtsilä’s in-house mesh mover model, which will be released as a part of OpenFOAM-dev in the future. Simulation results were compared in a comprehensive manner with ensembled averaged flow field data originating from particle image velocimetry (PIV). Overall, the simulation results from both mesh has shown acceptable agreement with the PIV data. However, results from GridPro mesh yielded more accurate and faster results compared to snappyHexMesh, due to GridPro’s higher-quality, block-structured, body-conformed meshing capabilities. The generated setups provide a pathway for further benchmarking of numerical methods and models of an IC engine CFD using OpenFOAM.