UW-Madison | College of Engineering | Search

SAE 2016 World Congress

Tuesday, April 12

Piston Geometry Effects on Compression Stroke Swirling Flow in a Light-Duty Optical Diesel Engine
16PFL-1192 (Oral Only) Zha, K., Kurtz, E.M., Park, C., Warey, A., Peterson, R.C., Perini, F., Busch, S., and Reitz, R.D.
Fluid flow Measurement & Analysis
(Session Code: PFL140)
Room 414A,
9:30 a.m.

For direct-injection, swirl-supported light-duty Diesel engines, in-cylinder flow asymmetry prior to fuel injection can lead to an asymmetrical mixture preparation process. Reducing mixing asymmetry is considered as one strategy to reduce unburned hydrocarbon (UHC) emissions under Low-Temperature Combustion (LTC) conditions. Previous planar PIV investigations have demonstrated that the swirl azimuthal asymmetry is initiated by the intake port orientation for conventional four-valve twin-port design. Interestingly, the observed flow asymmetry under two swirl ratios (2.2 and 3.5) in the late compression stroke suggests that the piston geometry effect also plays an important role in terms of swirl axis tilting. In order to assess the effects of piston geometry on swirling flow in the late compression stroke, experimental (planar PIV) and numerical investigations were carried out. A conventional re-entrant bowl piston and a stepped-lip bowl piston were used. Analysis of the distortion-corrected planar PIV results taken for two steady-state swirl ratios yields information about in-cylinder flow during the late compression stroke in an optical light-duty Diesel engine. The results of this experimental approach, combined with theoretical considerations (1-D simulation) and 3D CFD simulations, provide further insight into swirl center motion and swirl ratio temporal evolution and reveal the extents to which squish height and piston bowl geometry can affect swirl development and flow asymmetry.

Insights into Primary Atomization Behavior
(Oral only 16PFL-1053) Deshpande, S., Gurjar, S., Trujillo, M.F.
Fuel Injection and Sprays
(Session code PFL320)
Room 414A,
1:00 PM

Linear stability analyses are often used in Lagrangian-Eulerian spray simulations, a prominent simulation method, to model the dynamics occurring in the near-nozzle region. In the present work, these instability predictions are re-examined by first generalizing the treatment of interfacial conditions and related assumptions with a two-phase Orr-Sommerfeld (OS) system, and second, by employing highly resolved-Volume-of-Fluid (VoF) simulations. After presenting some validation exercises for both the VoF and OS solvers, the OS predictions are compared to earlier studies from the literature leading to reasonable agreement in the limit as the boundary layer thickness tends to zero. Results from VoF simulations of liquid sheet injection are used to characterize the range of scales of the liquid structures immediately before atomization. The mean value in this range is found to be approximately two to three orders of magnitude larger than the corresponding predictions from previous studies. A two-phase mixing layer under the same physical conditions is used to examine this disparity, revealing that within the linear regime, relatively good agreement exists between the VoF and OS predicted instability mechanisms. However, the most unstable mode in the linear regime is too small to cause a fracture or atomization of the liquid sheet and hence cannot be directly responsible for the atomization. The generation of a much larger mode, which emerges well beyond the linear regime, is the one causing breakup.

Near-Field Spray Characterization via Highly-Resolved Simulations
(Oral only 16PFL-1081) Gurjar, S., Trujillo, M.F.
Fuel Injection and Sprays
(Session code PFL320)
Room 414A,
1:30 PM

A qualitative and quantitative study of the atomization of a circular liquid jet is performed by employing highly resolved Volume-of-Fluid (VoF) simulations. Results reveal that the entire flow can be characterized as having 3 distinct regimes: intact liquid core domain with interfacial instabilities, large-scale structures and dense spray, and a dilute spray region. Different metrics are used to identify these three regions. Additionally, the level of numerical resolution of the simulations is also quantified. Within the second regime, it is reported that the slip velocity between the liquid and gas phases is found to increase along the jet axis, contrary to intuitive understanding of the flow. Similar trends are obtained for different inlet velocity profiles, thereby permitting the generalization of this phenomenon. The flow mass distribution obtained from simulation results is also compared against corresponding X-ray radiography-based experimental results, yielding decent agreement.

A Progress Review on Soot Experiments and Modeling in the Engine Combustion Network (ECN)
2016-01-0734. Skeen, S., Manin, J., Pickett, L.M., Cenker, E., Bruneaux, G., Kondo, K., Aizawa, T., Westlye, F., Dalen, K., Ivarsson, A., Xuan, T., Garcia-Oliver, J.M., Pei, Y., Som, S., Wang, H., Reitz, R.D., Lucchini, T., D'Errico, G., Farrace, D., Pandurangi, S.S., Wright, Y.M., Chishty, M.A., Bolla, M., and Hawkes, E.
Combustion in Compression-Ignition Engines: In-Cylinder Processes (Part 2 of 2)
(Session Code: PFL222)
Room 411B,
2:30 p.m.

The 4th Workshop of the Engine Combustion Network (ECN) was held September 5-6, 2015 in Kyoto, Japan. This manuscript presents a summary of the progress in experiments and modeling among ECN contributors leading to a better understanding of soot formation under the ECN “Spray A” configuration and some parametric variants. Relevant published and unpublished work from prior ECN workshops is reviewed. Experiments measuring soot particle size and morphology, soot volume fraction (fv), and transient soot mass have been conducted at various international institutions providing target data for improvements to computational models. Multiple modeling contributions using both the Reynolds Averaged Navier-Stokes (RANS) Equations approach and the Large-Eddy Simulation (LES) approach have been submitted. Among these, various chemical mechanisms, soot models, and turbulence-chemistry interaction (TCI) methodologies have been considered.

The Development of an Ignition Delay Correlation for PRF Fuel Blends from PRF0 (n-heptane) to PRF100 (isooctane)
SAE paper 2016-01-0551, 2016. DelVescovo, D., Kokjohn, S.L., and Rolf Reitz
Models for SI Combustion and Emissions (Part 3A)
(Session Code: PFL110)
Room 412B,
3:30 p.m.

A correlation was developed to predict the ignition delay of PRF blends at a wide range of engine-relevant operating conditions. Constant volume simulations were performed using Cantera coupled with a reduced reaction mechanism at a range of initial temperatures from 570-1860K, initial pressures from 10-100atm, oxygen mole percent from 12.6% to 21%, equivalence ratios from 0.30-1.5, and PRF blends from PRF0 to PRF100. In total, 6,480 independent ignition delay simulations were performed.

The correlation utilizes the traditional Arrhenius formulation; with equivalence ratio (φ), pressure (p), and oxygen mole percentage (xo2) dependencies. The exponents α, β, and γ were fitted to a third order polynomial with respect to temperature with an exponential roll-off to a constant value at low temperatures to capture the behavior expressed by the reaction mechanism. The location and rate of the roll-off functions were modified by linear functions of PRF. The activation energy term, λ is expressed as a combination of a third and second order polynomial with respect to temperature with an exponential roll-off function whose location and rate varied with a second order function with respect to PRF to capture the differences between PRF blends in the NTC region. The resulting correlation contains 41 constants with an average standard deviation of ±24% compared to the reaction mechanism values in the range of interest to engine applications.

Auto-ignition predictions were calculated using the Livengood-Wu auto-ignition integral and compared to experimental HCCI heavy-duty engine data. The predictions were calculated using two different methods, the first utilized the ignition delay correlation developed in this work, and the second method utilized linear interpolation between relevant points in a large matrix populated by 58,443 ignition delay values. Both predictions showed good agreement (±1.5 °CA) with the start of combustion from the HCCI engine data under the operating conditions tested.

Wednesday, April 13

Investigation of the Combustion Front Structure during Homogeneous Charge Compression Ignition Combustion via Laser Rayleigh Scattering Thermometry
SAE Paper 2016-01-0746 Matthew Blessinger, Jaal Ghandhi
Homogeneous Charge Compression Ignition, HCCI (Part 1 of 2) Room 411A,
8:00 a.m.

While assumed to be successive auto-ignition, the combustion propagation mechanism of low-temperature combustion regimes such as homogeneous charge compression ignition (HCCI) has not been conclusively demonstrated. To assist in addressing this question, the combustion front during HCCI combustion was investigated using planar laser Rayleigh scattering thermometry. The experiments were performed in a small-bore, high-speed engine with Bowditch-style optical access that was uniformly fueled by gaseous dimethyl ether. Image processing was seriously hindered by striations induced by beam steering at the entrance window, which rendered high-pressure HCCI data unusable. The Rayleigh scattering intensity, which is proportional to the local temperature, was obtained by correcting the raw images for camera bias, dark current, and stray scattered laser light; the Rayleigh scattering equation with corrections for the reactant-to-product cross-section change was used to convert the scattering intensity to temperature. Ethylene-fueled spark-ignition (SI) experiments, known to be flame-propagation, provided a baseline for comparison. Qualitative analysis of the SI and HCCI images confirmed previously known combustion aspects: a single wrinkled laminar flame front for SI and multiple ignition sites for HCCI. The relative intensity changes, i.e. temperature changes, across the combustion front were approximately 50% greater for SI than HCCI, which is consistent with the low-temperature nature of HCCI combustion. Broad areas with an intermediate temperature level were also captured in the HCCI images suggesting partial combustion. SI combustion fronts had sharp intensity gradients indicating thin flame fronts. HCCI combustion front intensity gradients varied widely, but some gradients were as thin as SI.

Exploring the Role of Reactivity Gradients in Direct Dual Fuel Stratification
SAE paper 2016-01-0774, 2016. Wissink, M., and Reitz, R.D.
RCCI and Dual-Fuel Low Temperature Combustion (Part 1 of 2)
(Session Code: PFL262)
Room 420A,
8:30 a.m.

Low-temperature combustion (LTC) strategies have been an active area of research due to their ability to achieve high thermal efficiency while avoiding the formation of NOx and particulate matter. One of the largest challenges with LTC is the relative lack of authority over the heat release rate profile, which, depending on the particular injection strategy, either limits the maximum attainable load, or creates a tradeoff between noise and efficiency at high load conditions. We have shown previously that control over heat release can be dramatically improved through a combination of reactivity stratification in the premixed charge and a diffusion-limited injection that occurs after the conclusion of the low-temperature heat release, in a strategy called direct dual fuel stratification (DDFS). This paper will focus on the role the of the reactivity gradients in the premixed charge, which are achieved by the relatively early injection of gasoline and the relatively late injection of diesel. Three regimes were identified for the diesel injection timing: premixed, reactivity-controlled, and diffusion-limited, with the reactivity-controlled regime being observed to offer superior control of combustion phasing and noise, while also having the best emissions and efficiency. It was also observed that increasing the energy fraction of the diesel fuel while in the reactivity-controlled regime resulted in increased efficiency with decreased peak heat release rate and noise by simultaneously advancing combustion phasing and increasing combustion duration, which is a method of control unique to the DDFS strategy.

Analysis of Natural Gas Dual-Fuel Combustion Regimes in a Heavy-Duty Engine
16PFL-0943 (Oral only) Walker, N.R., Kokjohn, S.L., and Reitz, R.D.
RCCI and Dual-Fuel Low Temperature Combustion (Part 1 of 2)
(Session Code: PFL262)
Room 420A,
10:00 a.m.

Natural gas is poised to become the next-generation primary energy feedstock critical to energy conversion and power generation on both domestic and global scales. As a result, recent research on the use of natural gas fuels in internal combustion engines has been extended to investigate the use of natural gas for advanced dual-fuel combustion strategies including diesel pilot ignition (DPI) and reactivity controlled compression ignition (RCCI). Recent studies have proven natural gas to be a beneficial premixed fuel allowing improved engine load capabilities and thermal efficiency improvements for advanced dual-fuel combustion strategies. Additional work focusing on the influence of equivalence ratio and dual-fuel ratio have indicated that both engine operating parameters strongly impact the closed-cycle combustion performance of both the DPI and RCCI combustion regimes while utilizing methane as the premixed fuel and F76 as the direct-injected fuel. In the experiments, the global equivalence ratio was varied over the range of 0.29-0.6; the dual-fuel ratio was varied between 0.86, 0.90, and 0.95, where the dual-fuel ratio is defined as the ratio of premixed fuel to direct-injected fuel on an energy basis. The results have been quantified with respect to three important areas of engine performance: emissions, efficiency, and noise. The ensuing analysis shows that both combustion regimes respond inversely proportional to dual-fuel ratio in terms of emissions and efficiency; however, the response of combustion noise is directly proportional to dual-fuel ratio for both DPI and RCCI combustion regimes. With respect to equivalence ratio, the DPI combustion regime was observed to achieve higher overall performance at richer equivalence ratios and higher dual-fuel ratios, whereas the RCCI combustion regime was observed to attain higher overall performance at leaner equivalence ratios and lower dual-fuel ratios. Based on the results of the experimental work, it is concluded that each natural gas dual-fuel combustion regime attains peak performance in different parts of the equivalence ratio/dual-fuel ratio map, from which arise a variety of implications for engine systems design of advanced dual-fuel combustion engines.

Investigating Air Handling Requirements of High Load Low Speed Reactivity Controlled Compression Ignition (RCCI) Combustion SAE Paper 2016-01-0782. Kavuri, C. Kokjohn, S.L., Reitz, R.D. RCCI and Dual-Fuel Low Temperature Combustion (Part 2 of 2)
(Session Code: PFL 262)
Room 420A,
2:00 p.m.

Past research has shown that reactivity controlled compression ignition (RCCI) combustion offers efficiency and NOx and soot advantages over conventional diesel combustion at mid load conditions. However, at high load and low speed conditions, the chemistry timescale of the fuel shortens and the engine timescale lengthens. This mismatch in timescales makes operation at high load and low speed conditions difficult. High levels of exhaust gas recirculation (EGR) can be used to extend the chemistry timescales; however, this comes at the penalty of increased pumping losses. In the present study, targeting the high load - low speed regime, computational optimizations of RCCI combustion were performed at 20 bar gross indicated mean effective pressure (IMEP) and 1300 rev/min. The two fuels used for the study were gasoline (low reactivity) and diesel (high reactivity). The effects of intake pressure and EGR on combustion and emissions were studied using a full factorial design of experiments of genetic algorithm optimizations. The optimizations were setup for three values of EGR (30%, 45% and 55%) and equivalence ratios (0.8, 0.9 and 1.0). The results showed that gross indicated efficiency (GIE) increases with boost and EGR. A high gasoline percent (> 93.5%) has to be maintained across the range of boost and EGR. Approximately 50% of the gasoline was premixed at the lower EGR's and this percentage increases with increasing EGR to maximize the efficiency. Direct injected gasoline, is injected post TDC at low EGR's and high boost pressures to control the peak pressure rise rates and is advanced as the EGR increases. The start of injection (SOI) of diesel is early at low EGR's to avoid stoichiometric, high temperature combustion and control the NOx emissions, and is brought closer to TDC as EGR increases. The pumping loop work for each case was estimated using a thermodynamic model prepared in CANTERA and the net indicated efficiency (NIE) for each case was calculated. The results showed that NIE had similar trends as GIE, but increasing the EGR beyond 55%, caused the NIE to decrease due to a high pumping loop penalty, indicating an optimum at this EGR and boost pressure.

An efficient level-set flame propagation model for hybrid unstructured grids using the G-Equation
SAE paper 2016-01-0582. Perini, F., Ra, Y., Hiraoka, K., Nomura, K., Yuuki, A., Oda, Y., Rutland, C.J., and Reitz, R.D.
Multi-Dimensional Engine Modeling (Part 2 of 4)
(Session Code: PFL120)
Room 410B,
3:30 p.m.

Cycle-to-cycle variability in gas-fueled large-bore spark ignition engines with pre-chamber ignition is crucial to their controllability when advanced combustion strategies are operated. Geometric design of the pre-chamber and pre-chamber gas feeding assembly is a key parameter for optimizing fuel-air mixing prior to ignition. Computational fluid dynamics can speed up the design process provided that 1) the reliability of the results is not affected by poor meshing and 2) the time cost of the meshing process for many geometries does not negatively compensate for the advantages of running a computer simulation. In this work, a flame propagation model that could run with arbitrary hybrid meshes was developed and coupled with the KIVA4-MHI CFD solver, in order to address these aims. The solver follows the G-Equation level-set method for turbulent flame propagation by Tan and Reitz, and employs improved numerics to handle meshes featuring different cell types such as hexahedra, tetrahedra, square pyramids and triangular prisms. Detailed reaction kinetics from the SpeedCHEM solver are used to map laminar flame speed, as well as non-equilibrium composition downstream and upstream of the flame surface. A generalized least-squares gradient reconstruction algorithm is employed to evaluate spatial derivatives with arbitrary node and cell connectivities, instead of the original WENO scheme proposed by Tan and Reitz. Finally, a new, extended version of the “marching cubes” algorithm for iso-surface tracking was developed and implemented for all four employed cell types. The solver was tested across different cell types and cell resolutions by simulating spherical ignition in a simple cylindrical combustion chamber. Validation was performed against experimental measurements of torch jet ignition and of pre-chamber gaseous fuel combustion in a large bore engine, with different grid resolutions.

Numerical Study of the RCCI and HCCI Combustion Processes Fuelled with Methanol, Ethanol, Butanol and Diesel
SAE paper 2016-01-0777. Zou, X., Wang, H., Zheng, Z., Reitz, R.D., and Yao , M.
RCCI and Dual-Fuel Low Temperature Combustion (Part 2 of 2)
(Session Code: PFL262)
Room 415B,
3:30 p.m.

In the current study, numerical study has been conducted to investigate the RCCI and HCCI combustion and emissions characteristics with various fuels, include methanol, ethanol, butanol and gasoline, which were used as the low reactivity fuels, and the diesel fuel was used as the higher reactivity fuel. A reduced Primary Reference Fuel (PRF)-alcohols chemical kinetic mechanism was coupled with a computational fluid dynamic (CFD) code to predict the RCCI and HCCI combustion and emission characteristics of these fuels under various operating conditions. This study shows that due to a combination of the lower reactivity and higher enthalpy of vaporization of alcohol fuels compared to gasoline-diesel, significantly higher quantities of diesel fuel were required to maintain optimal combustion phasing with the alcohol-diesel fuel blends. The simulations were also extended to model the homogeneous charge compression ignition (HCCI) combustion under the same operating conditions with these fuels. Detailed comparisons between HCCI and RCCI have been conducted and the results show that although comparable performance can be obtained with HCCI under low to medium load conditions, RCCI shows advantages under higher load conditions.

Thursday, April 14

Tools for Comparing Knock Signals from Multiple Transducer Types
oral only. Arsham J. Shahlari, Jaal Ghandhi
Powertrain Actuators and Sensors (Part 2 of 2) Room 415B,
3:00 p.m.

Vibration signals captured by accelerometers are the primary knock-detection tools for production engines, while in-cylinder pressure signals are considered the most accurate tool for studying knock. However, in-cylinder pressure transducers are generally too expensive for use on production engines. Hence, the comparison of these two knock detection tools is of interest. Magnitude Squared Coherence (MSC), is a measure of interdependence between two signals; it is a function of frequency and at each frequency it is a real number between zero and one. This signal-processing tool is commonly used to serve two main purposes. First, it is a measure of how strongly the two signals are related through a linear time invariant (LTI) system. Second, it is an indicator of the frequencies that are common between two signals, i.e., if MSC between two signals at a certain frequency is close to unity, the total interdependence between the two signals at that frequency is high (similarly, if MSC between the two signals is close to zero at a certain frequency, the total interdependence between the two signals is low at that frequency). The latter purpose is of interest in this study for a comparison between in-cylinder pressure signal and accelerometer signals from an SI engine under knocking conditions. Results are presented for data acquired in an SI engine with a wide range of knock intensity. The interdependence between the pressure signal and the accelerometer signal were significantly higher at chamber resonance frequencies than at other frequencies. Furthermore, the interdependence between the pressure signal and the accelerometer signal at chamber resonance frequencies were significantly higher for heavily knocking conditions than conditions with mild or no knock. The MSC calculated for heavily knocking conditions reached values higher than 0.9 at knock frequencies.

Effects of Numerical Schemes on Large Eddy Simulation of Turbulent Planar Gas Jet and Diesel Spray
SAE Paper 2016-01-0866. Tsang, C.W., Rutland, C.J.
Fuel Injection and Sprays (Part 5 of 5)
(Session Code: PFL320)
Room 414A,
2:00 p.m.

Three time integration schemes and four finite volume interpolation schemes for the convection term in momentum equation were tested under turbulent planar gas jet and Sandia non-reacting vaporizing Spray-H cases. The three time integration schemes are the first-order Euler implicit scheme, the second-order backward scheme, and the second-order Crank-Nicolson scheme. The four spatial interpolation schemes are cubic central, linear central, upwind, and vanLeer schemes. Velocity magnitude contour, centerline and radial mean velocity and Reynolds stress profiles for the planar turbulent gas jet case, and fuel vapor contour and liquid and vapor penetrations for the Diesel spray case predicted by the different numerical schemes were compared. The sensitivity of the numerical schemes to mesh resolution was also investigated. The non-viscosity based dynamic structure subgrid model was used. The numerical tool used in this study was OpenFOAM. Results showed that the first-order Euler implicit schemes predict wider range of length scales as seen in velocity and fuel vapor contour plots, while the second-order time integration schemes predict more small scale structures. The non-monotonic central schemes show less grid sensitivity than the monotonic schemes. Overall the cubic central scheme with the Euler implicit scheme gives the best performance. The first-order upwind scheme gave poor results, so it should be avoided to use in simulating high-pressure fuel injection process.