|Increasing complexity of vehicle design is driving the need for better simulation and more powerful computers. Wagner and Pannala. Click to enlarge.|
The complexity of new and future vehicles—driven by the need for increasing fuel efficiency and decreasing emissions with ever-changing drive-cycle demands and environmental conditions—is adding unprecedented flexibility in design and driving the need for better simulation and more powerful computers, observed Dr. Robert M. Wagner, Director of the Fuels Engines and Emissions Research Center, and Dr. Sreekanth Pannala, Senior Research Staff Member in the Computing and Computational Sciences Directorate at Oak Ridge National Laboratory in a keynote talk at the recent Global Powertrain Conference.
Advances in high performance computing (HPC) resources are leading to a new frontier in engine and vehicle development, Wagner and Pannala suggested, including the ability to produce detailed simulations to generate benchmark data; engineering simulations to explore the design space (e.g., injector optimization at ORNL); and reduced models for design optimization and control strategies. In general, HPC can help solve problems which were once thought unsolvable, they noted.
|3D DNS of auto-ignition with 30-species DME chemistry. (Bansal et al. 2011) Click to enlarge.|
In a 2009 paper describing the use of S3D—a flow solver for performing direct numerical simulation (DNS) of turbulent combustion that was developed at the Combustion Research Facility (CRF) at Sandia National Laboratories (CRF/Sandia)—in terascale direct numerical simulations of turbulent combustion, Chen et al. noted that:
Computational science is paramount to the understanding of underlying processes in internal combustion engines of the future that will utilize non-petroleum-based alternative fuels, including carbon-neutral biofuels, and burn in new combustion regimes that will attain high efficiency while minimizing emissions of particulates and nitrogen oxides.
Next-generation engines will likely operate at higher pressures, with greater amounts of dilution and utilize alternative fuels that exhibit a wide range of chemical and physical properties. Therefore, there is a significant role for high-fidelity simulations, direct numerical simulations (DNS), specifically designed to capture key turbulence-chemistry interactions in these relatively uncharted combustion regimes, and in particular, that can discriminate the effects of differences in fuel properties.—Chen et al.
|Growth of DNS capabilities. Sankaran et al. Click to enlarge.|
DNS is a tool for fundamental studies of the micro-physics of turbulent reacting flows and provides full access to time-resolved fields and physical insight into chemistry turbulence. It is a tool for the development and validation of reduced model descriptions that will then be used used in macro-scale simulations of engineering-level systems interactions.
Combustion currently provides 85% of our nation’s energy needs and will continue to be a predominant source of energy as fuel sources evolve away from traditional fossil fuels. Low emission, low temperature engine concepts of the future operate in regimes where combustion is poorly understood. In an effort to reliably predict efficiency and pollutant emissions for new engines and fuels, computer simulations are used to study fundamental turbulence-chemistry interactions.
Direct Numerical Simulations (DNS) are first principle, high- fidelity computational fluid dynamics simulations in which the reactive compressible Navier-Stokes equations are numerically solved on a computational mesh in which all of the spatial and temporal scales of the turbulence are resolved. In many practical turbulent combustion situations, turbulence strains the flame, causing molecular mixing of reactant streams. With increased mixing, chemical reactions are enhanced and overall efficiency increases up to a point, at which the loss of heat and radicals exceeds their rate of generation due to chemical reaction and the flame extinguishes resulting in increased emissions. Heat-release caused by the chemical reactions creates a complex feedback mechanism, affecting the intensity of the turbulence through density and viscosity changes across the flame.—Bennett et al.
The recent launch of Oak Ridge National Laboratory’s (ORNL) 20 petaflop Titan supercomputer (earlier post), highlights the importance of HPC as a key enabler for accelerating the development of high efficiency engines. One of the first six applications modified to run on the new Titan architecture is S3D.
S3D’s first Titan science problem is a 3‐dimensional DNS of HCCI combustion in a high‐pressure stratified turbulent ethanol/air mixture using detailed chemical kinetics (28 chemical species). Estimated Titan core-hours needed: 128M—an order of magnitude greater that the time required by three other initial problems to be run by other applications, and second only to the 150M core-hours required by a 100-year climate simulation to be run with tropospheric chemistry with 101 constituents at high spatial resolution (1/8 degree).
PreSICE. In 2011, the DOE convened a workshop including 60 US leaders in the engine combustion field from industry, academia, and national laboratories to identify research needs and impacts in Predictive Simulation for Internal Combustion Engines (PreSICE). The workshop focused on two critical areas of advanced simulation, as identified by the US automotive and engine industries:
|Fuel injection involves a cascade of complex processes. PreSICE report. Click to enlarge.|
Fuel spray processes. Fuel sprays set the initial conditions for combustion in essentially all future transportation engines, the workshop noted. However, designers now primarily use empirical methods that limit the efficiency achievable. The workshop identified three primary spray topics as focus areas:
- the fuel delivery system, which includes fuel manifolds and internal injector flow;
- the multi-phase fuel–air mixing in the combustion chamber of the engine, and;
- the heat transfer and fluid interactions with cylinder walls.
Stochastic processes. Current understanding and modeling capability of stochastic processes in engines remains limited and prevents designers from achieving significantly higher fuel economy, the workshop found. To improve this situation, the workshop participants identified three focus areas for stochastic processes:
- improve fundamental understanding that will help to establish and characterize the physical causes of stochastic events;
- develop physics-based simulation models that are accurate and sensitive enough to capture performance-limiting variability, and;
- quantify and manage uncertainty in model parameters and boundary conditions.
|Neroorkar et al. “Simulations and Analysis of Fuel Flow in an Injector Including Transient Needle Effects”, ILASS-Americas 24th Annual Conference on Liquid Atomization and Spray Systems, May 2012. Click to enlarge.|
Improved models and understanding in these areas will allow designers to develop engines with reduced design margins and that operate reliably in more efficient regimes, the workshop report stated. All of these areas require improved basic understanding, high-fidelity model development, and rigorous model validation. Such advances would greatly reduce the uncertainties in current models and improve understanding of sprays and fuel–air mixture preparation that limit the investigation and development of advanced combustion technologies.
Because of their relatively low cost, high performance, and ability to utilize renewable fuels, internal combustion engines—including those in hybrid vehicles—will continue to be critical to our transportation infrastructure for decades. Achievable advances in engine technology can improve the fuel economy of automobiles by over 50% and trucks by over 30%.
Achieving these goals will require the transportation sector to compress its product development cycle for cleaner, more efficient engine technologies by 50% while simultaneously exploring innovative design space. Concurrently, fuels will also be evolving, adding another layer of complexity and further highlighting the need for efficient product development cycles. Current design processes, using “build and test” prototype engineering, will not suffice. Current market penetration of new engine technologies is simply too slow—it must be dramatically accelerated.—PreSICE workshop report
|Model hierarchy proposed by DOE and national laboratories spans fundamentals to full vehicle. Source: Wagner and , prepared in cooperation between ORNL, Sandia National Laboratories, and DOE. Click to enlarge.|
Opportunities. Many concepts for higher efficiency engines historically have been held back by the available technology, Wagner noted in an earlier talk at the US Department of Energy’s (DOE) 2012 Directions in Engine-Efficiency and Emissions Research (DEER) conference. Examples of this include:
Direct Injection Spark Ignition. (e.g., Scussel, Simko, and Wade, “The Ford PROCO Engine Update”, SAE Technical Paper 780699, 1978).
Low Temperature Combustion. (e.g., Najt and Foster, “Compression-Ignited Homogeneous Charge Combustion”, SAE Technical Paper 830264, 1983; Akihama, Takatori, and Inaga, “Mechanism of the smokeless rich Early SI Direct Injection ￼diesel combustion by reducing temperature”, SAE Technical Paper 2001-01-0655, 2001.)
Dual-fuel Combustion. (e.g., Stanglmaier, Ryan, and Souder, “HCCI Operation of a Dual-Fuel Natural Gas Engine for Improved Fuel Efficiency and Ultra-Low NOx Emissions at Low to Moderate Engine Loads”, SAE Technical Paper 2001-01-1897, 2001; Singh, Kong, Reitz, Krishnan, Midkiff, “Modeling and Experiments of Dual-Fuel Engine Combustion and Emissions”, SAE Technical Paper 2004-01-0092, 2004.)
Advancing computational technology—from HPC supercomputers to on-board computers—is now opening unprecedented opportunities in combustion strategy and controls, Wagner and Pannala said. Four years from now, the next generation of supercomputer should arrive (OLCF-4), offering 400+ petaflops, they said, with the first 1 exaflop supercomputer (OLCF-5) following sometime after 2020, introducing the era of exascale computing. (Earlier post.) (Petaflop = 1 quadrillion floating point operations/second; Exaflop = 1 quintillion floating point operations/second.)
We can expect similar growth in computing capability of onboard vehicle hardware, the pair noted.
Simulation efforts fall into three broad areas:
- Predictive combustion: combustion optimization and methods development;
- Full engine simulation: engine system optimization and model-based onboard controls; and
- Full vehicle simulation: technology interactions, component optimization and supervisory controls
Each scale of simulation requires different level of fidelity; an increase in complexity results in an increase in the simulation space and accompanying computational requirements. There is thus a need, Wagner noted in his DEER talk, for faster simulation, faster optimization methods, and reduced models for on-board controls.
|Advanced control systems could enable combustion systems closer to the edge of stability. Wagner 2012. Click to enlarge.|
There is a significant opportunity for delivering the high-efficiency engines required via prediction and control for the forced stabilization of inherently unstable systems. Stability has been and continues to be a roadblock to many advanced combustion implementations.
The current approach, Wagner noted, is to maintain distance from the edge of stability to avoid unintended excursions. However, he said, dynamic instabilities are short-term predictable and conducive to control.
New control opportunities will help optimize the operation of the next generation of high efficiency engines; however, new approaches to calibration of the controls will be necessary for optimization of the next generation of engines, as the control parameter space is already growing rapidly and is expected to continue to do so. Predictive simulation is essential for optimal design and controls, Wagner said.
DOE’s Leadership HPC effort is now being used to investigate instability mechanisms from an engine design perspective, and also to accelerate design optimization. Example ORNL projects in progress with industry include:
Large Infrastructure computing for Multi-cycle Instability and Transient Simulations (LIMITS). This project addresses the limits of the fuel economy benefit of dilute combustion, and focuses on stochastic and deterministic processes that drive cycle-to-cycle instabilities.
Injector design optimization. This project seeks to improve understanding and design optimization of fuel injector hole patterns for improved engine efficiency and reduced emissions.
Robert Wagner, Sreekanth Pannala (2012) High Performance Computing: Accelerating the Development of High Efficiency Engines. Keynote at Global Powertrain Congress 2012
Robert Wagner (2012) Enabling the Next Generation of High Efficiency Engines. 2012 DEER
DOE Workshop on the Grand Challenges of Advanced Computing for Energy Innovation (31 July–2 August 2012)
Ramanan Sankaran, Jacqueline H. Chen, Ray Grout (2012) Direct Numerical Simulation of Turbulent Combustion: Fundamental Science towards Predictive Models.
Evatt R. Hawkes, Obulesu Chatakonda, Hemanth Kolla, Alan R. Kerstein, Jacqueline H. Chen (2012) A petascale direct numerical simulation study of the modelling of flame wrinkling for large-eddy simulations in intense turbulence, Combustion and Flame, Volume 159, Issue 8 Pages 2690-2703 doi: 10.1016/j.combustflame.2011.11.020
Ray W. Grout (2012) S3D Direct Numerical Simulation — Preparation for the 10–100 PF era.
Bronson Messer (2012) Early Science at the OLCF on Titan
Janine Bennett, Vaidyanathan Krishnamoorthy, Shusen Liu, Ray Grout, Evatt Hawkes, Jacqueline Chen, Jason Shepherd, Valerio Pascucci, Peer-Timo Bremer (2011) Feature-Based Statistical Analysis of Combustion Simulation Data IEEE Transactions on Visualization and Graphics.
J H Chen et al. (2009) Terascale direct numerical simulations of turbulent combustion using S3D Comput. Sci. Disc. 2 015001 doi: 10.1088/1749-4699/2/1/015001
David Lignell, C. S. Yoo, Jacqueline Chen, Ramanan Sankaran and Mark R. Fahey (2007) S3D: Petascale Combustion Science, Performance, and Optimization