Empowering Simulations with Supercomputing
The U.S. DOE has awarded supercomputing time to three biofuel-related projects through its Innovative and Novel Computational Impact on Theory and Experiment program. The awarded time will allow researchers to use computer simulations to perform virtual experiments that would otherwise be impossible or impractical to perform.
One of these three teams will be performing complex simulations to explore how biojet fuel performs in aircraft engines, as part of a project titled, “Large Eddy Simulation of Two-Phase Flow Combustion in Gas Turbines.” The project, led by Thierry Poinsot, a researcher at the European research institution CERFACS, has been awarded 10 million processor hours. According to Poinsot, the supercomputing time his team was awarded will be used to simulate the performance of existing aircraft and helicopter engines; specifically the conversion channels of these engines. “This is the place where the kerosene and the air are mixed and burned to get the power, it is really the heart of the engine,” he says. Existing aircraft engines have been optimized to effectively and efficiently burn kerosene, not biofuels. Before changing to a new fuel source, such as biojet fuel, it is important to understand how the engine will react. This is because any change in fuel can affect engine stability, such as oscillation, Poinsot says, which could lead to dangerous in-flight problems. Simulating biobased fuel use in existing aircraft engines with the DOE’s supercomputers will allow Poinsot and his team to identify possible problems in less risky conditions.
“We want to study combustion stabilities for normal fuels, the ones we have today,” Poinsot says. “Then we plan to use biofuels to see if they could lead to problems in terms of instabilities. That is the ultimate goal of the project.”
Although the biofuels component of the project is new, Poinsot says his team has been studying aircraft engines through simulations for several years. He also notes that he expects this line of study to continue well into the future. “Demand for these studies is huge in Europe at the moment, and it’s about the same in the states actually,” he says. “Every company is looking for a way to predict engine performance.”
In addition to publishing the results of the research in scientific journals, Poinsot says he expects the outcomes to provide real and immediate benefits to industry. “Since we are computing real engines, we hope to be able to give some guidance to companies working on these engines, to try to avoid instability,” he says. “That is the goal of the program of course, to try to control these things at the design stage.”
Biofuels projects at Oak Ridge National Laboratory and the University of Washington have also been awarded supercomputing hours. An ORNL project led by researcher Jeremy Smith, titled, “Cellulosic Ethanol: Simulation of Multicomponent Biomass System,” was awarded 30 million processor hours. The project aims to provide simulation models of biomass and biomass enzyme interactions, with the ultimate goal of allowing for improved second-generation biofuel yields. The University of Washington project, titled, “Towards Breakthroughs in Protein Structure and Design,” has also been awarded 30 million process hours. The project, led by UW researcher David Baker, will include simulations of “de-novo” enzyme design for fixation of carbon dioxide to create biofuels.