6/2/2014 – 2/5/2015
Over several months at the CRF, postdoc Brian Peterson investigated direct numerical simulation (NVO) strategies to enable improved low-load operation for low-temperature gasoline combustion. NVO involves modifying valve timings to first trap exhaust gases in the cylinder and then recompress them during the latter half of the exhaust stroke. The injection of a small amount of fuel during recompression releases a small amount of heat; the fuel is consequently reformed into partial combustion products. The intake charge heating and additional chemical species that occur under these conditions provide improved combustion stability. Brian studied a potentially beneficial chemical pathway that results from thermal decomposition of the injected fuel into more reactive components.
To this end, Brian used a custom dump-valve assembly housed in one of two available spark plug ports to sample cylinder contents shortly after the NVO reactions under a range of injected-fuel quantities (0–10 mg/injection) and bulk gas oxygen concentrations (2.5%–7%). Samples were then speciated by an in-house gas chromatograph. Combining these results with complementary chemical modeling, Brian showed that reactive species concentrations increased as the oxidizer fraction fell and the equivalence ratio increased. These results were presented at the SAE World Congress meeting in April 2015.
4/30/2012 – 2/26/2015
Sgouria (Iro) Lyra made several contributions to turbulent combustion while completing her postdoctoral appointment at the CRF, where she worked with Jackie Chen and Hemanth Kolla on several direct numerical simulation (DNS) configurations. For example, she identified the effects of turbulence, nonadiabaticity, and composition stratification on the flame structure and local dynamics of partial extinction and re-intensification in a study of a highly turbulent lean premixed hydrogen-air counterflow flame with product stratification. She worked with modelers to demonstrate good agreement between the velocity and scalar statistics in a posteriori comparisons of a one-dimensional turbulence model (modified for the premixed counterflow configuration) with the DNS.
In addition, in studying the structure of hydrogen-rich transverse jets in a vitiated turbulent flow, Sgouria established the statistical similarity between experimental and computational flame and flow structures and then extracted vorticity spectra from the windward shear layer in the DNS. This work showed the influence of heat release on the stability characteristics of the windward shear layer, which in turn impacts the jet/crossflow mixing of fuel and oxidizer. Finally, Sgouria performed DNS of a lean premixed flame of n-butane-air in intense turbulence using a reduced lumped kinetic model describing fuel cracking for high-temperature combustion of practical engine fuels, an effort that demonstrated that flame features controlled by the reaction zones are insensitive to the kinetic lumping.
With her strong computational and statistical background, Sgouria recently started a new position in the financial industry in London.
Riccardo Malpica Galassi
2/1/2015 – 2/28/2015
Riccardo Galassi, a PhD student at the Sapienza University in Rome, Italy, visited the CRF this February as part of an ongoing collaboration with his advisor Prof. Mauro Valorani, on chemical model reduction under uncertainty. During this visit, Riccardo worked on code development and on computations for analysis and reduction of an n-butane chemical mechanism with given uncertainty in reaction rate constants. Now back in Rome, he is advancing work on this area as part of his PhD research and continues to collaborate with the CRF remotely.