No abstract

Low-pressure exhaust gas recirculation (LP-EGR) is an EGR configuration in which clean exhaust gas is taken downstream of the turbine and aftertreatment, and then reintroduced upstream of the compressor (1). Employing LP-EGR on Diesel engines can improve fuel economy by reducing pumping losses, lowering intake manifold temperature and facilitating advanced combustion phasing (2, 3). The LP-EGR can also improve compressor and turbine performance by moving their operating points towards higher flow rate and higher efficiency points, which is reflected as a net reduction in pumping losses of the engine. In this study, we focus on effects of introducing LP-EGR on the compressor pressure ratio, and isentropic total-to-total efficiency. The flow field of LP-EGR and air mixing upstream of the compressor as well as the entire compressor stage were studied using a CFD RANS model. The model was validated against turbocharger gas stand measurements. A T-junction mixer was chosen as the design baseline, and various configurations of this mixer were evaluated. The impact of the geometric configuration of the mixer was studied by varying mixing length, EGR jet introduction angle, and EGR-to-air cross section area ratio over a wide range of relevant engine operating conditions. The flow field upstream of the compressor is strongly affected by the dimensionless quantity EGR-to-air momentum ratio. At intermediate momentum ratios, stream-wise counter-rotating vortex pairs (4) are induced in the flow. These vortices can reach the impeller inlet, and depending on vorticity and length scale, perturb the local velocity triangle. At low and high momentum ratios, creeping or impinging jets respectively are formed. In addition prewhirl can be induced by eccentric introduction of EGR. The EGR-induced prewhirl acts similar to an inlet guide vane and can alter the incidence angle at the impeller inlet. The performance of the compressor is altered by the EGR-induced flow field. Compressor pressure ratio is either increased or decreased depending on the direction of EGR-induced prewhirl with eccentric EGR introduction. The compressor efficiency decreases at low flow rates by introduction of concentric EGR due to perturbation of the velocity triangle at the impeller inlet. On the other hand, at low flow rates compressor efficiency can be improved by eccentric EGR introduction, which generates prewhirl in the direction of rotation of the impeller leading to improved incidence angle. The extent to which the compressor is influenced by the EGR-induced flow field is generally reduced by increasing the EGR mixing length, due to viscous damping and breakdown of large-scale EGR-induced vortices. The LP-EGR configuration provides a potential pathway towards improvement of compressor performance, not only by increasing compressor flow rate, but also by manipulation of the flow field. Given that the engine pumping losses are strongly dependent on compressor performance, specifically the compressor efficiency, this study indicates that LP-EGR provides an important path towards reducing pumping loss and improving fuel conversion efficiency.

In recent years, diesel exhaust gas aftertreatment has become a core combustion engine research subject because of both increasingly stringent emission regulations and incentives toward more fuel-efficient propulsion systems. Lean NOX traps (LNT) and selective catalytic reduction (SCR) catalysts represent two viable pathways for the challenging part of exhaust gas aftertreatment of lean burn engines: NOX abatement. It has been found that the combination of LNT and SCR catalysts can yield synergistic effects. Switches in the operation mode of the engine, temporarily enriching the mixture, are required to regenerate the LNT catalyst and produce ammonia for the SCR. This paper describes the design of a catalyst flow reactor that allows studying multi-brick catalyst systems using rapid exhaust gas composition switches and its initial validation. The flow reactor was designed primarily to study the potential of combining different aftertreatment components. It can accommodate two sample bricks at a time in two tube furnaces, which allows for independent temperature control. Moreover, the flow reactor allows for very flexible control of the composition and flow rate of the synthetic exhaust, which is blended using mass flow controllers. By using a two-branch design, very fast switches between two exhaust gas streams, as seen during the regeneration process of a LNT catalyst, are possible. The flow reactor utilizes a variety of gas analyzers, including a 5-Hz FTIR spectrometer, an emissions bench for oxygen and THC, a hydrogen mass spectrometer, and gas chromatographs for HC speciation. An in-house control program allows for data recording, flow reactor control, and highly flexible automation. Additionally, the hardware and software incorporate features to ensure safe testing. The design also has provisions for engine exhaust sampling.

One approach for nitrogen oxides (NOx) emission control of medium duty diesel engines is through the use of a combination Lean NOx Trap and Selective Catalytic Reduction (LNT-SCR) catalyst system. In this system, part of the NOx conversion occurs via an NH3 SCR catalyst that is dependent on the NO2 to NOx ratio of the feed gas with NO2 being a more advantageous oxidizer. One benefit of using this system is the conversion of NO to NO2 over the LNT which increases the NO2:NOx ratio of the feed gas to the SCR catalyst. An experimental study has been performed to investigate the NO2-NH3 reaction for an Fe-based zeolite SCR catalyst using a bench top flow reactor. The increase in NO2 concentration at the inlet of the SCR results in the formation of large quantities of N2O from 200°C to 400°C. Further experiments determined that N2O and NH3 react above 350°C. This has led to a hypothesis that one primary SCR reaction (Slow SCR) can be replaced with two reaction steps featuring NH3, NO2, and N2O. As a result, this paper proposes five NOx reduction reactions as part of a global mechanism, which would account for the observed experimental behavior.