Since many trips are of short duration and include a cold start, automotive engines run quite often without having reached their nominal temperature. This is known to have some major drawbacks, such as increased fuel consumption and higher emissions due to lower efficiency of after-treatment devices, but detailed description of these various effects is seldom presented in the literature. In this article, experiments were conducted on an automotive diesel engine by varying independently the coolant and oil temperatures between 30 °C and 90 °C. Three different operating conditions (low, mid and full load) were studied. The experimental set-up is briefly described as well as the uncertainty of the associated measurements and the development of analytic tools. Then, the evolution of volumetric efficiency, energy share, combustion heat release and exhaust emissions (NOx, particulate matter, CO, unburned hydrocarbons) are described in detail and analysed. Several strategies were considered, including some corrections used in the standard engine control unit to compensate for the low coolant temperature. Some effects of the coolant and oil temperature reduction were clear: increase in friction losses, volumetric efficiency and ignition delay and decrease in NOx emissions. On the contrary, the evolution of brake thermal efficiency, particulate matter, CO and unburned hydrocarbon emission depended on the operating point.

During the last decades, pollutant emissions from internal combustion engines used for transportation have become a major concern. Today, not only steady state emissions but also emissions during transients are regulated and have to be studied in order to be reduced. In this paper, we describe a new methodology developed to measure the instantaneous level of gaseous emissions from a internal combustion engine during transients, using an analyzer initially designed for steady state operation. Moreover, a new phenomenological thermodynamical combustion model is proposed in order to compute emissions during transients. The results of these two methods are compared on various transients. The measurement method seems to give good results (except for hydrocarbon (HC) measurements), as long as the speed and load variations are not too fast. Otherwise, the frequency of the analyzer which was used becomes the limiting factor. The new combustion heat release developed to simulate transients, coupled with an existing two-zone model for emission calculations, leads to satisfactory results for $CO2$ and $O2$ concentrations and $NOx$ emissions. The agreement with measurements is good for smooth transients and seems promising for faster dynamics. The initial goal was reached, although some improvements are still necessary concerning HC measurements and the fastest transients. These results could be helpful when trying to reduce the amount of pollutant emissions at the exhaust during transients, directly or with after treatment devices.

A new quasi-dimensional multi-zone diesel combustion model was developed, using the Musculus–Kattke transient spray model. The later is modified and extended in order to take into account fuel evaporation, to deal with multiple species and to account for temperature changes due to mixing and combustion. The spray model is then coupled with the CHEMKIN code to calculate the chemical reaction rates and to evaluate the ignition delay and the combustion rate. The results are first compared with those from the original Musculus–Kattke model in the case of a non-vaporizing inert spray. Good agreement is found. Then, experiments from the Engine Combustion Network database are used to evaluate the code on diesel-like combustion realized in a combustion vessel. Several tests are first conducted to set the values of the various numerical parameters of the model. Then a detailed analysis of the results is provided for a test case. Finally, two parametric studies are performed, relative to the ambient temperature and the oxygen concentration. The agreement of the model with the experiments is qualitatively good. However, some quantitative discrepancies appear, especially at low temperatures or for diluted combustions.

This paper aims at introducing an original, simulation based, methodology to optimize air charging system of a medium speed engine in order to improve fuel consumption/NO_x trade-off, over a given load profile. A preliminary study is performed on an engine originally equipped with fixed geometry turbocharger (TC) matched at full load. It shows that improving the TC efficiency has a positive impact at matching point but an adverse effect at lower loads. Therefore matching the TC at part load is more promising when the whole engine operating range is considered. However, this requires using a relief system to avoid unacceptable peak firing pressure at full load, and may alter regulated NO_x emission. A new, original methodology is then proposed. Using an advanced existing algorithm, it allows for simultaneous optimization of injection timing, main TC matching and relief system design. The objective is to minimize fuel consumption averaged over several operating points (corresponding to actual load profile). At the same time, NO_x emission, evaluated on regulatory cycle, is kept below a given limit and others reliability constrains are fulfilled. The methodology is assessed with several case studies: – First, various relief systems (blow-off valve, waste gate (WG) and parallel sequential turbocharger (PTC)) are compared on a given load profile – Then, the dimensioning of a TC+WG system is performed for two different load profiles In both cases, the proposed methodology leads to a charging system design customized for a given user profile, which enables a significant fuel consumption reduction.