نشریه سوخت و احتراق
سال ششم شماره 1 (پیاپی 11، بهار و تابستان 1392)

HCCI combustion engines due to their low fuel consumption and pollution have been studied by many researchers in recent years. Along with several experimental researches performed on these engines, mathematical modeling has also been utilized as a tool for prediction of their combustion, performance and pollution behaviours. Although use of detailed chemical kinetic mechanisms in the combustion models will result in a proper prediction of the above-mentioned behaviours, a significant increase in the computational time particularly for the more accurate combustion models will also occur. In order to overcome this challenge, mechanism reduction process in HCCI combustion engine conditions has been considered in the current study. In this direction, an optimized single-zone combustion model has been used as a necessary tool for simulation of combustion. Based on the studies performed in the current work, the Directed Relation Graph with Error Propagation (DRGEP) method, thanks to its low computational time while having high accuracy, has been selected as a proper method for the mechanism reduction process. Additionaly, instead of the traditional single-stage reduction approach, a novel approach based on gradual reduction with error propagation has been used. Therefore, the mechanism reduction process has been executed fully automatically in a wide interval of operating conditions of HCCI engine on GRI-Mech3.0 mechanism for simulation of combustion of natural gas, and on the Golovichev’s mechanism for simulation of combustion of n-heptane. At the end of this process, the sizes of these two mechanisms were reduced from 53 species and 325 reactions, and 57 species and 290 reactions to 24 species and 95 reactions, and 42 species and 146 reactions, respectively. The errors caused by reduction, however, were remained less than 1 percent at all conditions.

Large eddy simulation (LES) is performed to investigate oscillation behavior and time-averaged values of the fire behind a vertical wall. One-Equation sub-grid scale (SGS) model is used for turbulent closure. The combustion is assumed to be non-premixed. Also, modified eddy dissipation concept (EDC) and discrete ordinate methods (DOM) are used for incorporating combustion and radiation, respectively. The models are applied for the total heat release rate (HRR) of 36 kW and 54 kW. The numerical results are validated against experimental measurements. The time-averaged temperature and velocity are in a good agreement with the experiments. Generally, the accuracy of the predictions reduces considerably near the wall surface. The oscillating behaviors of the simulated quantities show the three-dimensional and asymmetric nature of the induced flow. In addition, the amplitude of the vertical velocity fluctuation reduces with increase in normal distance from the vertical wall. This fluctuation increases with increase in the height from the area where the fuel is burning.

In this study, a numerical model has been developed in AVL FIRE software to investigate combustion and injection inside the cylinder of a direct injection spark ignition CNG engine. In this regard two parts have been taken into consideration. In the first part of the study, gas injection via a single-hole injector into the cylinder with five different piston head geometries has been investigated. Using quantitative and qualitative representations of the results, the suitable combustion chamber geometry for DI application has been discussed. In the second part, combustion studies have been performed based on the selected geometry from the first part. Spark plug location and ignition timing have been investigated as two of the most important combustion variables. Five different configurations for the spark plug have been taken into consideration and the idea of using two spark plugs has been also tested, where it showed the best combustion characteristics. The spark timing has also been studied based on the selected configuration. The results show that ignition timing should be at 50 degrees before top dead center in order to have the best combustion characteristics.

Availability analysis provides useful information for optimizing the systems. In the present investigation, a multi-zone combustion model is developed to simulate diesel engine cycle.Then, the governing equations of availability analysis are applied in this model.The concept of chemical equilibrium based on Olikara and Borman method is used to calculate the concentration of combustion products.Chemical availability is considered as oxidation, reduction, and diffusion availabilities.The availability analysis is applied to the engine from the inlet valve closing (IVC) until exhaust valve opening (EVO).The effect of fuel injection timing is investigated by various availability terms. The results indicate that advancing time of injection increases work and heat transfer availabilities, but decreases irreversibility.

Over the recent decades, spray systems have been widely utilized in a variety of applications within the areas of engineering. In this regard, studying the evaporative behavior of droplets in spray can give us a good view point. One of the most common assumptions in droplets’ evaporation is to consider the Lewis number to be unity. In this paper the correctness of this assumption has been evaluated for water and n-Heptane single droplets. The results for droplets’ evaporation trend have been compared with some relevant experimental and numerical data. Evaporation time and droplets’ temperature have been investigated in both conditions. According to obtained results, although the unit Lewis number is not a good assumption for studying n-Heptane droplet evaporation, this assumption gives acceptable results for predicting the water droplets evaporation. Moreover, by assuming the unity Lewis number for water droplet, the calculated temperature of the droplet is close to wet bulb temperature of air.

Thermal characteristics of a porous radiant burner (PRB) including gas radiation effects are identified in the present study. This system operates on the basis of effective energy conversion method between flowing gas enthalpy and thermal radiation. In the PRB, the gas and solid phases are considered in non-local thermal equilibrium and combustion is modeled by considering a non-uniform heat generation zone. The porous media as a gray body, in addition to its convective heat exchange with the gas, can absorb, emit and scatter thermal radiation. In order to accurately determine the thermal characteristics of PRB, the gas radiation is also taken into account and a theoretical analysis is conducted for a two dimensional model, where convection, conduction and radiation take place simultaneously in porous medium and gas flow. Discrete ordinates method is used to obtain the distribution of radiative heat flux in the porous media and the coupled energy equations for the gas and porous medium in steady condition are solved numerically. The crucial influence of gas radiation effect on the system's performance is thoroughly explored. A comparison has been made between the present numerical results and those reported by other investigators to validate the simulation for the gas radiative effect.

In this paper, a turbulent lifted diffusion jet flame is studied using the composition probability density function approach in a two-domensional domain with detailed chemistry. The main purpose is to investigate the effect of density variations on scalar fields and lift-off height. For this purpose, the standard k- model as well as a modified turbulence model for variable density conditions are employed to investigate the impact of turbulence models on the flame behavior and the place of stabilization. The results show that the best agreement between the numerical results and measurements is achieved using the modified turbulence model. A comparison between the numerical results and measurements shows that the standard k- model over-predicts the spreading and decay rates in the jet. Using the velocity-pressure gradient term in the modified turbulence model resolves the relevant problem to a great extent and leads to better results than those of the standard k- model.