• Proceedings of the SAREK Conference
• /
• 2006.11a
• /
• pp.84-84
• /
• 2006

• Journal of the Korean Society of Combustion
• /
• v.22 no.3
• /
• pp.29-34
• /
• 2017

The multi-environment probability density function model has been applied to simulate a turbulent premixed flame in a swirl combustor. To realistically account for the unsteady flow motion inside the combustor, the formulations are derived for the large eddy simulation. The Flamelet generated manifolds is utilized to simplify a multi-dimensional composition space with reasonable accuracy. The sub grid scale mixing is modeled by the interaction by exchange with the mean mixing model. To validate the present approach, the simulation results are compared with experimental data in terms of mean velocity, temperature, and species mass fractions.

• Proceedings of the Korean Society of Propulsion Engineers Conference
• /
• 2007.04a
• /
• pp.71-74
• /
• 2007

Large Eddy Simulation(LES) has been conducted to insight interaction effects of turbulent flow and chemical reaction of a lean-Premixed swirl combustor. The unsteady turbulent flame is carefully simulated so that the motion of flow and flame can be characterized in detail. Fuel lumps escaping from the primary combustion zone move downstream and consequently produce local hot spots conveying large vortical structures in the azimuthal direction. The correlation between pressure oscillation and unsteady heat release is examined by the spatial and temporal Rayleigh parameter.

• Journal of the Korean Society of Combustion
• /
• v.4 no.2
• /
• pp.23-33
• /
• 1999

A turbulent non-premixed ethylene flame, which was set up in a vertical wind tunnel, was examined to understand the effect of turbulent mixing on formations of soot and gaseous species in the flames. Temperature and velocity profiles were measured using uncoated thermocouples and LDV system. Gaseous samples were withdrawn by using a water cooled stainless iso-kinetic gas sampling probe. The samples for inorganic compounds and light hydrocarbons were collected with sampling bottles and were analyzed by a gas chromatography. The samples for aromatic hydrocarbons were collected on a sorbent tube and were analyzed on a GC/MS system. Some of main results were followed. CO and $CO_2$ were measured relatively in early part of flame and the concentration of CO was greater than that of $CO_2$ all over the early flame region due to the scavenging of the oxidizing species OH by soot particles. Aromatic hydrocarbons were measured at x/D=122 along the radial direction and main important species were benzene, xylene, toluene, styrene, indene, naphthalene. The peak points of these compounds occurred at r/D=0.8 apart from the center of flame, around in which the concentration of $C_2H_2$ decayed relatively rapidly from the maximum value.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.9 no.4
• /
• pp.77-84
• /
• 2001

Measurements of local CO, $CO_2$, $O_2$, $N_2$, $C_3$H$_{8}$, NOx concentrations and flame temperature are made for partially premixed flame with and without acoustic excitation. The CO, $CO_2$, $O_2$, $N_2$, and $C$_3$H_8$ concentrations are determined by thermal conductivity detection (Gas-chromatograph) and NOx concentrations are determined by chemiluminescent detection (NOx analyser). To measure local sample concentration, sampling probe was made by quartz with inlet diameter of 0.25mm. In the case of excitation, the visual shape of the flame is changed from laminar flame to turbulent-like flame. The flame length is also reduced, and the flame width becomes broad. In the observation of emission concentration without acoustic excitation, meanwhile, the $CO_2$ and NOx concentrations peak at flame front where the mixture meets with surrounding air, and the CO concentration is increasing at maximum position of CO2 concentration and peaks at the centerline of the burner. In the case of acoustic excitation, the $CO_2$ concentration is widely occurred at nozzle of the burner and is higher relative to unexcitation. The CO concentration is much reduced, but NOx concentration is more increasing. And flame temperature is higher relative to unexcitation. These are caused by enhancing of mixing with surrounding air due to excitation. However, in the case of acoustic excitation, the total NOx concentration is reduced because of the shortened flame length which affects residence time.e.

• Proceedings of the Korean Society of Propulsion Engineers Conference
• /
• 2011.04a
• /
• pp.430-432
• /
• 2011

An experimental study of the flame response in a turbulent premixed combustor has been conducted in order to investigate mechanisms for combustion instabilities in lean premixed gas turbine combustor. A lab-scale combustor and mixing section system were fabricated to measure the flame transfer function. Measurements are made of the velocity fluctuation in the nozzle using hot wire anemometry and of the heat release fluctuation in the combustor using chemiluminescence emission. The results are analyzed to determine the phase and gain of the flame transfer function as a function of the modulation frequency and operating conditions.

• Journal of the Korean Society of Combustion
• /
• v.6 no.2
• /
• pp.65-70
• /
• 2001

To investigate statistics of flamelet in a turbulent premixed flame and to obtain components of their burning velocities in a vertical plane above a pipe-flow burner, the local motion of flamelets with respect to gas are measured by specially arranged diagnostics, composed of an electrostatic probe with four identical sensors and a two-color four-beam LDV system. With this technique, the three-dimensional local flame-front-velocity vector is measured by the electrostatic probe for the first time, and simultaneously the axial and radial components of the local gas-velocity vector in a vertical plane above the vertically oriented burner are measured by the LDV system. Two components of burning velocities of planar flamelets can be obtained from these results and are found to be distributed over different directions and to range in magnitude from nearly zero to a few times the planar, unstrained adiabatic laminar burning velocity measured in the unburnt gas. It may be concluded from these results that turbulence exerts measurable influences on flamelets and causes at least some of them to exhibit increased burning velocity.

• Journal of the Korean Society of Combustion
• /
• v.7 no.2
• /
• pp.62-68
• /
• 2002

To investigate statistics of flamelet in a turbulent premixed flame and to obtain components of their burning velocities in a vertical plane above a pipe-flow burner, the local motion of flamelets with respect to gas are measured by specially arranged diagnostics, composed of an electrostatic probe with four identical sensors and a two-color four-beam LDV system. With this technique, the three-dimensional local flame- front-velocity vector is measured by the electrostatic probe for the first time, and simultaneously the axial and radial components of the local gas-velocity vector in a vertical plane above the vertically oriented burner are measured by the LDV system. Two components of burning velocities of planar flamelets can be obtained from these results and are found to be distributed over different directions and to range in magnitude from nearly zero to a few times the planar, un strained adiabatic laminar burning velocity measured in the unburnt gas. It may be concluded from these results that turbulence exerts measurable influences on flamelets and causes at least some of them to exhibit increased burning velocity.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.14 no.6
• /
• pp.1661-1668
• /
• 1990

In a turbulent premixed flame stabilized by the streamline step, and dominated by a coherent eddy, a flame micro-structure was investigated with analyzing the signals of temperature, the ion current, and schieren phtographs simultaneously. Generally the contours of large scale coherent eddies of schlieren photographs was considered as the flame front, however, the main reaction can be occurred within the eddy as a structure of fine flamelets scale. The surrounding burned gas of flamelets could not propagate to a unburned mixture, obstructing flamelets from propagating to a unburned mixture. Consequently, it could restrain flashback. The main reaction region was found to be located at higher temperature of the burned gas rather than at maximum rms of fluctuating temperature. The peak probability of higher temperature was 6 times greater than that of lower temperature. As it was difficult to infer a flame structure from PDF distribution of the fluctuating temperature in form of bimodal shape, it should be taken into consideration with other informations related to the sensitive flame front, for instance, ion current.

• Journal of Advanced Marine Engineering and Technology
• /
• v.28 no.1
• /
• pp.98-108
• /
• 2004

This study concerns a large eddy simulation (LES) of turbulent flow around a bluff body inside a sudden expansion cylindrical chamber, a configuration which resembles a premixed gas turbine combustor The simulation code is constructed by using the general coordinate system based on the physical contravariant velocity components. The Smagorinsky model is employed and the calculated Reynolds number is 5,000 based on the bulk velocity and the diameter of the inlet pipe. The combined grid technique and cylindrical grid are tested in the numerical simulation with complex geometry. The predicted turbulent statistics are evaluated by comparing with LDV measurement data. The numerical flow visualizations depict the behavior of turbulent mixing process behind the flame holder.