skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Turbulent premixed combustion; Further discussions on the scales of fluctuations

Abstract

The prediction of turbulent combustion is classically performed by numerical integration of modeled equations. For each existing approach, a crucial quantity is a time scale for the destruction of temperature (or concentration) fluctuations. Usually, this time scale is simply taken as proportional to the corresponding time scale for velocity fluctuations, but the question of the influence of the reaction on this scale arises, and has already been discussed. The authors discuss this point in the particular case of wrinkled premixed flames. Algebraic closure formulas, as well as closures based on the so-called {epsilon}{sub y} equation, or on the equation for the flame surface by unit of volume of Marble and Broadwell, are proposed and discussed. It is found, in particular, that in the case of wrinkled flames with infinite Damkohler and Reynolds numbers the reaction does not play any direct role on the time scale.

Authors:
 [1]
  1. (Laboratoire de Thermodynamique, UA CNRS No. 230, Faculte des Science et des Techniques de Rouen, 76130 Mont-Saint Aignan (FR))
Publication Date:
OSTI Identifier:
6806603
Alternate Identifier(s):
OSTI ID: 6806603
Resource Type:
Journal Article
Resource Relation:
Journal Name: Combustion and Flame; (USA); Journal Volume: 80:3
Country of Publication:
United States
Language:
English
Subject:
37 INORGANIC, ORGANIC, PHYSICAL AND ANALYTICAL CHEMISTRY; 42 ENGINEERING; FLAMES; COMBUSTION KINETICS; FUELS; MIXTURES; TURBULENT FLOW; EQUATIONS; REYNOLDS NUMBER; TEMPERATURE GRADIENTS; CHEMICAL REACTION KINETICS; DISPERSIONS; FLUID FLOW; KINETICS; REACTION KINETICS 400800* -- Combustion, Pyrolysis, & High-Temperature Chemistry; 420400 -- Engineering-- Heat Transfer & Fluid Flow

Citation Formats

Borghi, R. Turbulent premixed combustion; Further discussions on the scales of fluctuations. United States: N. p., 1990. Web. doi:10.1016/0010-2180(90)90106-2.
Borghi, R. Turbulent premixed combustion; Further discussions on the scales of fluctuations. United States. doi:10.1016/0010-2180(90)90106-2.
Borghi, R. Fri . "Turbulent premixed combustion; Further discussions on the scales of fluctuations". United States. doi:10.1016/0010-2180(90)90106-2.
@article{osti_6806603,
title = {Turbulent premixed combustion; Further discussions on the scales of fluctuations},
author = {Borghi, R.},
abstractNote = {The prediction of turbulent combustion is classically performed by numerical integration of modeled equations. For each existing approach, a crucial quantity is a time scale for the destruction of temperature (or concentration) fluctuations. Usually, this time scale is simply taken as proportional to the corresponding time scale for velocity fluctuations, but the question of the influence of the reaction on this scale arises, and has already been discussed. The authors discuss this point in the particular case of wrinkled premixed flames. Algebraic closure formulas, as well as closures based on the so-called {epsilon}{sub y} equation, or on the equation for the flame surface by unit of volume of Marble and Broadwell, are proposed and discussed. It is found, in particular, that in the case of wrinkled flames with infinite Damkohler and Reynolds numbers the reaction does not play any direct role on the time scale.},
doi = {10.1016/0010-2180(90)90106-2},
journal = {Combustion and Flame; (USA)},
number = ,
volume = 80:3,
place = {United States},
year = {Fri Jun 01 00:00:00 EDT 1990},
month = {Fri Jun 01 00:00:00 EDT 1990}
}
  • The authors studied a high velocity premixed turbulent flow of air and methane (..mu.. = 55 m/s, T = 560K, equivalence ratio ..cap alpha.. = 0.8) ignited and stabilized by a parallel flow of hot gases (..mu.. = 110 m/s, T = 2000K). Also studied were the mixing of the two flows without combustion (..cap alpha.. = 0) and a configuration where the combustion was stabilized by a recirculation zone in a downward facing step. In part I, previous experimental results (visualizations, LDV, averaged concentrations) are recalled. Through some characteristic data (turbulence intensity, and profiles of velocity and CH/sub 4/more » and CO concentrations), the structure of the flame and its scales and the increase of turbulence within the flame zone are discussed and compared to other works in turbulent combustion. In Part II, instantaneous measurements of fluctuating temperatures are presented. These measurements have been performed using broadband CARS on N/sub 2/. The CARS system, the choice of the optical arrangement (BOXCARS, parallel polarizations of laser beams), and the data reduction are described. Temperature pdfs were obtained at various locations in the stabilization region of the flame. The comparison between pdfs with and without combustion, the bimodal shape of the pdf in the mean flame zone, and the position of the hot and cold peaks as a function of the location within the flow show that the characteristic chemical time is not infinitely small. This observation confirms that the flame is not purely wrinkled. In the case of the recirculation zone, this zone cannot be considered as a well stirred reactor; the fluctuations of the temperature are bimodal in the mixing layer between the recirculation zone and the external flow.« less
  • No abstract prepared.
  • Direct numerical simulations (DNS) are conducted in 3D to investigate the evolution of flame surface density (FSD) in turbulent premixed combustion. A parametric study is performed with respect to turbulent intensity and Lewis number to investigate all component terms in the FSD transport equation. A higher turbulent intensity leads to a higher turbulent burning velocity due to increased flame area, while the mean consumption speed remains close to the laminar flame speed. A lower Lewis number leads to a higher turbulent burning velocity, with increases in both total flame area and mean consumption speed. There are two source terms tomore » govern FSD: tangential strain and propagation term, given as a product of displacement speed and curvature. The mean strain rate varies linearly with the turbulent intensity, but shows no noticeable dependence on the Lewis number. The correlation between curvature and displacement speed does not depend on the turbulent intensity, but shows significant influence of the Lewis number. The propagation term decreases with increasing turbulent intensity to become a larger negative sink in the rear of flame brush with flame elements of smaller radii of curvature and higher displacement speeds. A lower Lewis number leads to a larger positive propagation term in the front due to an increased displacement speed to produce more flame area through diffusive thermal instability. (author)« less
  • Using hydrogen or carbon dioxide as an additive, we investigate the bending effect of turbulent burning velocities (S{sub T}/S{sub L}) over a wide range of turbulent intensities (u{sup '}/S{sub L}) up to 40 for lean premixed methane combustion at various equivalence ratios ({phi}), where S{sub L} is the laminar burning velocity. Experiments are carried out in a cruciform burner, in which a sizable downward-propagating premixed CH{sub 4}/diluent/air flame interacts with intense isotropic turbulence in the central region without influences of ignition and unwanted turbulence from walls. Simultaneous measurements using the pressure transducer and pairs of ion-probe sensors at various positionsmore » of the burner show that effects of gas velocities and pressure rise due to turbulent combustion on S{sub T} of lean CH{sub 4}/H{sub 2}/air flames can be neglected, confirming the accuracy of the S{sub T} data. Results with increasing hydrogen additions ({delta}=10, 20, and 30% in volume) show that the bending of S{sub T}/S{sub L} vs u{sup '}/S{sub L} plots is diminished when compared to data with {delta}=0, revealing that high reactivity and diffusivity of hydrogen additives help the reaction zone remaining thin even at high u{sup '}/S{sub L}. In contrast, the bending effect is strongly promoted when CO{sub 2} is added due to radiation heat losses. This leads to lower values of S{sub T}/S{sub L} at fixed u{sup '}/S{sub L} and {phi}, where the slope n can change signs from positive to negative at sufficiently large u{sup '}/S{sub L}, suggesting that the reaction zone is no longer thin. All S{sub T} data with various {delta} can be well approximated by a general correlation (S{sub T}-S{sub L})/u{sup '}=0.17Da{sup 0.43}, covering both corrugated flamelet and distributed regimes with very small data scatter, where Da is the turbulent Damkoehler number. These results are useful in better understanding how turbulence and diluents can influence the canonical structures of turbulent premixed flames and thus turbulent burning rates. (author)« less