Flame Speed Measurement Technique
Among other parameters, the HPCR can be used to determine the burning velocity of a wide variety of gaseous and liquid fuels under conditions pertinent to those in a gas turbine combustor.
Figure 1 shows an exploded schematic of the optical combustion measurement section which will be used to carry out the tests.
Figure 1: Exploded Schematic of the Optical Combustion Measurement Section
The section is rated to the following conditions:
Air flow rate: < 5 kg/s
Stationary flame burner techniques have been utilized by many researchers to characterize laminar and turbulent fuel burning rate. This traditional method has recently been developed at Cardiff University to develop and optimize the technique utilizing advanced imaging diagnostic techniques, prior to its employment in the more challenging environment of the High Pressure Combustor Rig.
STATIONARY FLAME BURNER TECHNIQUE
Unlike closed-bomb methods for flame-speed characterisation, the stationary flame technique has the advantage that the burning rate is given explicitly, without the requirement of quantifying the equivalence ratio and then deriving burning rate from the direct measurement of flamespeed. For the stationary flame burner technique the fuel and air is premixed to accurate equivalence ratios through utilisation of flow meters. Though the ultimate aim is to utilise the technique for alternative fuel characterization, only methane has been utilised in these calibration tests. The flame is stabilised at the exit of a simple ‘Bunsen-like’ burner, through balancing the flowrates of the fuel/air mixture with the natural burning rate of the fuel mixture at the particular equivalence ratio. The average burning rate across the flame front is simply calculated from the following Equation:
S = Q/A – Where S is the burning rate, Q is the bulk volumetric flowrate and A is the average area of the flame front.
A wide variety of optical and laser based techniques were found suitable to analyse the seeded flow from the burner including Laser Doppler Anemometry (LDA), Particle Image Velocimetry and high speed imaging. However, the high-speed imaging gave the best method of determining the flame front area, this method is described in detail below.
The pre-mixed fuel and air mixture is seeded with aluminium oxide, with an average particle diameter less than 1 micron, that accurately follows the flow. A laser sheet from a pulsed laser is used to strobe the centre plane of the burner, a camera perpendicular to the plane records the Mie-scattering from the particles, as shown in Figure 2 .
The technique works on the principle that the seed within the reactant zone is denser than the seed in the product zone. Therefore when enough images are recorded and averaged then only the reactant zone can be observed.
The camera used to record the images is a high-speed high-resolution camera that can record up to 65,000 fps, which will be more relevant to turbulent flames, but is utilised here to develop the methodology. First, 800 raw images of the flame are averaged. The average image is then ‘binarised’ using user defined pixel values; the maximum and minimum extent of the reactant zone are first determined, and the actual flame front then assumed to be mid-way between the two. The binary images are processed using an algorithm that accurately calculates the surface area of a 3D shape that is created by revolving the binary image about the burner axis.
Figure 3 below shows the results of a range of equivalence ratios for a laminar natural gas flame at atmospheric pressure with approximately the same volume flowrate. The results clearly show the change in flame front area and hence burning velocity, which is shown plotted against equivalence ratio in Figure 4.
A 16 bar gas turbine the facility (up to 5 kg/s air flow, and 900K) now situated at Gas Turbine Research Centre - GTRC, has been successfully relocated from QinetiQ (Pyestock, near Farnborough, UK) to South Wales.
The facility has been upgraded in several respects, including providing the capability of studying gaseous and liquid fuels, as well as incorporating a new optical section enabling access for a wide range of diagnostic tools including LDA, PIV and PLIF for non-intrusive in-flame measurements. An important component of future work programmes will be turbulent flame speed and burning rate measurements at elevated temperatures and pressures on a wide range of different fuel mixes, both liquid and gaseous.
Tests on a low pressure burner with similar proportions to the burner at the GTRC have shown that the proposed technique of measuring flame speed works in principle, and the data agrees well with data previously published by authors using a variety of techniques.