Emission Spectra as a Combustion Diagnostic Tool
Dermeval Carinhana Junior1*, Juliana C. de Oliveira2#, Marcelo G. Destro1, Alberto M. dos Santos1 1
Instituto de Estudos Avançados (IEAv), São José dos Campos, SP, Brasil 2 Universidade Braz Cubas (UBC), Mogi das Cruzes, SP, Brasil *
[email protected]
Abstract In the present work, spontaneous emission spectra of Liquefied Petroleum Gas (LPG) flames were used for combustion diagnostics. CH* and C2* radicals were chosen as flame probe because these species shown a large optical emission in the visible spectrum region. Additionally, the occurrence of these radicals is strongly dependent of the flame composition. In rich flames, for which the equivalence ratio (φ) is higher than 1, the production of C2* is enhanced, while the CH* production is more favored in lean flames (φ < 1). Several flame compositions were studied, with φ values varying from 0.81 up to 2.20. All flames were carried out in under atmospheric pressure conditions. Emission spectra were measured from 400 to 700 nm. In this spectral range, several electronic bands of both CH* and C2* radicals can be observed, but only the (A2∆-X2Π) and (d3Πga3Πu), respectively, were used in this work. Two flame regions were investigated: at 0.5 mm from the top of the burner and at 2.5 mm. The results showed that the ratio CH*/ C2* decreases with φ values following an exponential decay. This decay is more pronounced at 2.5 mm, which suggests that C2* radical presents a broader distribution along the flame axis than CH* species. The existence of a well-established mathematical relation between optical emissions of the studied radicals and the equivalence ratio indicated that this ratio can be used in the active control of flame composition.
Introduction
Spontaneous (or natural) emission in flames is due to chemical reactions that generate chemical species in excited states. This phenomenon, also known as chemiluminescence, has been employed for monitoring the heat oscillations produced in combustion systems, for monitoring fuel/oxidant equivalence ratio (φ), and to supply data for theoretical models, consisting in an important tool for active control in burning processes [1]. Among the several radicals present in flames, CH* and C2* showed the most intense emission in visible spectrum region, which able these radicals for chemiluminescence measurements. CH* most intense emission band system is observed around 431 nm, that corresponds to the (A2∆-X2Π) electronic transition. The bands of C2* radicals, known as Swan system, are observed in 473.7, 516.2 and 563.5 nm, that corresponds to the vibrational bands of the (d3Πg-a3Πu) electronic transition [2]. An important advantage of the use of natural emission in flame measurements is the fact that the radical species emit near the region where they were generated, due to their small emission lifetime. This can be used to relate the information obtained from flame emission spectra to local flame properties as temperature and relative species concentrations [3]. The last one has been used as an important tool for combustion diagnostics, which is the aim of this work.
Experimental Setup
We used a premixed type home-built burner. Flames were produced from a gaseous mixture of LPG/nitrogen/oxygen. LPG used was a propane/n-butane equimolar mixture. The burner supply was controlled by calibrated flow meters. φ values were set from 0.81 up to 2.20. The burner was assembled on a optical translation stages with mobility in the three orthogonal directions (x, y and z). This setup permitted the spectra to be recorded at different heights along the principal flame axis. The light collection was made using a quartz lens, with f = 100 mm and 2” diameter. The flame image was projected in a mask with a 1:1 magnification, where a 1 mm diameter pinhole was placed to allow light from the flame to be introduced in a fiber optics bundle, which was coupled to the spectrometer entrance slit.
The spectrometer was a HR 4000 spectrometer (Ocean Optics) with a fixed 600 lines.mm-1 grating, covering the spectral range from 190 to 1100 nm. Emission spectra were obtained from 400 to 700 nm, at two different heights: 0.5 mm from the top of the burner and at 2.5 mm. A scheme of experimental apparatus is shown in Figure 1. Lens Mask Bundle
Aquisition
Burner
Monochromator N2 O2 LPG
Figure 1: Experimental apparatus used in emission measurements.
.
Results and Discussions
Flame total height increased with the increasing φ values. For large φ values, the higher fuel concentration increases the residence time of the reagents due to the lower reaction rates. In other words, in this condition the flame inner cone becomes depleted in oxidant, resulting in a partial fuel oxidation. Therefore, besides the original fuel excess, intermediate fuels are formed (in this case, CO and H2). Since the outer cone flame is governed by transport processes (diffusive zone), there is a natural increase in the cone height [4]. Figure 2 shows a typical flame spectrum obtained at 0.5 mm and φ = 1.28. As discussed, only the most intense radical bands were considered in the following results.
1.0 0.9 0.8
CH
*
*
C2
I (a.u.)
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 400
450
500
550
600
650
700
λ (nm)
Figure 2: Flame spectrum at 0 mm and φ = 1.28
For each flame height and φ value three different spectra were obtained. Band intensities errors were estimated based on the average of these measurements. Figure 3a and 3b shows the variation of relative CH* and C2* emission as function of flame composition, given by φ values, at 0.5 mm and 2.5 mm respectively.
500
a)
*
CH * C2
400
I (a.u.)
300 200 100 0
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
1.6
1.8
2.0
2.2
φ
b)
*
CH * C2
1000
I(u.a.)
800 600 400 200 0
0.8
1.0
1.2
1.4 φ
Figure 3: Relative radical intensity in function of flame composition. a) 0.5 mm and b) 2.5 mm. In Figure 3 we can observe that relative radical intensities are affected by flame composition. CH* production is higher than C2* up to φ = 1.2~1.3. After this range, the production of C2* is more intense. C2* production species is also more influenced by flame height. At the 2 mm this radical presents a more pronounced increase of the emission than CH* species. The relationship between both radical emission intensities can be observed in Figure 4. This figure shows the exponential decay behavior of the CH*/C2* ratio with increasing φ values. This decay is more pronounced at 2 mm. This indicates that, as discussed, C2* radical shows a broader intensity distribution along the flame axis than CH*. 2.5
0 mm 2 mm
2.0
*
*
CH /C2
1.5 1.0 0.5 0.0
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
φ
Figure 3: CH*/C2* intensity ratio in function of flame composition. a) 0 mm and b) 2 mm.
Conclusions
The existence of a well-known relationship between CH*/C2* ratio as function of φ values indicates that this ratio can be assigned as an active control parameter of combustion systems. As the used bands are the most intense of radical spectrum, a simple optical detection system, like band-pass filters can be used to determined the species ratio and, as consequence, the on-line flame composition.
Acknowledgements #
The authors thank the CNPq for its scholarship - PIBIC Proc. Number 110829/2006-9.
References [1] [2] [3] [4]
N. Docquier and S. Candel, Prog. Energ. Combust. Sci. 28, 107 (2002). A. G. Gaydon, “The Spectroscopy of Flames”, 1st ed., London: Chapman and Hall (1957). H. Coitout and G. Faure, Spectrosc. Lett. 29, 1201 (1996). K. K. Singh, C. Zhang, J. P. Gore, L. Momgeau and S. H. Frankel, Proc. Combust. Flame 30, 1707 (2005).
