Species Concentration and Temperature Measurements in a Lean, Premixed Flow Stabilized by a Reverse Jet

-The chemical and thermal structure and the emission performance of an aerodynamic f1ameholder are presented and examined. Recirculation is established by injecting a premixed jet into an opposing main stream of premixed reactants. The injection of the jet directly into the recirculation zone provides a control of the stabilization zone mixture ratio, temperature, and size not found in bluffbody flameholding. The size and stoichiometry ofthe recirculation zone is dictated by the jet velocity and mixture ratio respectively. A parametric study of the controlling variables (main and jet stream velocities, main and jet stream equivalence ratios) reveals the partitioning between the recirculation zone and wake in both the heat release and pollutant production. An examination of the emission indexes and flowfield profiles of temperature and species concentration estab lishes the influence and control of jet and mainstream conditions on pollutant production. A reduction in jet velocity and/or an enrichment of the jet, for example, effects a substantial change in NO, emission. Further, jet enrichment extends the lean blow-off limit of the mainstream. There exists a point, however, beyond which the reaction is not supported in the wake and further leaning of the mainstream results in a substantial emission of unspent fuel.


INTRODUCTION
The present demand to develop energy-efficient and low-emission combustion systems requires tailoring the combustor aerodynamics to more effectively control the temperature field, the distribution of fuel, and the limits of flame stability. Aerodynamic flameholding offers some advantages in this regard as an alternative to conventional bluffbody or sudden expansion flameholding.
A reverse jet flameholder is shown in Figure I. The incoming mainstream of premixed fuel and air is opposed by a high velocity jet positioned along the longitudinal axis. The jet creates the zone of recirculating flow necessary to stabilize the reaction. The jet stream, also composed of premixed fuel and air, constitutes a few percent of the total flow [Present address: KVB Engineering, Inc., P.O. 19518, Irvine, CA 92714. t Present address: Pacific Environmental 'Services, 1905 Chapel Hill Road, Durham, NC 27707. §To whom correspondence should be addressed: Mechanical Engineering, University of California, Irvine, CA92717. 211 entering the combustor, but contributes as much as one third to the mass within the recirculation zone. As a result, a wide range of stable combustion conditions may be achieved by independently varying the mixture ratios and velocities of the jet and maintstream (Figure 2). Most notably, by enriching the jet, the lean blow-off limit can be significantly extended. The reverse jet ("opposed jet") flameholder was first introduced as a candidate for flameholding in afterburners (Schaffer, 1954). Jets injected at an angle from the wall, were proposed for stabilizing the reaction during afterburner operation while avoiding the attendant pressure drop associated with conventional, physical flameholders when afterburning was not in use. Adoption proved to be infeasible, however, upon the discovery that the flameholding performance of the reverse jet drops sharply when the jet is located at small angles (ca. 5°) to the opposing flow (Duclos et 01., 1957).
For detailed flowfield maps, radial traverses were taken at twelve axial locations, with the distance between axial locations ranging from 2.54 mm (O.l-inch) in the nose of the recirculation zone to up to 24.13 mm (0.95-inch) between the jet tube exit and combustor "exit plane" (a plane arbitrarily located 13 mm upstream of the combustor outlet and 67 mm downstream of the jet tube exit). Radial locations were at 3.05 mm (0.12-inch) increments between the jet tube wall and the chamber wall.
Analysis of the sample gas was performed using a packaged emission analysis system (Scott Research, Model 113). Passage of the gas through an ice bath allowed all concentration measurements to be taken on a dry basis. Nitrogen oxides (NO, NO x ) concentrations were measured using a chemiluminescence analyzer (Scott Research, Model 125). Nondispersive infrared analyzers (Beckman, Models 315B and 315BL respectively) were used to measure carbon dioxide (C02) and carbon monoxide (CO) concentrations. Total hydrocarbon (HC) concentration was measured by a flame ionization detector (Scott Research,Model 215

Combustor
The experimental facility is shown in schematic in Figure 4. The combustor consisted of a 51 mm (2-inch) i.d. quartz cylindrical chamber with an overall length of 457 mm (l8-inches). The jet stream issued from a 1.32 mm (0.052-inch) i.d. hole at the end of a 6.35 mm (0.25-inch) o.d., watercooled, stainless steel jet tube. The jet tube exit was located upstream of the combustor outlet. Combustion air was supplied by the building compressed air system and was dried and filtered prior to introduction to the combustor. Commercial grade propane was supplied from pressurized cylinders. A complete description of the test facility and test analysis system is available (Peterson and Himes, 1978).

Species and Temperature Measurements
Temperatures were measured using an unshielded, fine wire, platinumjplatinum-H percent rhodium thermocouple mounted on a micrometer traverse EXPERIMENT FIGURE 4 Schematic of combustor facility.
structure for a range of parametric variations of the four primary controlling variables: Main and jet stream equivalence ratios, main and jet stream velocities. Exit plane and detailed flowfield profiles are presented and analyzed for NOx , carbon monoxide (CO), total hydrocarbons (HC), and temperature. The goals are to provide (I) insight into the performance of a reverse jet, aerodynamic f1ameholder, (2) guidance for practical applications of aerodynamic f1ameholding, and (3) a data base for future code testing.

Vm
( Table I) that the mainstream equivalence ratios (<Pili) were biased to fuel lean mixture ratios because of the interest in lean mainstream emission performance. The jet equivalence ratios (!Pi) were biased to the fuel rich mixture ratios to extend the lean limit of the main stream mixture. The range of main stream (V III) and jet velocities (Uj) allowed an examination of the effects of recirculation zone size and stoichiometry on flame structure and pollutant emission.
Finally, emission indexes are presented to summarize the emission behavior of the combustor at the conditions considered, and to provide a practical perspective to the utility of opposed jet flameholding. A complete set of the data and results is available (McDannel, 1979).

A. Base Case
The detailed flowfield maps and exit plane profiles for the base case are presented in Figure 5. Two distinct regions can be deduced from the results ( Figure 6). One is the recirculation zone, which is a zone of strong backmixing driven by the jet flow. The other is a radially propagating reaction in the wake of the recirculation zone. For example, the oxidation of hydrocarbons ( Figure 5a) and formation of carbon monoxide (CO) occurs within the recirculation zone where there is intense mixing of reactants with hot products, and along the radially propagating wake reaction front. Within the wake, temperature, oxygen, and residence time are sufficient to ensure nearly complete HC consumption, and to initiate the oxidation of CO to carbon dioxide (C0 2) as demonstrated by the decrease in CO concentration adjacent to the jet tube proceeding downstream toward the exit plane. Proceeding toward the combustor wall, the concentrations of HC and O 2 approach those of the reactants. As a result, the source of the hydrocarbons emitted at the exit plane is the area outside of the wake.
Oxides of nitrogen (NOx ) are formed thermally in both the recirculation zone and wake as a result of elevated temperatures, sufficient residence time, and available oxygen. Area-averaged concentrations calculated at both the "jet exit plane" and combustor "exit plane" (Figure la), indicate that 75 percent of the total NO x emitted is formed in the recirculation zone for this base condition.
The exit plane profiles (Figure 5b) show the general structure of the wake. Within the wake and proceeding from the jet tube to the combustor wall, HC and oxygen concentrations and temperature remain relatively constant, while CO concentrations increase slowly and NOx concentrations decrease. At approximately rlR =0.55, the concentrations of HC, CO 2, and O 2 change sharply. Oxygen and HC rise, CO 2 drops and CO peaks. Eventually, the HC and O 2 rise to the reactant concentrations.
Finally, it is noteworthy that the NO/NO x ratio drops abruptly at the flame front ( Figure 5b). This is attributed to the rapid mixing of hot products and cold reactants at the flame front which produce radical relaxation reactions and associated populations of hydroperoxy radicals (H0 2) sufficient to oxidize NO to N02. Unfortunately, these events can be influenced by the probe, and the extent to which the measured levels of N02 are real or artifacts of the probe remains unanswered. However, an evaluation (Chen, et al., 1979) of similar observations in a premixed combustor (Oven, et al., 1979) concluded that, although measurements within high temperature reaction zones (e.g., within the recirculation zone and wake) are likely biased by probe-induced oxidation of NO, elevated levels of N02 in areas of rapid flame quench (e.g., the flame front) are likely real and not artifacts of the probe.

B. Parametric Study Mainstream and Jet Velocities
The major effect of changing mainstream and jet velocity is to change the size of the recirculation and wake region. This is demonstrated in the present study by independently increasing the mainstream velocity (U m ) and decreasing the jet velocity (Uj). The effect of either is to decrease the size of the recirculation and wake regions.
The visual appearance of the flame for the base case and two variations on the base case is shown in Figure 7. Both the penetration of the jet and the radial propagation in the wake are restricted by increasing the mainstream velocity or by decreasing the jet velocity. This is confirmed by the detailed temperature maps presented in Figure 8.
A decrease in the size of the recirculation and wake reaction zones produce a net reduction in residence time and, hence, a net reduction in the NOxproduction (Figure 9).

Mainstream and Jet Equivalence Ratios
The mainstream equivalence ratio (<Pm) is the dominant variable controlling the heat release and, ultimately, the pollutant emission. The effect on heat release is shown in Figure 10.
HC, CO, T. As indicated by He exit plane profiles, the wake reaction propagates further radially as the mainstream mixture is enriched. This is a consequence of the decrease in air dilution as <Pm is increased from 0.6 to 1.0. The increase in wake reaction as <Pm is enriched from 1.0 to 1.2 is attributed to an increased availability of hydrocarbon radicals. Peak flame velocities for propaneair flames generally occur at equivalence ratios rich of stoichiometric (Fristrom and Westen berg, 1965). The highest temperatures occur for the base case I (</>m = 1.0). Peak temperatures are about 300 0 K lower for </>m = 1.2 and 500 0K lower for 0.8.
Carbon monoxide concentrations increase with mainstream equivalence ratio as the amount of available oxygen to oxidize CO to CO 2 decreases. For all conditions, the temperature is sufficient for the oxidation to occur. For equivalence ratios of 0.8 to 1.0, peak concentrations correspond to the location of the wake reaction front. Inside the front, CO is oxidized to CO 2 • This accounts for the increase in temperature in the wake. Ahead of the front, CO diffuses into the cold reactant gases.
At </>m = 1.2, the absence of oxygen in the wake results in relatively constant CO concentrations and an absence of a distinct CO peak at the flame front.
For all cases except </>m =0.6, temperatures are fairly constant within the wake reaction zone dnd drop at the radially propagating wake reaction front. For </>m =0.6, the temperatures drop immediately adjacent to the axis, and HC concentrations remain elevated while CO concentrations fall instead of rise when proceeding from the axis to the chamber wall. This suggests that reaction in the wake is suppressed and CO formation, for example, is restricted to the recirculation region upstream with radial diffusion in the wake. Note that CO concentrations for </>m =0.6 are not appreciably lower than for </>m =0.8. In fact, near the jet tube, concentrations are lower for </>m =0.8. The additional oxygen available at </>m =0.6 is offset by lower temperatures.
Varying the jet stream equivalence ratio (</>J) allows determination of the effect of recirculation zone mixture ratio. The effect on exit plane profiles of HC and temperature is pronounced only at </>m = 0.6. Higher jet equivalence ratios result in lower HC concentrations near the jet tube wall and higher (b) NO, flow maps. temperatures, and this effect diminishes as distance from the jet tube wall increases. This is attributed to higher temperatures in the recirculation zone that result from recirculation zone mixture ratios closer to stoichiometric. This effect is not as pronounced for the other equivalence ratios because each effectively sustains a fully developed reaction in the wake.
The effect of jet equivalence ratio on carbon monoxide is noticeable only near the jet tube. For <Pm =0.6, richer jet mixtures result in lower CO emissions because of higher temperatures in combination with the elevated concentration of oxygen. This same trend occurs, but to a much lesser extent, for <Pm =0.8. At <Pm = 1.0 and 1.2 this trend is reversed. For these cases oxygen, and not temperature, limits the CO oxidation.  NOx. The NO x profiles are presented in Figure  11. Peak concentrations are highest for the base case (<Pm = 1.0). The concentrations are slightly lower at <Pm = 1.2. For the leaner cases, there is a significant drop in concentrations.
For all cases the shape of the NO x exit plane profile is similar, decreasing almost linearly from the jet to the combustor wall. The shape indicates that most of the NO x is formed in the recirculation zone, and diffuses by turbulent transport downstream. This is confirmed in the detailed flow maps presented in Figure 1\ b. These trends correspond well with the trends observed for temperature (Figure 10d) reflecting the temperature dependence of NOx formation reactions.
Note that the <Pm = 1.2 exit plane profiles ( Figure  II) intersect the <Pm = 1.0 profiles with higher concentrations near the chamber wall. This is attributed to the additional production of NO x in the larger wake reaction zone associated with the rich mainstream. Although the recirculation zone is larger as well, the data indicate that NO x production in the recirculation zone is not increased, a consequence of suppressed oxygen availability and temperature. Jet equivalence ratio (,h) directly impacts both the mixture ratio and temperature of the recirculation zone. As a result, the effect of jet equivalence ratio on NO x production is predictable. Production of NO x is increased with jet enrichment for lean mainstreams (</>m =0.6, 0.8), decreased with jet enrichment or jet leaning for stoichiometric mainstream (</>m = 1.0), and increased with a lean jet or decreased with a rich jet for a rich (</>m = 1.2) mainstream.
The effect of mainstream equivalence ratio on the NO/NO x ratio is shown in Figure 12 for a stoichiometric jet. (Other jet equivalence ratios are omitted for clarity.) The rapid drop in the NO/NOx ratio occurs at the flame front for each of the cases (</>m = 1.2, 1.0, 0.8) wherein a wake reaction was supported. The low NO/NO x ratio for </>m=0.6 is attributed to the quench zone surrounding the hot recirculation zone in the absence of a wake reaction,

C. Emission Indexes
The emission indexes for NO x , CO, and HC are presented in Figure 13 as a function of mainstream equivalence ratio (</>m). The parameters are jet equivalence ratio (</>1) and mainstream velocity (U m). The procedure used to compute the emissibn index involved correcting the data for water vapor in the exhaust, calculating the area-averaged ekit plane concentrations and the area-averaged mass emission, and taking the ratio of the mass emissibn to the fuel mass input.
I The emission index data reflect the observations derived from the detailed results above, and place  stability by direct, reactant Injection into the recirculation zone. The data presented provide insight into the performance of the reverse jet flameholder, give guidance for practical application, and establish a data base for future testing of elliptic codes.
The reverse jet combustor, chemically and aerodynamically, consists of two distinct regionsthe recirculation zone and the wake. The mainstream and jet velocity influence the size of these two regions, while the mainstream and jet mixture ratios affect the overall chemistry and heat release.
The influence of the jet on the emission of NO x is one of the more interesting characteristics of the flameholder. Jet changes which reduce the size, lower the temperature, and decrease the residence time are favorable to the reduction of NO x emission. For example, an enrichment of jet equivalence ratio affects recirculation zone temperature and mixture ratio, and will produce either an increase or decrease in the net emission of NO x depending on the mainstream equivalence ratio. Such changes do not significantly affect He and CO emissions. A primary benefit from an enriched jet is to extend the lean blow-off limit and thereby maintain combustor stability simultaneous with a reduction in the emission of NO x . However, a practical limit exists beyond which the emission of unburnt fuel is excessive. In the present experiment, this limit occurred at a mainstream equivalence ratio of approximately 0.8. the performance of the combustor into a practical perspective. For example, NO x emissions are highest at <Pm = 1.0, and are only slightly lower at 1.2. Values are 50 percent lower at <pm =0.8 than at stoichiometric, and those at 0.6 are from 5 to 15 times lower than those for the base case (depending on jet equivalence ratio). The high temperatures that favor NO x formation also favor hydrocarbon oxidation. In the absence of a developed wake reaction at <Pm =0.6,60 percent of the fuel is emitted unburned. As suggested by the analysis of the detailed flow maps, jet equivalence ratio has a minimum impact on HC and CO emission, but does affect NO x emission.