Scott Stouffer

The Impact of Heat Release in Turbine Film Cooling

The Ultra Compact Combustor is a design that integrates a turbine vane into the combustor flow path. Because of the high fuel-to-air ratio and short combustor flow path, a significant potential exists for unburned fuel to enter the turbine. Using contemporary turbine cooling vane designs, the injection of oxygen-rich turbine cooling air into a combustor flow containing unburned fuel could result in heat release in the turbine and a large decrease in cooling effectiveness. The current study explores the interaction of cooling flow from typical cooling holes with the exhaust of a fuel-rich well-stirred-reactor operating at high temperatures over a flat plate. Surface temperatures, heat flux, and heat transfer coefficients are calculated for a variety of reactor fuel-to-air ratios, cooling hole geometries, and blowing ratios. Results demonstrate that reactions in the turbine cooling film can result in increased heat transfer to the surface. The amount of this increase depends on hole geometry and blowing ratio and fuel content of the combustor flow. Failure to design for this effect could result in augmented heat transfer caused by the cooling scheme, and turbine life could be degraded substantially.

Heat Release in Turbine Cooling I: Experimental and Computational Comparison of Three Geometries

Marc D. Polanka,∗ Joseph Zelina,∗ Wesly S. Anderson, and Balu Sekar∗ U.S. Air Force Research Laboratory, Wright-Patterson Air Force Base, Ohio 45433 Dave S. Evans Air Force Institute of Technology, Wright-Patterson Air Force Base, Ohio 45433 Cheng-Xian Lin University of Tennessee, Knoxville, Tennessee 37996 and Scott D. Stouffer University of Dayton Research Institute, Wright-Patterson Air Force Base, Ohio 45433

Modeling Soot Formation in a Stirred Reactor

The formation of carbonaceous particulate matter and polycyclic aromatic hydrocarbons has recently been studied (Stouffer, et al, 2002 and Reich, et al, 2003) in a toroidal well-stirred reactor using ethylene as the fuel, with and without the additive ethanol. In the later work, modeling of the gas-phase species was performed and compared to the experimental trends. In the present study, a modified version of the CHEMKIN-based code for ‘perfectly stirred reactors’ has been used to model soot particle formation, including computations of particle mass and smoke number. Detailed soot formation routines have been extracted from Hall and coworkers (1997), who modeled soot formation in flames. Experimental trends are accurately modeled by the code with quantitative accuracies generally within 50%. The importance of accurate knowledge and control of reactor temperature is discussed. In fact, scatter in the original experimental study can be largely attributed to inadequate temperature control. Speculation for differences between the model and experiment are offered while additive effects and the well known ‘soot bell’ are discussed. For the initial experiments examined by Stouffer et al, the effect of the additive is largely due to temperature differences.© 2004 ASME

Spray characteristics and flame structure of jet A and alternative jet fuels

A 2D phase Doppler anemometer is used to characterize alternative jet fuel droplets and compare them to Jet A fuel droplets in the National Jet Fuel Combustion Program referee single cup combustor near lean blowout. The two alternative jet fuels selected were chosen for their unusual properties: one with low cetane number and one with a flat boiling curve. Measurements are made on all three fuels at steady-state combustion at a pressure of 30 psia, swirler pressure drop of 3 percent, and a global equivalence ratio of 0.096. The results show differences in the droplet diameter distributions of the different fuels. This is particularly prominent in the flat boiling curve fuel, which has a Sauter mean diameter between 12 and 37 microns larger than the corresponding points for the other two fuels. OH* chemiluminescence imaging, conducted at 20kHz, is used to compare flame structures between the fuels as well as to correlate spray characteristics to flame location. The averaged OH* images show significant differences between the flame structures of the three fuels. The results of the study motivate further investigation into the correlation between alternative jet fuel spray characteristics and flame behavior for use in evaluating and predicting alternative jet fuel performance.

Large-eddy simulations of fuel effects on gas turbine lean blow-out

Towards the implementation of alternative jet fuels in aviation gas turbines, testing in combustor rigs and engines is required to evaluate the fuel performance on combustion stability, relight, and lean-blow out (LBO) characteristics. The objective of this work is to evaluate the effect of different fuel candidates on the operability of gas turbines by comparing a conventional petroleum-based fuel with two other alternative fuel candidates. Numerical investigations are performed to examine the performance of these fuels on the stable condition close to blow-out and LBO-behavior in a referee gas turbine combustor. Large-eddy simulations (LES) are performed at stationary conditions near LBO to examine effects of the fuel properties on evaporation, gaseous-fuel deposition, flame anchoring.

Soot Reduction Research Using a Well-Stirred Reactor

** †† ‡‡ §§ A comprehensive research program involving industry, academia and Government laboratories is developing a fundamental understanding of the complex interactions of fuel additives with the processes leading to particulate matter emissions from military gas turbine engines. The goal of this program is to eventually select promising additive compounds that would reduce particulate matter (PM) emissions. One experimental platform for assessing the performance of additives is the Well-Stirred Reactor (WSR) research combustor at the Air Force Research Laboratory. The WSR provides a unique capability for simulating the chemical kinetics within the primary zone of a gas turbine engine combustor. The current study presents results from six different compounds (nitromethane, nitroethane, nitropropane, cyclohexanone, pyridine, and quinoline) as additives for soot reduction. The effect of the temperature on the chemical kinetics vs. the chemical effect of the additives is addressed for the nitroalkane additives.

Effects of a Reacting Cross-Stream on Turbine Film Cooling

Film cooling plays a critical role in providing effective thermal protection to components in modern gas turbine engines. A significant effort has been undertaken over the last 40 years to improve the distribution of coolant and to ensure that the airfoil is protected by this coolant from the hot gases in the freestream. This film, under conditions with high fuel-air ratios, may actually be detrimental to the underlying metal. The presence of unburned fuel from an upstream combustor may interact with this oxygen rich film coolant jet resulting in secondary combustion. The completion of the reactions can increase the gas temperature locally resulting in higher heat transfer to the airfoil directly along the path line of the film coolant jet. This secondary combustion could damage the turbine blade, resulting in costly repair, reduction in turbine life, or even engine failure. However, knowledge of film cooling in a reactive flow is very limited. The current study explores the interaction of cooling flow from typical cooling holes with the exhaust of a fuel-rich well-stirred reactor operating at high temperatures over a flat plate. Surface temperatures, heat flux, and heat transfer coefficients are calculated for a variety of reactor fuel-to-air ratios, cooling hole geometries, and blowing ratios. Emphasis is placed on the difference between a normal cylindrical hole, an inclined cylindrical hole, and a fan-shaped cooling hole. When both air and nitrogen are injected through the cooling holes, the changes in surface temperature can be directly correlated with the presence of the reaction. Photographs of the localized burning are presented to verify the extent and locations of the reaction.

Design Studies of Turbine Blade Film Cooling with Unburned Fuel in Cross Stream Flow

Film cooling plays a critical role in providing effective thermal protection for components of modern gas turbine engines. A Reynolds-Averaged Navier-Stokes (RANS) approach is employed to simulate the complex turbulent reactive flow exhibited by film cooling flows emanating from a surface. The widely used SST k-ω turbulence model is used to model the turbulent flow. A simplified two-step propane-air reaction scheme has been employed to model the combustion process and study the underlying physics of mixing between film cooling and cross stream flow driving secondary combustion. The Eddy-dissipation concept (EDC) approach is used to account for the turbulencechemistry interaction. The three-dimensional geometry is modeled using a hybrid mesh. The reacting flow field and the resulting film cooling effectiveness are predicted for circular, angled circular, and fanned film hole geometry for two equivalence ratios, one blowing ratio, and both air and N2 film cooling. Numerical results between air and N2 film cooling generally agree well with experimental data in terms of relative temperature change, non-dimensionalized with respect to the N2 film temperature. Results indicate that hole geometry plays a key role in the effectiveness of the film cooling design. Film cooling provided by the normal circular hole is considerably lower than that provided by the angled and fanned hole for both lean and in rich conditions. Air injection feeds secondary combustion that substantially increases the wall temperature on the flat surface for a considerable distance downstream of the hole. However, the shaped hole produces a larger effective film area in the immediate vicinity of the cooling hole both axially and laterally when compared to the normal circular and angled circular configurations. For fuel rich conditions a distinct hot area downstream of the coolant hole generated by the secondary combustion feed by coolant air injection has been predicted. This results in negative cooling effectiveness in certain areas of the flat surface, specifically for the shaped hole. The N2 coolant air injection provides no O2 to feed secondary combustion for the unburned fuel exiting the combustor at high equivalence ratios.

Impact of an Upstream Film-Cooling Row on Mitigation of Secondary Combustion in a Fuel Rich Environment

In advanced gas turbine engines that feature very short combustor sections, an issue of fuel-rich gases interacting with the downstream turbine components can exist. Specifically, in combustors with high fuel-to-air ratios, there are regions downstream of the primary combustion section that will require the use of film-cooling in the presence of incompletely reacted exhaust. Additional combustion reactions resulting from the combination of unburnt fuel and oxygen-rich cooling films can cause significant damage to the turbine. Research has been accomplished to understand this secondary reaction process. This experimental film-cooling study expands the previous investigations by attempting to reduce or mitigate the increase in heat flux that results from secondary combustion in the coolant film. Two different upstream cooling schemes were used to attempt to protect a downstream fan-shaped cooling row. The heat flux downstream was measured and compared between ejection with air compared to nitrogen in the form of a heat flux augmentation. Planar Laser Induced Fluorescence (PLIF) was used to measure relative OH concentration in the combustion zones to understand where the reactions occurred. A double row of staggered normal holes was unsuccessful at reducing the downstream heat load. The coolant separated from the surface generating a high mixing regime and allowed the hot unreacted gases to penetrate underneath the jets. Conversely, an upstream slot row was able to generate a spanwise film of coolant that buffered the reactive gases off the surface. Essentially no secondary reactions were observed aft of the shaped coolant hole ejection with the protective slot upstream. A slight increase in heat transfer was attributed to the elevated freestream temperature resulting from reactions above the slot coolant. Creating this full sheet of coolant will be a key toward future designs attempting to control secondary reactions in the turbine.

Numerical Modeling of Combustion Performance for a Well- stirred Reactor for Aviation Hydrocarbon Fuels

An extensive numerical investigation is conducted to characterize the combustion performance of a toroidal wellstirred reactor (WSR) burning a stoichiometric mixture of C 2H4 and air. The three-dimensional flow inside the toroidal combustor and exhaust port is simulated by employing the steady RNG k-e RANS governing equations. A comprehensive, steady state computational model which employs either one-step (1SGM), two-step (2SGM), or reduced mechanism (RM) with 15 reactions, 18 species, and 10-QSS species is used to simulate the combustion process with species- and temperature-dependent thermodynamic and transport properties. The interaction between turbulence and chemistry is assumed to be negligible (i.e. laminar chemistry is assumed valid). The RM is firstly validated by comparing the predicted detailed flame structure with measurements in an axisymmetric laminar premixed flame. There is good quantitative and qualitative agreement between measurements and predictions. The 20°-off-the-radius of-the-torus 48 fuel-air jets enter the combustor at Mach (Ma) and Reynolds number (Re 2R ), and static temperature of 0.86, 10520, and 420 K, respectively. Beyond the jets’ potential cores the velocity decreases, while the jets expand at nearly constant spread angle (ξ1/2 /R) of 14.5°. This spread angle is substantially larger than that predicted in an axisymmetric jet as well as that predicted by boundary layer theory. This is due to the presence of a tangential flow velocity component in the WSR. Two counter-rotating forced vortices are formed on each side of the jet on the combustor cross-sectional area. Along the jets centerline the temperature first decreases slightly and then increases until it reaches a maximum. Multiple premixed-like flames are established at the fuel-air injector exits. These flames are stabilized in high velocity regions and positioned at the same relative location with respect to their corresponding fuel-air injectors. Whereas these flames are thin near the injectors, their thickness increases substantially at the flame tip due to preferential diffusion (i.e. Lewis number). Moreover, results indicate that mixing time ( τmix ) is the fastest at the jets, in-between-jets zones, and in the exhaust ports and lowest in regions downstream the jets and counter-rotating vortexes. The average number of turbulent mixing operations per residence time (N TMO ) is in the range of ~41-59. It is also found that N TMO increases with temperature. Furthermore, an order of magnitude analysis indicates that the WSR is operating in the well-stirred reactor turbulent regime. Comparisons of combustion products and reactor temperatures among the WSR, an axisymmetric jet, PSR, and PaSR suggest that for the correct assessment of homogeneity in the WSR a comprehensive mechanism is needed. Even though 1SGM and 2SGM provide reliable information regarding the velocity and temperature flow fields as well as turbulent regime, they do not provide valuable information for further improvement on the homogeneity of the WSR.