Carbonaceous Deposit Density in a Fuel-Film-Cooled Rocket Combustor
2022; American Institute of Aeronautics and Astronautics; Volume: 38; Issue: 2 Linguagem: Inglês
10.2514/1.b38608
ISSN1533-3876
AutoresPhilip M. Piper, Timothée L. Pourpoint,
Tópico(s)Catalytic Processes in Materials Science
ResumoOpen AccessTechnical NotesCarbonaceous Deposit Density in a Fuel-Film-Cooled Rocket CombustorPhilip M. Piper and Timothée L. PourpointPhilip M. Piper https://orcid.org/0000-0002-5287-5330Air Force Research Laboratory, Edwards Air Force Base, California 93516*Aerospace Engineer, Combustion Devices Branch (RQRC).Search for more papers by this author and Timothée L. PourpointPurdue University, West Lafayette, Indiana 47907†Professor, School of Aeronautics and Astronautics. Associate Fellow AIAA.Search for more papers by this authorPublished Online:14 Feb 2022https://doi.org/10.2514/1.B38608SectionsRead Now ToolsAdd to favoritesDownload citationTrack citations ShareShare onFacebookTwitterLinked InRedditEmail AboutI. IntroductionHydrocarbon fuel-film-cooled rocket combustors can develop beneficial carbonaceous deposits that insulate regenerative cooling jackets from the chamber hot gases. The deposits form in conditions atypical for combustion soot literature: at pressures over 5 MPa, in forced convection flows at Mach >0.2, and on highly cooled walls with heat fluxes over 10 MW/m2.Accurate modeling of these conditions requires thermophysical properties and deposition parameters of the carbonaceous deposits. Previous studies have demonstrated multilayer deposition in fuel-film-cooled combustors likely occurring due to a combination of heterogeneous condensation of polycyclic aromatic hydrocarbons (PAHs) and thermophoretic diffusion of combustion soot [1]. Figure 1 shows a typical annular sample chamber liner (upper images) and sectioned sample (bottom image) while distinguishing between heterogeneous condensate (abbreviated to dense) and soot layers. The grayer dense layer was always present beneath the darker soot layer. Differences in chemical and physical structure cause the dense and soot deposit layers to have different properties. Therefore, all relevant thermophysical properties will be required for both layers to develop accurate heat transfer models.Graphite has a theoretical single-crystal density of 2266 kg/m3; however, typical engineering graphite has a density of around 1800 kg/m3 due to its small crystallite size [2]. Most combustion soot density measurements refer to effective density, which is typically taken as the mass of a soot particle divided by its (assumed spherical) volume [3,4]. For instance, soot agglomerate effective density from diesel engine exhaust was measured by Park et al. [5], with values ranging from 300 to 1200 kg/m3. Larger agglomerate mobility diameters result in lower effective densities. Note that effective density is not the same as absolute or bulk density.Fig. 1 Posttest images of the inlet (top left) and outlet (top right) of a sample tube chamber liner. High-resolution image of sectioned sample (bottom).Density can be quantified in several different manners for powders and other materials with voids. Absolute density excludes the volume of both closed and open pores and can be thought of as a measure of the density of only the solid portion of the powder. Ouf et al. [6] compare a diverse set of soot absolute densities from more common combustion applications like diffusion flames and automotive engines. Bulk density includes the open and closed pore volumes and is therefore always lower in value compared to absolute density.We measured densities of the dense and soot layers formed in a fuel-film-cooled H2O2–kerosene rocket combustor operated at 4.8 MPa [7]. Absolute densities were measured with a pycnometer and bulk densities were measured with a scanning electron microscope (SEM) and optical profilometer (for layer depth and volume) as well as a high accuracy scale (for mass).II. CombustorThe experimental hardware consisted of a kerosene–H2O2 bipropellant fuel-film-cooled rocket combustor that operated at pressures up to 4.8 MPa and run times up to 15 s [7]. The combustor was designed to have an axisymmetric heat sink sample chamber and removable chamber liners.Figure 2 shows the fuel-film cooled combustor used for the tests outlined in this experimental effort. Flow through the combustor started with 90 wt.% hydrogen peroxide injected through a 25-mm-inner-diam catalyst bed packed with silver-plated wire mesh screens to decompose the H2O2 to approximately 1030 K hot oxygen and steam. Fuel was radially injected through four 0.64 mm holes into the hot oxygen and steam flow to promote mixing at a stoichiometric mixture ratio of 7.5. A 0.38-mm-thick yttria-stabilized zirconia thermal barrier coating protected the flame holder section of the combustor from the high-enthalpy combustion gases at temperatures up to 2800 K when operated at stoichiometric conditions. Downstream of the step, a fuel-film injector introduced a low-velocity annular kerosene flow through a 0.25 mm axisymmetric gap to cool the inner wall of an annular sample chamber liner (same as that shown in Fig. 1). Sample chamber liners were thin, precision-machined tubes with inner and outer diameters of 29.8 and 32.0 mm, respectively. The sample length of 74 mm was long enough that the fuel-film decomposition occurred on the sample. A graphite converging–diverging nozzle with a 1.14 cm throat diameter choked the flow to achieve chamber pressures up to 4.8 MPa.K-type 1.59 mm thermocouples were embedded within blind holes of the sample chamber at four axial distances and 14 azimuthal locations. The one-dimensional inverse heat transfer problem was solved in our heat flux measurement paper, where we calculated peak metal wall temperatures (at axial location D, as shown in Fig. 2) of approximately 474, 573, and 569 K for copper, stainless steel, and nickel alloy chamber liners, respectively. These wall temperatures changed with axial distance from the fuel-film injector and bipropellant run time due to the combustor being heat sink cooled. [7]Fig. 2 CAD model of the fuel-film-cooled kerosene–oxygen combustor used in this study.III. MethodsA Micromeritics AccuPyc II 1340 gas pycnometer was used with helium to determine the absolute density of soot carbon deposit powder samples. The pycnometer 3.5 cm3 sample cup was calibrated with a metal sphere standard, as per the operator manual. Ten volume measurements were made per sample, resulting in precisions higher than the stated accuracy of the pressure transducer and volume calibration. Reported absolute density uncertainties are calculated from this volume precision and the scale uncertainty of 0.1 mg. Sample collection involved lightly scraping soot deposits from chamber samples, requiring over 15 cm2 of area to obtain accurate measurements. The dense deposits could not be removed in sufficient volume for gas pycnometry analysis; therefore absolute density is not reported for the dense layer.In situ bulk density, which accounts for soot pore volumes, was determined by measuring the mass and the volume of each layer. The layer mass was measured gravimetrically with 0.1 mg accuracy. Measurements were made on the posttest samples, after cleaning the soot layer with alcohol, and after cleaning the dense layer ultrasonically. Light cleaning with isopropyl alcohol removed the soot layer while leaving the dense layer mostly intact. The more tenacious dense layer was removed by ultrasonic cleaning in Ensolv, an n-propyl bromide solvent, for 60 minutes or until the metal substrate was fully visible. Samples were allowed to dry at room temperature and pressure for over 48 h between cleaning and measurement steps. SEM and optical profilometry depth measurements occurred on the metal sample substrate, before scraping for mass measurements. Layer volume was calculated by measuring the sample surface area and assuming an axisymmetric deposition profile where a trapezoidal rule approximate definite integral was used to convert layer height to layer volume. The soot layer volume was calculated by subtracting the dense layer volume (measured with SEM [8]) from the total volume (measured with profilometer [9]).Figure 3 shows SEM cross sections cut with a gallium focused ion beam and imaged with an Everhart–Thornley detector [8]. The distinction between dense and soot layer was often ambiguous, as demonstrated when cross-sectioning with the SEM focused ion beam column. For instance, some dense layer SEM cross sections showed multiple dense layers with gaps between them, which would manifest as a lower bulk density when integrated across the entire sample surface area.The largest uncertainties in this analysis were from the soot-dense layer transitions (see right cross section in Fig. 3) and the axisymmetric deposition profile assumption. Soot-dense layer transitions occupied, at most, 15% of the total layer height for each SEM image. Deposition profile volumes were calculated to vary by 27% along the azimuth based on soot and dense layer measurements at multiple azimuthal locations [8,9]. Uncertainties are not individually reported for bulk density measurements due to these two large sources of volume uncertainty.Fig. 3 Representative well-defined (left) and ambiguous (right) soot-dense carbon layer transitions from two different axial positions of the same sample.IV. ResultsAbsolute density measurements from tests 40, 58, and 63, which occurred at a variety of conditions (see Table A1), resulted in similar values to those expected for engineering graphite [2], as shown in Table 1. Standard deviation is reported based on the mass scale and pycnometer volume calculations, where the latter accounts for calibration and zeroing. The largest component of uncertainty was sample volume, where test 40 had the highest volume and 58 had the lowest volume. Absolute and bulk densities of the soot and dense layers were approximately constant across all tests. Dense layer bulk densities approached the absolute density of the soot layer. Soot layer open porosities of approximately 0.9 were calculated as the ratio of soot layer bulk density to absolute density when both measurements were present.V. DiscussionThe test conditions were different in film flow rate, bipropellant run time, sample material, and total mixture ratio. However, measured absolute and bulk densities were largely similar across these conditions. Note that tests 57 and 58 had the closest conditions, where differing sample materials of S30400 and N06600 resulted in similar wall temperatures, but the bulk densities from these two tests had the largest spread likely due to small sample volumes resulting in large uncertainties.Ouf et al. [6] measured and collected absolute densities from a variety of soot-based applications and correlated absolute density with organic to total carbon ratio (OC/TC). Based on their work, we would expect our soot to have a low OC/TC ratio due to its high graphite-like density. This matches findings from our spectroscopy work, which noted a lack of PAHs in the soot layer [1].The layer depth and density are primary factors in determining thermal resistance of the carbonaceous deposits. Because the soot layer was almost an order of magnitude less dense than the dense layer, it is expected to be the primary contributor to thermal insulation in fuel-film cooled combustors at chamber pressures and wall temperatures near our reported values.VI. ConclusionsThe thermal modeling of hydrocarbon-fueled rocket engines requires accurate density measurements of the solid carbonaceous layer deposited on the walls of the combustion chamber. The absolute and bulk densities of a multilayer carbonaceous deposit formed in a fuel-film-cooled rocket combustor at a steady-state chamber pressure of 4.8 MPa were measured using a gas pycnometer, scanning electron microscope, optical profilometer, and gravimetric analysis. Absolute densities of approximately 1900 kg/m3 for the soot upper layer matched those previously published in the combustion literature and approached values for engineering graphite [2]. Dense layer bulk densities from 1336 to 1639 kg/m3 were close to the soot absolute density, suggesting low void fractions, as confirmed by SEM cross sections. Some variation in the dense layer bulk density was expected due to ambiguous dense-to-soot layer transition that the bulk density measurement method could not distinguish. Soot bulk densities around 200 kg/m3 were significantly lower than those of the dense layer due to their highly porous structure. Thermal models for hydrocarbon fuel-film-cooled rocket combustors may benefit by accounting for the soot layer’s thermal insulation, due to its low density.E. L. PetersenAssociate EditorAppendix: Test MatrixAcknowledgmentsThis work is a collaboration between CFD Research Corporation (CFDRC) and Purdue University under a Phase II Small Business Technology Transfer (STTR) project (contract number FA9300-17-C-2501) sponsored by the U.S. Air Force at Edwards Air Force Base to develop chemical kinetics models for the prediction of carbon deposition in fuel-film-cooled rocket engines. The first author also acknowledges financial support from the Department of Defense’s Science, Mathematics, and Research for Transformation (SMART) scholarship, which is funded by the Under Secretary of Defense Research and Engineering (USD/R&E) and the National Defense Education Program Basic Research (NDEP/BA-1). The authors would like to thank Matthew Tanner for his help with running the scanning electron microscope and Stephen Schneider for enabling use of his lab group’s optical profilometer. Distribution Statement A: Approved for Public Release; Distribution is Unlimited. PA Clearance AFRL-2021-1700. References [1] Piper P. M., Orth R. M., Zemlyanov D. and Pourpoint T. L., “Carbonaceous Deposits in a Fuel-Film Cooled Rocket Combustor: Spectroscopy,” Carbon, Vol. 187, Feb. 2022, pp. 173–186. https://doi.org/10.1016/j.carbon.2021.11.001 CrossrefGoogle Scholar[2] Ho F. H., Graphite Design Handbook, General Atomics, San Diego, CA, 1988. https://doi.org/10.2172/714896 Google Scholar[3] Abegglen M., Durdina L., Brem B. T., Wang J., Rindlisbacher T., Corbin J. C., Lohmann U. and Sierau B., “Effective Density and Mass-Mobility Exponents of Particulate Matter in Aircraft Turbine Exhaust: Dependence on Engine Thrust and Particle Size,” Journal of Aerosol Science, Vol. 88, Oct. 2015, pp. 135–147. https://doi.org/10.1016/j.jaerosci.2015.06.003 CrossrefGoogle Scholar[4] Johnson T. J., Olfert J. S., Symonds J. P. R., Johnson M., Rindlisbacher T., Swanson J. J., Boies A. M., Thomson K., Smallwood G., Walters D., Sevcenco Y., Crayford A., Dastanpour R., Rogak S. N., Durdina L., Bahk Y. K., Brem B. and Wang J., “Effective Density and Mass-Mobility Exponent of Aircraft Turbine Particulate Matter,” Journal of Propulsion and Power, Vol. 31, No. 2, 2015, pp. 573–582. https://doi.org/10.2514/1.B35367 LinkGoogle Scholar[5] Park K., Cao F., Kittelson D. B. and McMurry P. H., “Relationship between Particle Mass and Mobility for Diesel Exhaust Particles,” Environmental Science & Technology, Vol. 37, No. 3, 2003, pp. 577–583. https://doi.org/10.1021/es025960v CrossrefGoogle Scholar[6] Ouf F.-X., Bourrous S., Fauvel S., Kort A., Lintis L., Nuvoli J. and Yon J., “True Density of Combustion Emitted Particles: A Comparison of Results Highlighting the Influence of the Organic Contents,” Journal of Aerosol Science, Vol. 134, Aug. 2019, pp. 1–13. https://doi.org/10.1016/j.jaerosci.2019.04.007 Google Scholar[7] Piper P. M., Gabl J. R., Dawson T. E., Mehta R. S. and Pourpoint T. L., “Carbonaceous Deposits in a Fuel-Film Cooled Rocket Combustor: Heat Flux Measurements,” Journal of Propulsion and Power, Vol. 37, No. 5, 2021. https://doi.org/10.2514/1.B38275 Google Scholar[8] Piper P. M. and Pourpoint T. L., “Carbonaceous Deposits in a Fuel-Film Cooled Rocket Combustor: Electron Microscopy,” Journal of Propulsion and Power, 2021. https://doi.org/10.2514/1.B38514 Google Scholar[9] Piper P. M. and Pourpoint T. L., “Carbonaceous Deposits in a Fuel-Film Cooled Rocket Combustor: Optical Profilometry,” Journal of Propulsion and Power, Vol. 37, No. 5, 2021. https://doi.org/10.2514/1.B38276 Google ScholarTablesTable 1 Absolute and bulk densities of fuel-film-cooled combustor carbonaceous depositsTest No.Soot layer absolute density, kg/m3Dense layer bulk density, kg/m3Soot layer bulk density, kg/m3Soot layer open porosity401929±89——————57——1336202——581960±30216392440.88631843±17314192030.89Table A1 Relevant portions of the fuel-film combustor test matrixTest No.Test dateFuelBiprop time, sBiprop pressure, MPaCore mixture ratioTotal mixture ratioSample materialFuel-film mass flow rate, g/s4006/18/2018RP-2154.837.442.89C1010052.55703/13/2019RP-264.627.483.52S3040037.85803/13/2019RP-264.637.473.52N0660037.86304/29/2019RP-2104.587.363.47N0660037.8The full test matrix can be found in our previous publication [7]. Previous article FiguresReferencesRelatedDetails What's Popular Volume 38, Number 2March 2022 CrossmarkInformationCopyright © 2022 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved. All requests for copying and permission to reprint should be submitted to CCC at www.copyright.com; employ the eISSN 1533-3876 to initiate your request. See also AIAA Rights and Permissions www.aiaa.org/randp. TopicsCombustion ChambersCombustorsCooling TechnologyHeat ConductionHeat ExchangersHeat TransferRocket EngineRocketryThermal Control and ProtectionThermal InsulationThermal Modeling and AnalysisThermodynamic PropertiesThermodynamicsThermophysical PropertiesThermophysics and Heat Transfer KeywordsRegenerative CoolingRocket CombustorAerospace EngineeringScanning Electron MicroscopeCombustorsThermal ResistanceFocused Ion BeamChamber PressureKeroseneEnthalpy of CombustionAcknowledgmentsThis work is a collaboration between CFD Research Corporation (CFDRC) and Purdue University under a Phase II Small Business Technology Transfer (STTR) project (contract number FA9300-17-C-2501) sponsored by the U.S. Air Force at Edwards Air Force Base to develop chemical kinetics models for the prediction of carbon deposition in fuel-film-cooled rocket engines. The first author also acknowledges financial support from the Department of Defense’s Science, Mathematics, and Research for Transformation (SMART) scholarship, which is funded by the Under Secretary of Defense Research and Engineering (USD/R&E) and the National Defense Education Program Basic Research (NDEP/BA-1). The authors would like to thank Matthew Tanner for his help with running the scanning electron microscope and Stephen Schneider for enabling use of his lab group’s optical profilometer. Distribution Statement A: Approved for Public Release; Distribution is Unlimited. PA Clearance AFRL-2021-1700.PDF Received5 August 2021Accepted12 January 2022Published online14 February 2022
Referência(s)