George C. Harting

Fuel decomposition and boundary-layer combustion processes of hybrid rocket motors

Using a high-pressure, two-dimensional hybrid motor, an experimental investigation was conducted on fundamental processes involved in hybrid rocket combustion. HTPB (Hydroxyl-terminated Polybutadiene) fuel cross-linked with diisocyanate was burned with GOX under various operating conditions. Large-amplitude pressure oscillations were encountered in earlier test runs. After identifying the source of instability and decoupling the GOX feed-line system and combustion chamber, the pressure oscillations were drastically reduced from +/-20% of the localized mean pressure to an acceptable range of +/-1.5% Embedded fine-wire thermocouples indicated that the surface temperature of the burning fuel was around 1000 K depending upon axial locations and operating conditions. Also, except near the leading-edge region, the subsurface thermal wave profiles in the upstream locations are thicker than those in the downstream locations since the solid-fuel regression rate, in general, increases with distance along the fuel slab. The recovered solid fuel slabs in the laminar portion of the boundary layer exhibited smooth surfaces, indicating the existence of a liquid melt layer on the burning fuel surface in the upstream region. After the transition section, which displayed distinct transverse striations, the surface roughness pattern became quite random and very pronounced in the downstream turbulent boundary-layer region. Both real-time X-ray radiography and ultrasonic pulse-echo techniques were used to determine the instantaneous web thickness burned and instantaneous solid-fuel regression rates over certain portions of the fuel slabs. Globally averaged and axially dependent but time-averaged regression rates were also obtained and presented.

HEAT FLUX AND INTERNAL BALLISTIC CHARACTERIZATION OF A HYBRID ROCKET MOTOR ANALOG

The results of lab-scale hybrid rocket motor test firings and solid fuel pyrolysis experiments were combined with several analytical and numerical techniques to develop a semi-empirical method to analyze and correlate the experimental data with descriptive non-dimensional parameters. The experimental regression rate database was combined with a solid-fuel pyrolysis model and the energy flux balance equation to determine the total heat flux reaching the fuel surface. The component heat fluxes due to convection, radiation from the gas-phase combustion products, and radiation from soot were determined using various methods. The results of the analysis showed that under some conditions the magnitude of the overall radiant heat flux to the fuel surface was quite significant compared to the convective heat. Though CO2 represented the most important radiating gas-phase combustion product, radiation from soot was found to account for 80 to 90% of the total radiant heat flux. Dimensionless regression rate correlations indicated that the classical hybrid boundary layer correlation must be modified to account for the effects of radiation and variable fluid properties across the boundary layer. The Boltzmann number and velocity ratio between the flame and bulk flow were employed to make these corrections. In addition, it was found that the deduced Stanton number can sometimes be larger than the reference Stanton for turbulent flow over a flat plate due to various effects in the hybrid shear flow. The effects of variable transport properties were not found to be important in correlating the data, and no evidence for the existence of rate-limiting chemical kinetics was observed even for relatively large mass fluxes and low pressures. Finally, for the same percent weight mass addition, Alex powder (an ultra-fine aluminum powder, 0.05 to 0.1 |J.m) caused a substantially greater increase in regression rate than did conventional aluminum powder (5 to 15 urn).

Pyrolysis and combustion of solid fuels in various oxidizing environments

An experimental study was conducted to determine the dependence of the regression rate of two solid-fuel formulations (cured HTPB and JIRAD fuel) on operating conditions near the head-end of a hybrid motor. Cylindrical fuel samples were burned in a windowed combustor at pressures ranging from 0.79 to 3.55 MPa. The burning was sustained by a diffusion flame created over the fuel surface by an impinging oxidizer jet. The gaseous oxidizer was a mixture of oxygen and nitrogen with the oxygen mass fraction ranging from 0.21 to 1.00. For both fuels, the regression rate increased with oxidizer mass flow rate, oxygen mass fraction, and oxidizer temperature, but decreased at higher pressures in the range tested. Measured regression rates, surface temperatures, and operating parameters were used to validate a simple power-law regression rate correlation for each fuel. A modified form of Marxmans analysis was also developed to consider the effects of species diffusion, heterogeneous surface reactions, and fluid-dynamic/heat transfer processes. The effect of heterogeneous reactions was found to be important in the range of parameters tested. For both fuels, the activation energies of the heterogeneous reactions were lower than the pyrolysis activation energies, with HTPB being more reactive than the JIRAD fuel. Nomenclature A pre-exponential coefficient [mm/s] B Spalding transfer number cf skin friction coefficient CP specific heat [kJ/kg-K ] Ea activation energy [kcal/mole] Ah enthalpy difference between the flame and the surface [kJ/kg] AHf° heat of formation [kJ/mole] hpy effective heat of pyrolysis [kJ/kg] hpych chemical enthalpy of fuel pyrolysis [kJ/kg] If heat flux feedback due to radiation [W/m] Krf, concentration of reactants below the flame m mass flow rate [g/s or kg/s] O/F oxidizer-to-fuel ratio P pressure in chamber [MPa or Pa] Q heat release [kJ/kg] rb solid-fuel regression rate [mm/s] Re Reynolds Number Ru universal gas constant [kcal/K-mole] T temperature [K] u velocity of gas [m/s] Y species mass fraction (j) velocity ratio A, thermal conductivity of the species [W/m-K] p density [g/cm] Subscripts b flame boundary e edge of the boundary layer EX oxidizer below the flame f solid-fuel F fuel F,O species and oxidizer g gas-phase h heterogeneous 1 initial condition/ i* species o initial condition O,T oxidizer and energy 02 oxygen OX oxidizer P products of combustion py pyrolysis r reactants of combustion ref reference condition s solid-fuel surface st stoichiometric condition «> free-stream quantity

REGRESSION RATE AND HEAT TRANSFER CORRELATIONS FOR HTPB/GOX COMBUSTION IN A HYBRID ROCKET MOTOR

Distinguished Professor and Director of the High Pressure Combustion Laboratory. AIAA Fellow. § Ph.D. Student. Student Member AIAA. 1 Ph.D. Candidate, AFRL Palace Knight. Member AIAA * Visiting Professor, on sabbatical leave from Rafael, Israel. Associate Fellow AIAA. ** Research Assistant. * Hybrid Propulsion Engineer, Advanced Propulsion Systems. Member AIAA. t Hybrid Propulsion Engineer, Advanced Propulsion Systems. Copyright

Regression-Rate and Heat-Transfer Correlations for Hybrid Rocket Combustion

A data analysis program was developed to correlate the experimental solid-fuel regression-rate and thermal-pyrolysis data obtained in a lab-scale hybrid motor, a thermal-pyroly sis test apparatus, and a GC/MS system. The code employs an interfacial energy flux balance, various methods to determine the component heat fluxes due to radiation and convection, transient mass and energy conservation equations for the local gas properties, and chemical equilibrium computations for the flame properties. Separate dimensionless correlations were developed to accurately predict the regression rates in the upstream developing-flow region of the motor and the downstream developed-flow region. Radiation, variable fluid properties across the boundary layer, and axial variations in the gas temperature and velocity gradients had significant effects on the regression rate behavior in both regions. The radiant heat flux from soot contributed significantly to the overall surface heat flux under relatively low mass flux and low O/F ratio conditions. Correlations were also developed for the deduced Stanton and Nusselt numbers. The regression rate correlations predicted independent data from a tube burner to within ±5% error.