Modeling and Simulation of Fabrication, Transport, and Combustion of Nano-Engineered Energetic Materials (NEEM) in Flow Environments Vigor Yang The Pennsylvania State University Presented at Nano-Engineered Energetic Materials (NEEM) MURI Kickoff Meeting University Park, PA 16802, September 14, 2004
Research Scope Study Nano-Engineered Energetic Materials (NEEM) behavior in two-phase flow environments at meso and macro scales Optimize NEEM fabrication techniques, especially the rapid expansion of supercritical solution (RESS) approach
Fabrication of Nano-Sized Energetic Materials Rapid expansion of supercritical solution (RESS) Supercritical anti-solvent precipitation (SAS)
Visualization of Near-Field Jets ( I ) Shadowgraph images of supercritical methane/ethylene jets at various P inj /P chm. x CH4 =0.9, T inj =T chm =300 K, and d=1.0 mm. P inj /P c 1.38 1.29 1.22 1.17 P inj /P chm 37.5 9.2 4.5 2.2 The supercritical jet behaves like a highly under-expanded ideal gaseous jet at high P inj /P chm with Mach disk and barrel shock visible in image. The structure of these shock waves gradually decreases in size as P inj /P chm decreases.
Visualization of Near-Field Jets ( II ) Shadowgraph images of supercritical methane/ethylene jets at various P inj /P chm. x CH4 =0.1, T inj =285 K, T chm =300 K, and d=1.0 mm. P inj /P c 1.11 1.07 1.04 P inj /P chm 34.0 8.1 2.0 The jets injected at a temperature close to the critical temperature exhibit opaque shadowgraph images, indicating two-phase mixtures within the jets. The dome-shaped near-field jets indicate the existence of shock structures, which are masked by the opaque appearances.
Visualization of Near-Field Jets ( III ) Shadowgraph images of methane/ethylene jets at various injection temperatures. x CH4 =0.1 and d=1.0 mm. P inj /P c 1.15 1.16 T inj /T c 1.23 1.03 P inj /P chm 35.7 36.7 γ (C p /C v ) 1.56 5.51 The T inj /T c =1.03 jet has a large jet expansion angle and an opaque appearance due to condensation. The difference in jet expansion angles may come from the substantial difference in specific heat ratios and the pressure rise due to the release of latent heat during condensation.
Isentropic Expansion Paths Entropy-pressure diagram for methane/ethylene mixture with x CH4 =0.1. 220 The path originated at a temperature close to the critical point can readily penetrate into the two-phase region during isentropic expansion. 200 CH 4 - C 2 H 4 MIXTURE, x CH4 = 0.1 T = 350 K 300 K VAPOR SUPER- CRITICAL FLUID T inj /T c =1.23 The expansion path initiated at a temperature away from the critical point may exhibit an idea gas expansion. Condensation phenomenon is more sensitive to injection temperature than injection pressure. S (J/Mole/K) 180 160 140 TWO-PHASE MIXTURE BUBBLE- POINT LINE DEW-POINT LINE LIQUID T inj /T c =1.03 C.P. 276.5 K 265 K 250 K 235 K 220 K 200 K 120 1 2 3 4 5 6 7 P (MPa)
Characteristics of Supercritical Fluid Jet p = 2.0 MPa p = 1.0 MPa p = 4.0 MPa p = 4.0 MPa Mayer et al. AIAA 1996-2620 T LN2 = 105 K T = 300 K GN2 u LN2 = 10 m/s D in = 1.9 mm Thermodynamic non-idealities and transport anomalies in transcritical regime - rapid property variations - large density gradient Diminishment of surface tension and enthalpy of vaporization Pressure-dependent solubility High Reynolds number
LES Formulation of Supercritical Fluid Dynamics Favre-filtered conservation equations ρ ( ρ u~ ) + = t x j ( ρu~ j ) ( ρu~ iu ~ j + pδ ij τ ij ) + = t x j ~ ~ ( ρe + q) [( ρe + P) u ~ j uiτ ij] + = t x j 0 ( R ij ( K + L x j ij j + Q x + C j J ij ) + q j ) Closure requirements Thermodynamic and transport properties Subgrid-scale turbulence interaction Chemical kinetics R, L, ω& i Z, C, µ, λ, D C p im
Density Gradient Field (p = 9.3 MPa,, T = 300 K, u in = 15 m/s, T in = 120 K, D in = 254 µm)
Comparison of Laminar Premixed Flames (Risha, et al., PSU) ~ 1 cm 1.9 cm methane / air aluminum/air aerosol flame (φ = 1) (φ = 1, β=309 g/m 3 )
Condensed-Phase Particle Analysis (Risha, et al., PSU) 1 Aluminum 0.8 0.6 0.4 0.2 0 Oxygen -r CL r
Aluminum Particle Models Numerical Analysis Nano-sized aluminum particles as large gaseous molecules to be regarded as a limiting case when particle size approaches zero. The particle laden-flow was modeled as a one-dimensional, laminar, steady flow of a premixed gas mixture at constant pressure. The model was solved with CHEMKIN software package and the PREMIX subroutine. Thermochemistry data for aluminum containing species was obtained from JANNAF database or Swihart and Catoire Transport Properties data from Svehla for Al, AlO and kinetic theory for AlOAl, OAlO, AlOAlO, Al 2 O 3 (l), AlH, AlOH, etc. Aluminum submodel consists of 9 species and 16 reactions and Al/HCl/H 2 O/CO 2 /O 2 mechanism from Swihart and Catoire Methane / air model from Kee et al. Analytical Analysis The analytical method extended from Goroshin et al. for fuel-lean mixtures. The particle burning times for micro-sized and larger particles are estimated using d 1.8 model by Beckstead.
Species and Temperature Profiles of 1D Stoichiometric Al-Air Air Laminar Flame 4000 0.8 3500 3000 N2 T 0.7 0.6 Temperature, K 2500 2000 1500 1000 500 S L =5.82m/s AL(L) O2 AL*20 ALO*20 ALOAL*20 AL2O3(L) 0.5 0.4 0.3 0.2 0.1 Mole Fractions 0 0.04 0.05 0.06 0.07 x(cm) 0
Species and Temperature Profiles of 1D Stoichiometric Al-H 2 O Laminar Flame Temperature, K 4000 3500 3000 2500 2000 1500 1000 500 S L =0.22m/s H2O T H2 AL(L) AL*50 ALOH AL2O3(L) ALO*50 ALOAL 0 0.2 0.4 0.6 0.8 x, cm 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Mole Fraction
Effect of Particle Diameter on Flame Speed (Huang, et al., PSU) 10 1 10 1 φ =0.8 10 0 10 0 S L,m/s 10-1 10-1 H2O Air 10-2 10-2 Analytical prediction Numerical prediction Experimental data 0 10-6 10-5 10-4 particle diameter, m
Summary -- Experimental Findings Preliminary Experimental Findings Measured laminar flame speeds of Al particle/air mixtures were found to be independent of equivalence ratio as also found by Goroshin et. al. due in part to a small variation of flame temperature. Flame speeds asymptotically increased with increasing oxidizer shear velocity. This is thought to be caused by the enhanced particle break-up due to the higher jet momentum. The effect of hydrogen addition was studied to provide some high-temperature steam in the aerosol. The addition of H 2 caused the flame speed to increase due to its low molecular weight. Thermophoretic sampling was performed to determine the transition of Al to Al 2 O 3 through the flame zone. From the data analyzed, high-resolution micrographs showed the inner portion of the flame consisted of 5-8 mm aluminum particles and its outer flame consisted of smaller stoichiometric Al 2 O 3 particles.
Summary -- Theoretical Findings Initial Theoretical Findings In the molecular limit, Al flame speeds with air are predicted to be significantly higher than those with H 2 O, as a result of the greater reaction rate of Al with O 2 versus H 2 O. The kinetic bottleneck results from the AlOH formation in the Al/H 2 O system. For nano-sized particles, the Al/H 2 O flame speed is predicted to be slightly greater than that of Al/air, which is due to the high diffusivity of H 2 in the products.
Combustion-Wave Structure of HMX/GAP Pseudo-Propellant Propellant T sf = ~ 2000 K Rapid Consumption of HCN and NO Secondary Flame Zone T dark zone = ~ 1250 K Decomposition, Evaporation, and Gas-Phase Reactions (Bubble) T s = ~ 700 K T melt = 558 K Major Species in Dark Zone: N 2, H 2O, NO, CO, HCN, H CO, N O, H, CO 2 2 2 2 GAP Polymer Residue Dark Zone Primary Flame Zone (HMX Vapor /Liquid Interface) Foam Layer (HMX Melt Front) Heterogeneous Solid Phase HMX GAP
RDX Spray with and without CO 2 Laser-Support (Tim Parr, NAWC) high CO 2 laser energy 1 mm low CO 2 laser energy CO 2 off
Program Structure and Interactions Synthesis & Assembly PSU UIUC nano - macro Theoretical Modeling & Simulation USC PSU macro - nano NEEM Experimental Characterization & Diagnostics nano - macro UIUC PSU
Shock Precipitation (SP) Processing of Nano-sized Oxidizers (Ken Kuo,, PSU) In shock precipitation, the energetic material is dissolved in a liquid solvent at ambient or elevated temperature conditions and ambient pressure. Generally, a solvent in which the material is very soluble (e.g., acetone) is selected to maximize the batch size and throughput. The oxidizer/solvent solution is then rapidly mixed with a liquid in which the oxidizer is not soluble (e.g. water). Fine particles form very rapidly upon mixing, and can then be separated from the liquid. Parameters that control particle size distribution will include: type of solvent, concentration, temperature, and rate of mixing/dispersing the solvent into the secondary fluid. Oxidizer crystals to be considered include: RDX, ADN (ammonium dinitramide), HNF and FOX-7 (1,1-diamino-2,2-dinitroethylene).
Supercritical Fluid (SCF) Processing of Nanosized Oxidizers (Ken Kuo,, PSU) One major advantage is the ease of separation of the SCF from the particles (and co-solvent, if used). Victor Stepanov of ARDEC has obtained mean RDX particle sizes of ~130 nm with narrow size distribution. Mr. Stepanov has expressed strong interest in collaborating with our planned effort to improve his existing methods for achieving high throughput of nano-energetic materials. Two SCF methods will be investigated for application to energetic materials: Rapid Expansion of Supercritical Solutions (RESS) material of interest is dissolved in a SCF. The solution is then expanded rapidly into a low-pressure chamber causing rapid nucleation of fine particles. Supercritical Anti-Solvent precipitation (SAS) addresses one of the main shortcomings of RESS (very low solubility of many materials in common SCF solvents such as CO 2 ) with the use of a second solvent having much higher solubility. Rapid mixing via opposed-jet impinging flows will be applied to increase the rate of nucleation and thus reduce the particle size while increasing yield.