Varunkumar S

This paper is concerned with a new and novel approach to modeling composite propellant burn rate behavior. It is founded on the fact that composite propellant combustion is largely boxed between the premixed limits – of pure AP and fine AP-binder (HTPB, here) whose burn behaviors are taken as known. The current strategy accounts for particle size distribution using the burn time averaging approach. The diffusional effects are accounted for through a calibrated heterogeneous quasi-one-dimensional model (HeQu1-D for short) that allows for the flame temperature dependence on the local AP size-binder thickness geometry. Fine AP-binder homogenization is adopted as in recent models with refinement on the particle size as a function of pressure. The specialty of the present approach is that it invokes local extinction for fuel rich conditions for specific particle sizes when the heat balance causes the surface temperature to drop below the low pressure deflagration limit of AP; this feature allows for the prediction of extinction of propellant combustion. Combining these ideas into a MATLAB®calculation framework that uses a single dataset on properties of AP and binder consistent with burn rate vs. pressure of pure AP and fine AP-binder system allows for making the predictions of propellants with multiple particle sizes and different fractions. Comparisons of burn rate data over nearly thirty compositions from different sources appear excellent to good. It is found that it is important to treat the full particle size distribution to achieve better predictions. Low burn rate index (∼0.25) observed with addition of SrCO3is captured by extending the model to include the effect of binder melt; the gas phase effect is accounted for by calibration against catalytic effect on the fine AP-binder propellant. An interesting deduction from the model is that the temperature sensitivity of propellants should not exceed that of AP. The robustness of the current model and speed of determining the burn rate behavior allow for the possibility of determining the particle size distribution required to meet the burn rate specifications of a specific propellant for practical applications before actually embarking on making the propellant.

Combustion of powdered biomass fuels under Moderate to Intense Low-oxygen Dilution (MILD) conditions is investigated using a combination of experiments and CFD, and comparisons are presented with earlier studies on coal MILD operations. The biomass MILD regime is established by the recirculation of high-temperature product gases into the incoming reactants driven by the velocity differences between the central biomass carrying low-speed primary air jet (2.6 to 12 m/s) and a pair of off-axis asymmetric high-speed jets (supersonic). Temperature profile is used as a parameter to assess the combustion characteristics of the reactor. Experiments are conducted with a wide range of fuel feed rates and airflow rates, corresponding to an overall equivalence ratio (ϕ) of 0.9 to 4. The influence of primary jet velocity on biomass MILD operations is not as pronounced on reactor stability as observed with coal due to lower ignition times of biomass particles. The influence of high-speed jets flow rate on reactor stability is marginal with biomass as with coal. Extremely rich operating conditions (ϕ ∼ 3) resulted in reactor quenching, and uniform temperatures around 1200 °C were observed around stoichiometry. Results from the simulations performed using in-house unified ignition devolatilization (UID) model complement the experimental explorations. Temperature profiles from the computations are in close agreement with the experimental results (within 15%–20%). Ignition index (τ), a ratio of particle flow time (tf) to particle ignition time (tig), explains the differences in reactor operations with coal and biomass as fuels.

The Heterogeneous Quasi One-dimensional (HeQu1-D) model for AP/HTPB composite propellant combustion is extended to aluminized propellants. Following the serial burning approach of HeQu1-D, a statistical particle path through an aluminized propellant is taken to consist of aluminized binder-matrix coated AP particles of various sizes. Extending the idea of homogenization of fine-AP particles, aluminum particles are assumed to be homogenized with the binder-matrix. Large Al particles (nominal size ≥ 15 µm) in the binder-matrix are assumed to get ejected into the gas phase and hence do not contribute to heat feedback to the burning surface. With these modifications, the model is shown to accurately predict the burn rate variation with pressure and initial temperature of propellants using Al of nominal sizes in the range of 15–50 µm (termed conventional aluminum, CAl). The experimentally observed reduction in propellant burn rate with substitution of AP by CAl is shown to be due to the increase in fuel richness and energy sink effect of melting of Al. Catalytic effects due to addition of burn rate modifiers (Fe2O3 in space shuttle booster propellant, for instance) are also accounted for by modifications to gas phase rate parameters. Increase in burn rate up to 25% with Al particles of a few µm (3–6 µm) in comparison with non-aluminized propellant is explained by recognizing that there will be non-negligible fraction of sub-micron Al due to the associated particle size distribution. This leads to heat release by sub-micron Al combustion close to the surface and an enhancement in the heat feedback, by a combination of convective and radiative mechanisms. Dramatic increase in burn rate with sub-micron Al (a factor of 4–5 as compared to CAl) is captured by invoking radiation from fine-Al/Al2O3 particles to propellant surface. Agglomeration of sub-micron Al particles is invoked to explain the saturation of burn rate enhancement with reduction in Al particle size and the effectiveness of sub-micron Al substitution in smaller fractions (bi-modal Al). Predictions for over fifteen different propellants with varying fractions of fine and coarse aluminum from earlier literature have been presented. Comparisons of the predictions from the model with experimental results from different sources is shown to be excellent-to-good for most cases. The hitherto unknown radiation dominated ablation (r˙>50mm/s) of coarse AP particles and the effect of Al particle size on propellant temperature sensitivity are brought out. The MATLAB® code based on the model offers opportunity for design of AP-HTPB composite propellants with combination of fine and coarse aluminum with confidence.

The present study focuses on the experimental determination of the global extinction strain rate (ag) for different syngas-air combinations using the Tsuji type configuration. To study the effect of porous burner diameter (D), ag values were obtained for four values of D at atmospheric pressure. The experimentally obtained ag for a given fuel-oxidizer combination decreases with an increase in burner diameter (D). This trend is consistent with the limited data available in the literature for hydrocarbon fuels. Other geometric and flow-field effects namely, (1) plug flow, (2) flow-field blocking by the burner, and (3) heat loss by the flame to sidewalls that can affect ag were also experimentally quantified. The results from this study show that the plug flow boundary condition is always satisfied for oxidizer inlet distance > 2 times the largest porous burner diameter. Burner diameter less than 1/4 times side wall length (as is the case for all burners used in this study) does not significantly modify the flow. Hence, these two flow-field modifications do not affect ag. However, heat loss from the flame to the ambient through the side walls can cause a 4–9 % decrease in ag. Experiments showed that, CO/H2 mixtures diluted with N2 yield 1.6–2.25 times higher ag in comparison to CO/H2 mixtures diluted with CO2. Increasing H2 from 1 to 5 % leads to 2.5–3.8 times increase in ag, compared to 5 to 10 % increase in H2 which leads to only 1.3–1.7 times increase in ag for 70 % of N2 (v/v) in fuel mixture. Global extinction strain rate (ag) increases by 1.5–2.4 times with 10 % increase in CO for fuel mixtures consisting of H2 (1 and 5 % by v/v), CO2 (50, 60 and 70 % by v/v) and N2 (50, 60, 70 and 80 % by v/v). The change in overall reactivity (ωo) due to different diluents is used to quantitatively explain the variation of ag for different fuel compositions. These effects are also qualitatively explained using OH radical concentration change with H2 % in the fuel mixtures.

In this chapter fundamental results from recent studies on packed-bed gasification of ligno-cellulosic biomass are presented. Theory based analysis tools developed to enable interpretation of the experimental results for a wide range of oxidizer combinations (mixtures of O2 –CO2 –steam–N2 ) and different types of biomass (pelletized agro-residues, wood chips, coconut shells etc.) are also included. In spite of being a well developed configuration for gasification reactors, several fundamental results on packed-bed biomass gasification had remained unknown till the completion of the recent studies described in this chapter. One such finding is the role of volatiles stoichiometry on the flame propagation rate, CO2 and net-steam conversion, syngas and char yield and levels of higher hydrocarbons (aka tars) in the syngas. That the transition from gasification to char oxidation regime roughly coincides with the volatiles stoichiometry point is a result of significant practical importance. Other fundamental results of significance include (1) the invariance of H2 yield (around 40 g/kg of biomass) over a wide range of volatiles equivalence ratio (2.5–1.2) under oxy-steam gasification operation, (2) identification of stagnation point flame extinction/transition as the mechanism of transition from gasification to char oxidation regime and (3) elucidation of the role of ratio of ignition to de-volatilization times on the stability of flame propagation. The implications of these results for existing practical systems and possible new technologies for value addition to ligno-cellulosic biomass are outlined.

This paper presents a scalable ultra-rich carbonization process in counter-current packed bed (CCPB) for synthesizing biochar/charcoal with a yield of more than 33% and fuel consumption rate of more than 200 g/m2s (which is at least 2 times more than the maximum achievable with conventional gasification) from coconut shells (CS). The novelty of the process lies in the use of reactant pre-heating to slightly below the devolatilization temperature of CS as a strategy to achieve steady and controlled conversion of CS to biochar/charcoal. The basis for this idea is that the ignition time (tig) of biomass particles decreases significantly with pre-heating while the time for devolatilization (tv) remains more or less the same. Single particle experiments with CS as fuel and atmospheric air as oxidizer show that tv/tig increases from 1.3 to 7.7 with a corresponding increase in pre-heating temperature Ti from 30 to 170 ∘C. The bed operation, in terms of flame propagation, equivalence ratio, and charcoal yield, with pre-heating to slightly below devolatilization temperature in CCPB systems, shows several interesting behaviors not observed in traditional gasifier systems. Two prominent examples of such behaviors are steady flame propagation under extremely rich conditions (volatile equivalence ratio, ϕv > 10) and reduction in peak bed temperature (Tpb) well below 600 ∘C. The biochar yield of the developed process is at least ∼ 25% more than traditional CCPB gasification system. The feedstock flexibility, self-sustainability, and scalability of the developed process are also brought out.

This paper is concerned with a new and novel approach to modeling composite propellant burn rate behavior. It is founded on the fact that composite propellant combustion is largely boxed between the premixed limits – of pure AP and fine AP-binder (HTPB, here) whose burn behaviors are taken as known. The current strategy accounts for particle size distribution using the burn time averaging approach. The diffusional effects are accounted for through a calibrated heterogeneous quasi-one-dimensional model (HeQu1-D for short) that allows for the flame temperature dependence on the local AP size-binder thickness geometry. Fine AP-binder homogenization is adopted as in recent models with refinement on the particle size as a function of pressure. The specialty of the present approach is that it invokes local extinction for fuel rich conditions for specific particle sizes when the heat balance causes the surface temperature to drop below the low pressure deflagration limit of AP; this feature allows for the prediction of extinction of propellant combustion. Combining these ideas into a MATLAB® calculation framework that uses a single dataset on properties of AP and binder consistent with burn rate vs. pressure of pure AP and fine AP-binder system allows for making the predictions of propellants with multiple particle sizes and different fractions. Comparisons of burn rate data over nearly thirty compositions from different sources appear excellent to good. It is found that it is important to treat the full particle size distribution to achieve better predictions. Low burn rate index (∼0.25) observed with addition of SrCO3 is captured by extending the model to include the effect of binder melt; the gas phase effect is accounted for by calibration against catalytic effect on the fine AP-binder propellant. An interesting deduction from the model is that the temperature sensitivity of propellants should not exceed that of AP. The robustness of the current model and speed of determining the burn rate behavior allow for the possibility of determining the particle size distribution required to meet the burn rate specifications of a specific propellant for practical applications before actually embarking on making the propellant.

In this paper the phenomenon of flame jump vis-a-vis steady propagation in biomass packed beds in counter-current mode is discussed. By analyzing the fuel flux and propagation rate data from experiments with a range of oxidizers, namely, air, O2-N2, O2-CO2, and O2-steam mixtures, parameter regimes of steady propagation, and flame jump are identified. A theoretical basis for this classification is developed by analyzing the thermo-chemical conversion of single particles subject to flow and thermal conditions in a packed bed. The ratio of the ignition ((Formula presented.)) to devolatilization ((Formula presented.)) times is shown to emerge as the controlling parameter in determining the flame propagation regimes. It is found from the theoretical analysis that steady propagation occurs for (Formula presented.) < 2 and transition to flame jump occurs if (Formula presented.) 2. Operational zones of a packed bed biomass system is mapped using the predicted ratio of (Formula presented.) as a function of volatiles-based equivalence ratio ((Formula presented.)). Implications of these results to practical ligno-cellulosic biomass combustion and gasification systems, especially using oxygen-steam mixtures for hydrogen generation, are brought out.

A gasification-based single step process is reported for synthesizing moderately activated char (surface area in the range of 100–600 m2/g) from ligno-cellulosic biomass. The novelty of the process lies in the use of mixtures of O2-CO2 and O2-steam as gasification agents to obtain significantly higher activation in a single-step as compared to gasification with air (<100 m2/g) in a self-sustained process. Experiments were conducted in a counter-current packed bed reactor with three biomass (agro-residue pellets, wood based pellets and coconut shells). Yield and Brunauer-Emmett-Teller (BET) surface area are reported for char obtained from experiments covering an overall equivalence ratio range of 3.9–1.2. The volatiles equivalence ratio (ϕv), identified as the unifying parameter for analysis from the earlier studies of the authors, emerges as the relevant parameter to characterize the relationship between yield and activation as well. With mixtures of O2-CO2, as ϕv decreases from 2.7 to 1.1 (fuel-rich to stoichiometric) the yield decreases from 100 to 10% (of fixed carbon content) with a corresponding increase in BET surface area from 10 to 570 m2/g. With mixtures of O2-steam, as ϕv decreases from 2.1 to 1.04 the yield decreases from 92 to 5% with a corresponding increase in BET surface area from 106 to 561 m2/g. The zone of transition of activation from reaction rate limited to diffusion limited conditions is identified from the experimental results. It occurs at around ϕv ~ 1.7 (T ~ 1450 K) for O2-CO2 mixtures and around ϕv ~ 2.1 (T ~ 1200 K) for O2-steam mixtures. The activation obtained with the current single step process is an order of magnitude higher compared to that from other single step processes like pyrolysis and air-gasification. Char obtained from the current process can be used as activated biochar and for other applications requiring only moderate activation (100

In this paper, experimental investigations on oxy-steam gasification of biomass in downdraft packed bed reactor are presented; propagation regimes and the associated hydrogen yield will be the focus of the current work. Steady flame propagation is shown to be established for a range of oxygen mass flux (16–120 g/m2s) covering ‘gasification’ and ‘combustion’ regimes with O2 fractions of 23, 30 and 40% by mass (rest steam). By restricting the upstream bed temperatures between 120 °C (to avoid steam condensation) and 150 °C (to prevent bulk devolatilization of bed), the intrinsic H2 yield from biomass is determined over an equivalence ratio (Φ) range of 3.5 to 1.2. This is shown to correspond to a ‘volatiles’ based equivalence ratio (ϕv) of 2.2 to 1. Interestingly, the H2 yield over this entire range is within 30–40 g/kg of biomass. Using equilibrium calculations, it is shown that the ‘unburnt volatiles’ is the major H2 source when ϕv > 2 and as ϕv → 1, ‘volatiles’ H2 drops close to zero and the major contribution is through the reaction, C + H2O → CO + H2. Increase in char conversion from about 20% at ϕv ∼2.1 to almost 100% as ϕv → 1, the corresponding increase in peak bed temperature and decrease in ‘higher hydrocarbons’ are consistent with the observed H2 yield. The important insight is, operating close to ϕv = 1, under slightly rich condition, leads to tar free exit gas with little or no compromise on H2 yield. This hitherto unknown result is perhaps the reason why all earlier works focused only on highly fuel rich conditions and/or very high steam temperatures (∼800 °C), tolerating higher tar content.