Ulrich Bonne

Radiative Extinguishment of Diffusion Flames at Zero Gravity

The purpose of this study was to determine the influence of radiative heat transfer on the growth of small fires in zero-gravity environments such as spacecrafts and free-falling chambers. Flat diffusion flames were used to model such fires in the laboratory. Calculations in quantitative agreement with experiments confirmed that radiative extinguishment does occur, without first depleting available fuel and oxidant. Radiative heat transfer cools such flames within less than 1 see to the point of visible flame (soot luminosity) disappearance ( 900°K). The flame thickness reached at this point is smaller than 4 cm. Changes in the partial pressure of oxygen and in the type of hydrocarbon fuel used do not alter the extinguishment data appreciably. The pressure dependence of cooling time and flame thickness is weaker than p 1/2 . Very sooty diffusion flames will reach lower temperatures and become extinguished earlier than nonsooty flames; soot concentrations below a mass fraction of 10 −3 or a volume fraction of 10 −7 do not contribute significantly to the flame emissivity and cooling rate of the CH 4 flames used. Although a zero-gravity, steady-state solution exists for the diffusion equation in spherical coordinates, steady spherical diffusion flames, in the absence of convection currents, “exist” only when the fuel is provided in the form of particles that are small compared to the flame-quenching distance. The data obtained are consistent with previous work by Kumagai and Isoda on droplet combustion in free-falling chambers and by Kimzey et al. on extinguishment of small fires during short periods of weightlessness in aircraft flights of parabolic trajectory.

Actuation-based microsensors

We have evaluated fluid flow effects induced by off-the-shelf mini-actuators with available microstructure flow, pressure and temperature sensors and demonstrated the feasibility of exciting new sensing approaches. Specifically, we discuss new approaches for experimental and analytical determinations of: ? sub-millisecond flow sensor response time versus flow velocity and microsensor structure, ? compact and affordable composition correction, CV, for volumetric fluid flow sensors, ? concentration of binary mixtures, based on measurement of CV, and ? fluid properties based on actuator-induced flow or compression, such as viscosity or ? = cp/cv, respectively. The micromachined thermal flow sensors, i.e.?thermal microanemometers, that we used consisted of either: (a)?off-the-shelf, front-etched microbridge sensor chips of ~1.7?1.7?mm, with bridges of ~0.2?0.25?mm, or (b)?developmental, very rugged, MicrobrickTM sensor chips of equal size but without the etched cavities. For actuators, we used commercially available, 10-12?mm OD, membrane-based, low-cost, earphone speakers, with resonances in the 2?kHz region. We found the useful operating frequency range of both sensors and actuators, with due consideration to resonance effects, to be in the 40-100?Hz range and the one most free of disturbances for the actuators used. The flow sensors themselves showed the capability of operating beyond 500?Hz, especially the rugged version, which showed response times down to ~0.2 ms. This MicrobrickTM sensor is burst-proof and designed for operation in harsh environments featuring gas or liquid mass fluxes up to?500?g?cm-2?s-1, with condensible vapors and suspended sand or dust. With the above devices we demonstrated a new approach for on-line fluid flow sensor composition correction, which is needed to correct errors caused by fluid composition changes. Previously developed, time-consuming and costly composition correction for thermal flow sensors relied on either individual calibration or via measurement of thermal conductivity, specific heat and Prandtl number. Those methods can now be replaced by this one-step, on-line, low-cost, actuation-based normalization, which can be adapted as well to other flow sensing technologies, such as orifice flow sensors. Using the same mini-actuators to induce flows in laminar flow restrictors, we also report on the demonstration of a very compact and affordable approach to the measurement of viscosity, which is a coveted gaseous fuel property for feed-forward combustion control. The demonstration included the design and fabrication of associated circuitry to prove satisfactory operation after temperature cycles, shock and vibration, and to provide an accurate, temperature-compensated output, despite changes in supply voltage, gas pressure or temperature.

Sensing fuel properties with thermal microsensors

We report on measurements of combustion-relevant fuel properties for on-line, feedforward control with small, rugged and fully compensated microsensor-based systems. Such silicon microstructure sensor systems have been demonstrated to determine gaseous and liquid fuel properties such as stoichiometric oxygen demand, octane number, heating value, density and other properties of interest. The measurement approach consists of a three-step process: (1) Measurement of changes in electrical quantities when the sensing elements come in contact with the fluid, (2) Conversion of these quantities into primary sensor outputs, yi, such as thermal conductivity, specific heat, temperature and pressure, and (3) Correlation between these and the properties of interest, Y(yi). By coupling this property sensor to an equally rugged and small thermal flow microsensor, millisecond-range response time signals of mass or volume flow, or stoichiometric oxygen demand rate are provided for feed-forward control, without exposing any sensor to harsh exhaust gas environments. Having presented results with gaseous fuels elsewhere, we update these here but concentrate on the determination of octane and cetane number of liquid fuels. Analysis results show that the correlations between these combustion performance properties and physical fuel properties are as good as the ones between octane and critical compression ratio or between cetane and ignition delay. However, all those correlations appear to be limited presently by the accuracy or at least consistency of available data, which are needed for calibration of the sensor system. In checking the temperature dependence of one of the correlations for octane, we found the system output to shift by 15% when using hexadecane as a reference fuel, but only by 1% with iso-octane as reference, for a 10 degree(s)C shift in temperature.

Molecular Modeling of Analyte Adsorption on MEMS GC Stationary Phases

Future microelectromechanical systems (MEMS), nanoelectromechanical (NEMS), and micro-optical electromechanical systems (MOEMS) require distinct understanding of interfacial effects in order to predict their performance and to reliably manufacture these devices. We show here that molecular modeling offers a unique tool for simulating and understanding critical working interfaces by specifically modeling the atomic mechanics during performance. This paper offers examples of how molecular modeling may be used for improving materials used in MEMS devices using as example the comparative performance of materials for stationary phases in gas chromatographs. This comparison was based on derived interaction enthalpies between analytes and stationary phases and using simulations of surface separation by employing molecular dynamics. The separation performance was compared to experimental GC data., showing that qualitative comparison of separation was present from the molecular scale and confirming that molecular modeling may be a useful tool to pre-select stationary phases for specific activity

Control of overall thermal efficiency of combustion heating systems

A method for analyzing the energy performance of central combustion heating systems has been derived. The method consists of a combination of measuring a minimum number of key system parameters and using them as input for HFLAME, the digital model used to simulate the dynamic performance of the heating system. HFLAME is based on the stack loss method, uses only input that can be obtained during field measurements, and relies on the steady state operation of the burner to calibrate flows and temperatures. This methodology is being applied to the study of a variety of furnace/boiler designs, modes of operation and control options, aimed at reducing the energy consumption for residential space heating. One of the main reasons for the substantial, difference between steady state furnace efficiency and average seasonal system efficiency was found to be caused by the furnace draft flows during the burner off-period. Power burners can reduce the losses, LD, caused by those flows considerably by virtue of functioning like leaky dampers in the off-period. Other reasons for the generally high LD losses are: low average furnace load (oversizing), the use of pilot flame ignition systems, frequent system cycles and high fan/pump control set points. Although typical seasonal efficiency data for gas fired heating systems are about 60%, and 66% for those fired with oil, we estimate that efficiencies of up to 69%, respectively 76%, can be achieved by attending to careful selection, sizing, installation and adjustment.

Hydrocarbon Flames as Reference uv Light Sources

Premixed hydrocarbon–air flames stabilized on Meker burners are proposed for use as reference sources of ultraviolet (uv) light. A measure for the uv light intensity output is given by the flame chemi-ionization. The ratio of optical energy output per unit of electrical charge generated is a function of the wavelength band of the flame emission spectrum but is independent of size of the flame, applied electric field required for measuring flame ionization, burner–electrode configuration, temperature, and of small fluctuations of pressure and composition of the hydrocarbon fuel. An approximate, average value of 8 · 10−3 J/C was determined for the 2000–2100-A spectral band of a stoichiometric, premixed methane–air flame.