John F. Widmann

Measurement of the Optical Extinction Coefficients of Post-Flame Soot in the Infrared

Abstract The optical extinction coefficients of post-flame soot have been measured in the wavelength range 2.8 to 4.1 μm. A laminar diffusion burner was combined with an infrared spectrograph and gravimetric measurements to determine the mass specific extinction coefficient, σs, and the dimensionless extinction coefficient, Ke. Using ethene gas as the fuel, the burner was operated at four global equivalence ratios (φ = 0.8, 1.0, 2.0, and 3.0) to examine the effect of the fuel-air ratio on the extinction coefficient. The extinction coefficient was found to decrease with increasing values of the global equivalence ratio for φ = 1.0, 2.0, and 3.0. The results for φ = 0.8 and φ = 1.0 were in agreement to within the uncertainty of the measurements. Measurements were obtained using propane gas as the fuel (φ = 1.0) and resulted in extinction coefficients equivalent to those of ethene. Transmission electron microscopy (TEM) images revealed differences in the morphology of the particles, consistent with the quantitative differences observed in the extinction data. The data indicate that the equivalence ratio has a strong effect on the optical properties of post-flame soot agglomerates.

Improving phase Doppler volume flux measurements in low data rate applications

Phase Doppler interferometry (PDI) measurements in low number density sprays necessitate a compromise between collecting a large number of samples for adequate statistics and practical data acquisition times. This paper investigates the effect of insufficient sample statistics on the calculated probe area, and the resultant uncertainty in the volume flux measurement. Several methods of improving the probe area calculation and volume flux measurement are investigated using experimental data obtained from water sprays produced by residential fire sprinklers. It is shown that the corrections result in statistically significant improvements in the volume flux measurements.

A Benchmark Experimental Database for Multiphase Combustion Model Input and Validation

Experimental data were obtained for the purpose of validating multiphase combustion models and submodels. A reference spray combustor was fabricated to permit well-defined input and boundary conditions, enabling measurements to characterize the fuel spray, combustion air, wall temperatures, gas temperatures, and species concentrations. The level of complexity of the reference spray combustor is consistent with a benchmark case in which the level of coupling between the flow field and the remaining physics (e.g., chemical reactions, heat transfer) is not as extensive as in the complete system, but is also not insignificant. The characteristics (i.e., size, velocity, volume flux, etc.) of the methanol fuel spray were determined using phase Doppler interferometry. Fourier transform infrared spectroscopy was used to measure the species concentrations in the reactor emissions, and the conversion of methanol in the reactor. Thermocouple measurements provide gas temperature data at the reactor exit. Particle image velocimetry was used to characterize the inlet combustion air, and also to provide validation data farther downstream.

Evaluation of the planck mean absorption coefficients for radiation transport through smoke

A literature review of experimental measurements of mass specific extinction coefficients, σ s for combustion-generated particulate has been conducted to obtain an improved value of the constant in the frequently used expression for the Planck mean absorption coefficient through smoke, s m = C 0 f v T . This expression, which results from the product of the spectral absorption coefficient for smoke particles and the Planck spectral distribution curve integrated over all wavelengths, is used to simplify radiative transport calculations in combustion systems. It is shown that a value of C 0 = 2370 - 240 m m 1 K m 1 is consistent with the available data and that this value is approximately two times the values commonly used. Using the available experimental values of σ s to calculate C 0 eliminates the problems associated with the large uncertainties in soot refractive index encountered when Rayleigh theory is used to predict C 0 . It is also shown that a value of K e = 8.8 - 0.9 for the dimensionless extinction coefficient is consistent with the available experimental data. These values are valid for stoichiometric and overventilated combustion only.

Aerodynamic study of a vane-cascade swirl generator

Abstract The air flow through a vane-cascade swirl generator is examined both experimentally and numerically to characterize the inlet combustion air flow entering a reference spray combustion facility at NIST. A three-dimensional model is used to simulate the aerodynamics in the 12-vane-cascade swirl generator that imparts the desired degree of angular momentum to the air in the annulus leading into the reactor. A numerical simulation using the renormalization group method (RNG) k – e turbulence model results in a velocity profile consistent with experimental measurements, and correctly predicts a recirculation zone that is experimentally observed at the exit of the annular passage. The standard k – e turbulence model does not compare as well with the experimental data and fails to predict the recirculation zone at the exit. This work is part of a larger project at NIST in which benchmark data are collected for input and validation of multiphase combustion models, and the results presented provide a well-characterized inlet condition for the spray combustion reactor. The good agreement between the experimental data and the simulation with the RNG k – e turbulence model provides further validation for this model in confined, annular flows.

Non-intrusive measurements in fire sprinkler sprays

The results of phase Doppler interferometry (PDI) and particle image velocimetry (PIV) measurements to characterize the water sprays produced by fire sprinklers are presented. The large size of the water drops produced by fire sprinklers, and the relatively large coverage area of the spray, present significant challenges when attempting to characterize these sprays. These difficulties are especially relevant when using PDI because large drops and large coverage areas may result in attenuation of the transmitting laser beams. For the fire sprinkler investigated, it was determined that attenuation of the laser beam resulted in over-counting of drops due to burst splitting, a situation in which the Doppler signals from single drops are misinterpreted as being due to multiple drops. This effect was minimized by carefully choosing the operating conditions of the PDI processing electronics. PIV measurements provide velocity profiles that can be used as input for fire dynamics simulations to predict the effect of sprinkler sprays on fires. The results from the PIV measurements show good agreement with the velocity measurements obtained from PDI in the core of the spray, but poorer agreement along the sprinkler axis. The discrepancy was attributed to recirculation zones present in the experimental facility and possible biasing of the PIV measurements towards the larger drops.

Benchmark Experimental Database for Multiphase Combustion Model Input and Validation: Characterization of the Inlet Combustion Air

The air flow through a vane-cascade swirl generator is examined both experimentally and numerically to characterize the inlet combustion air flow entering a reference spray combustion facility at NIST. A three-dimensional model is used to simulate the aerodynamics in the 12-vane cascade swirl generator that imparts the desired degree of angular momentum to the air in the annulus leading into the reactor. A numerical simulation using the Renormalization Group method (RNG) k-e turbulence model results in a velocity profile consistent with experimental measurements, and correctly predicts a recirculation zone that is experimentally observed at the exit of the annular passage. The standard k-e turbulence model does not compare as well with the experimental data and fails to predict the recirculation zone at the exit. This work is part of a larger project at NIST in which benchmark data are collected for input and validation of multiphase combustion models, and the results presented provide a well-characterized inlet condition for the spray combustion reactor. The good agreement between the experimental data and the simulation with the RNG k-e turbulence model provides further validation for this model in confined, annular flows.

A correction method for spray intensity measurements obtained via phase Doppler interferometry

A phase Doppler interferometer (PDI) was used to measure the intensity (i.e., the expected number of droplets measured per unit time) of a swirling methanol spray flame. Gaps observed in the data, which have been previously reported (McDonell and Samuelsen 1995), correspond to periods where the PDI was inactive. The intensity of the spray as a function of position within the flame was determined from the distribution of interarrival times (time between subsequent droplets entering the probe volume of the PDI), and a statistical simulation was developed to correct for the regularly occurring intervals in which the PDI does not accept data. The duration of this dead time was determined to be ca 5.2 ms for all points in the flame for which data were collected. The intensity of the spray varied by over two orders of magnitude throughout the flame. A statistical model is presented to correct for the observed behavior of the PDI, and the good agreement between the simulations and the experimental data indicate that the spray can be adequately modeled by Poisson processes.

PIV MEASUREMENTS OF DROPLET TRANSPORT IN A HOMOGENEOUS TURBULENT FLOW OVER A CYLINDER

Particle image velocimetry (PIV) measurements were carried out in the droplet-laden, homogeneous turbulent flow over both an unheated and heated cylinder, and body-centered cube (BCC) arrangement of spheres. Transport of both water droplets and aerosol particles was characterized upstream and downstream of these obstacles. Data were recorded for the cylinder at ambient temperature and after being heated to 423 K to estimate the effects of the hot cylinder surface on droplet transport. The results indicate that smaller droplets are entrained into the recirculation region behind the cylinder while the larger droplets impact the cylinder surface, accumulate and drip off, and/or rebound off the surface and disperse radially outward into the free stream. The Weber number was too low to lead to droplet shattering. Significant cooling of the cylinder surface from the initial preset temperature, which resulted from spray impingement, was observed for the heated cylinder, in addition to the presence of a vapor stream downstream of the cylinder along the shear layer region between the recirculation zone and free stream. For the BCC (with a blockage ratio of about 64 %), there was both transport of droplets and seed particles around and through the BCC, as well as significantly more liquid accumulation and dripping than for the cylinder.

A laboratory apparatus for the measurement of optical extinction coefficients of postflame soot in the infrared

An experimental apparatus for the measurement of optical extinction coefficients of postflame soot aerosol in the infrared region of the electromagnetic spectrum is presented. Reproducible soot aerosol is generated using a laminar diffusion burner, and the mass concentration of the aerosol is determined gravimetrically. The instrument, which utilizes an infrared spectrograph for the extinction measurements, provides data over the wavelength range 2.5–5.0 μm. The infrared spectrograph, which provides a near-instantaneous measurement of the incident radiation over a range of wavelengths, is ideal for applications in which the deposition of soot over time on the optical windows can interfere with the measurement. Furthermore, the wavelength dependence of the extinction coefficient can be determined independently of the gravimetric measurements, which represents the dominant source of uncertainty in the measurement of the mass specific extinction coefficient. The spectrograph reported herein has been characte...