L. J. Forney

Three-wire Thermocouple: Frequency Response in Constant Flow

Theory and experimental measurements are compared with a novel three-wire thermocouple. Signals from three wires of unequal diameters are recorded from the thermocouple suspended in constant flow with a periodic temperature fluctuation. It is demonstrated that the reconstructed signal from the three-wire thermocouple requires no compensation for to_<5to1, where ,oI is the natural frequency of the smaller wire. The latter result represents a significant improvement compared to previous work with two-wire thermocouples. A correction factor has also been derived to account for wires of arbitrary diameter.

Computational simulation of turbulent mixing with mass transfer

Abstract In this paper, we present a series of computational simulations of three-dimensional turbulent mixing with mass transfer for various pipe mixing arrangements. The simulations are carried out with the adina software, in which general-purpose finite element and finite volume formulations along with the k – ϵ turbulent model are used for incompressible Navier–Stokes flows with mass transfer. Based on the predicted pressure and velocity profiles and the standard deviation of tracer (or fiber) spatial distributions at certain distances downstream from the injection point, we compare the mixing performances of various transverse, concentric, and multijet mixers as well as four silo mixing units. In addition, we deduce certain design information pertaining to different mixing configurations.

Turbulent Jet Reactors

Reactor scale-up and numerical simulation of reactor performance require an estimate of the time scale of turbulent mixing. In the present paper, time scales for mixing within free turbulent jets are discussed. In particular, it is demonstrated that both the Kolmogoroff time ( v/ ɛ) 1/2 and eddy dissipation time k /ɛ are proportional throughout the isotropic flowfield of fully developed free jets that are commonly used in reactor design. The constant of proportionality between both times is shown to depend on the jet Reynolds number. Scaling laws are derived in the limiting case of either jet-to-tube velocity u /v→1 or u /v→∞ plus one intermediate value u /v→3. Both time scales are also compared with the yield from a mixing-controlled parallel reaction occurring within a free jet where it is demonstrated that the smaller Kolmogoroff value correlates the yield in the limit of a small scale reactant feed. The eddy dissipation time, however, is shown to correlate data for larger but fixed reactant feed streams, but a complete correlation was not possible.

Numerical Study of Multi-Jet Mixing

The mixing process of turbulent streams within a pipeline due to single and multiple transverse jets is analyzed in this paper. The process is simulated using the incompressible Navier-Stokes equations with mass transfer and the standard k – ϵ turbulence model. The mixing effectiveness is characterized by evaluation of the second moment of mixing at various downstream locations. The numerical results for single and double jets agree well with both existing experimental data as well as semi-empirical expressions. Using the simulated data, a new correlation is developed to predict the mixing performance of both single and multiple jet configurations. The numerical results suggest that multiple jets ensure better mixing in addition to decreasing power requirements with an increasing number of jets. In particular, the case of four jets ( n = 4) provides distinctly superior mixing compared to n

Measurement of frequency response in short thermocouple wires

Experimental measurements are made for the steady‐state frequency response of a supported thermocouple wire. In particular, the effects of axial heat conduction are demonstrated for both a supported one material wire (type K) and a two material wire (type T) with unequal material properties across the junction. The data for the amplitude ratio and phase angle are correlated to within 10% with the theoretical predictions of Fralick and Forney (1992) for a laser‐heated junction with no bead. This is accomplished by choosing a natural frequency ωn for the wire data to correlate the first‐order response at large gas temperature frequencies. Moreover, it is found that a large bead size (roughly twice the wire diameter) will increase the amplitude ratio at low frequencies and distort the phase angle. In the latter case, it is necessary to increase the thermocouple wire length by 20% in the theory to correlate the data when the junctions have beads (silver soldered).

Aerosol fractionator for large-scale sampling.

The design of a versatile instrument for aerosol assessment is described. The single stage device incorporates the principal of virtual impaction and a continuously variable slit width to provide for ease of calibration, low internal wall losses, and potential for large‐scale sampling, as well as the option of either fractionating particles into two size classes for mass or chemical analysis or determining entire particle size distributions. Initial tests indicate the efficiency curve of the device is solely a function of particle Stokes number ψ for typical geometries and flowrates and is characterized by the cutoff value ψ1/250=0.36 ±0.04 and a steepness of ψ1/284/ψ1/250 =1.32±0.06. Moreover, wall losses are shown to peak at less than 15% at the critical Stokes number ψ1/250, which is an improvement over existing devices now used for large‐scale sampling.

Jet Trajectories of Transverse Mixers at Arbitrary Angle in Turbulent Tube Flow

Anumerical procedure is presented to evaluate the tracer trajectory that results from single jet injection into a pipe at arbitrary angles. Numerical and asymptotic results from the jet conservation expressions, as well as empirical formula are compared with experimental data. It is found that the numerical procedure provides an accurate prediction of the jet trajectory, particularly, when the injection angle is either normal or upstream. Moreover, the numerical solutions provide estimates of both the jet diameter and concentration (or temperature) decay. Example trajectories are computed for a range of injection angles 30° ≤ θ 0 ≤ 150° and the profiles are found to be insensitive to the jet-to-pipe velocity ratio when jet coordinates are scaled with the jet momentum length.

Multiwire thermocouples in reversing flow

Measurements are recorded for multiwire thermocouples consisting of either two or three wires of unequal diameters. Signals from the multiwire probe are recorded for a reversing gas flow with both a periodic temperature and time constant fluctuation. It is demonstrated that the reconstructed signal from the multiwire thermocouple requires no compensation provided ω/ω1<2.3 for two wires or ω/ω1<3.6 for three wires where ω1 (=2πf) is the natural frequency of the smaller wire based on the maximum gas velocity. The latter results were possible provided Fourier transformed data from the wires were used and knowledge of the gas velocity phase angle was available.

Single jet mixing at arbitrary angle in turbulent tube flow

An asymptotic procedure is presented to evaluate the tracer trajectory in a two-stream turbulent pipe mixing unit with an oblique branch. The proposed mixing jet trajectory estimate near the injection point is compared with the existing experimental data and used to calculate the critical mixing configurations. In addition, it is shown that the well-known tracer jet profile can be recovered for the case of a normally issued tracer turbulent jet

UV Disinfection of E. coli Between Concentric Cylinders: Effects of the Boundary Layer and a Wavy Wall

UV inactivation of E. coli in a plug flow reactor between concentric cylinders was investigated. The concentration boundary layer thickness was computed for laminar, turbulent and Taylor-Couette flow in terms of the respective mass transfer Sherwood number. It is demonstrated that the concentration boundary layer is thin and that the mass transfer coefficient is large and comparable in size for both turbulent and laminar Taylor-Couette flow in contrast to laminar flow. Computation of the fluence distribution for each flow pattern indicate that turbulent and especially Taylor-Couette flow subject E. coli to an equal flux of photons corresponding to ideal plug flow. However, experiments with turbulent flow that require large axial velocities indicate that very long reactor lengths are necessary to inactivate E. coli. Finally, rotor wavy wall modifications are explored to increase the inactivation of microorganisms in Taylor-Couette flow.