Neutralization of Chemical / Biological Hazards Using Turbulent Mixing


U.S. Department of Defense: Defense Threat Reduction Agency (DTRA)

Objectives

The objective of this study is to devise accurate techniques for predicting turbulent mixing with a view toward estimating the effectiveness of strategies for neutralizing bio/chem hazards. One scenario assumes scalar plumes emanate from multiple sites within confined or open air domains.  In this case, Vorcat is better positioned to model the essential effect of small scale vortices on scalar mixing than traditional grid-based closure and large eddy simulation techniques relying on unphysical diffusive models.

We conducted a study of turbulent mixing phenomena in the context of a model problem displaying many of the essential aspects occurring under more general circumstances.  The geometry of our model problem is sketched in Figure 1 whereby a vertical jet containing species B is simulated over a sufficiently long time period to establish a turbulent equilibrium condition in a “test section" of the computational domain. The latter consists of a rectangular parallel piped domain over the vertical jet within which the vortex element and scalar species particles reach an essentially equilibrium state, i.e., there is no significant drift in the mean velocity or scalar statistics.

With the vertical jet flow as a background, a puff of species A is  injected into the test section in the form of a horizontal jet positioned to the side center of the test section.  During the finite duration of the puff jet, N = 10 particles per time step of species A are put into the flow. The jet runs for 100 time steps, giving a total duration of ∆t = 0:1. In this a total of 1000 species, A particles are put into the flow. Several independent realizations of the experiment are obtained by releasing the puff at different times in the simulation.

A key interest in this study is in developing a means for quantifying turbulent mixing. We adopted a methodology which is particularly well suited for our grid-free Lagrangian simulation. Several aspects of measuring mixing for both single species and two species are discussed in detail in our Technical Report. Sample results  based on probability density function (pdf) of minimum distance between particles are shown below, followed by conclusions.

Sample Results

Geometry of model problem showing test section. The only boundary condition is the assigned jet and puff velocities

Geometry of model problem showing test section. The only boundary condition is the assigned jet and puff velocities

Scalar particles at indicated times in the region 2:833 < z < 3:1667 of the test section: species A (red) in the horizontal puff; species B (cyan) in the vertical jet.

Scalar particles at indicated times in the region 2:833 < z < 3:1667 of the test section: species A (red) in the horizontal puff; species B (cyan) in the vertical jet.

Mean of Probability Density Function (pdf) of particle separation: ----, coarse simulation; - - -, fine simulation.

Mean of Probability Density Function (pdf) of particle separation: ----, coarse simulation; - - -, fine simulation.

 
Scalar particles at indicated times in the region 2:833 &lt; z &lt; 3:1667 of the test section: species A (red) in the horizontal puff; species B (cyan) in the vertical jet

Scalar particles at indicated times in the region 2:833 < z < 3:1667 of the test section: species A (red) in the horizontal puff; species B (cyan) in the vertical jet

pdf at t = 1:22, ----; and t = 1:24 - - -.

pdf at t = 1:22, ----; and t = 1:24 - - -.

 

 

 Conclusions

We conclude that Lagrangian mixing measures built around the pdf of particle separation offer a promising avenue to categorize turbulent mixing in engineering applications. We have also demonstrated several of its advantages including:

  1. its capacity to remain free of random variations affecting other mixing measures,

  2. its straightforward generalization to mixing between any number of species,

  3. its sensitivity to subtle changes in mixing levels such as was witnessed in the jet expansion,

  4. its responsiveness to gross changes in mixing as was seen in the puff/jet interaction.

Besides these advantages, the Lagrangian approach opens up the door to a wide range of more advanced statistics which can be made instrumental in the formulation of turbulent combustion and other chemical interaction models. For example, the residence times of close interaction between individual particles of differing species can be monitored, and then used to affect the elimination or modification of particles. From this the effectiveness of mixing in neutralizing a toxic species can be modeled. Clearly, the range of such studies is unlimited - only requiring the establishment of particular guidelines based on the particular chemistry of interaction.