Modeling person-to-person contaminant transport in enclosed environments

Chun Chen, Purdue University


It is essential to predict person-to-person contaminant transport in enclosed environments to improve air distribution design and reduce the infection risk from airborne infectious diseases. This study aims to improve and accelerate the simulation of person-to-person contaminant transport in enclosed environments. This investigation first conducted experimental measurements of person-to-person contaminant transport in an office mockup and the first-class cabin of a functional MD-82 aircraft. The experimental data of steady-state airflow, temperature, and gas contaminant concentration fields obtained in the office were used to validate the steady-state computational fluid dynamics (CFD) models. In the aircraft cabin, the transient particle concentrations were measured at the breathing zones of each passenger. The experimental data were used for evaluating the transient particle models in this study. When applying the CFD models, most of the existing studies assumed that the index person coughed or sneezed directly without covering the mouth. In reality, however, people usually cover their mouths with a hand or a tissue when they cough or sneeze. Currently, no simple method is available in the literature for predicting the exhaled airflow from a cough with the mouth covered. Therefore, this study developed simplified models for predicting the airflow on the basis of the smoke visualization experiment. This investigation then applied the developed simplified models to assess the influence of mouth coverings on the receptor's exposure to exhaled particles. It was found that covering a cough with a tissue, a cupped hand, or an elbow can significantly reduce the horizontal transport of exhaled particles. As a popular particle model, the Lagrangian model needs to track a large number of particles in the calculations in order to ensure accuracy. Traditionally, modelers have conducted an independence test in order to find a reasonable value for this particle number. However, the unguided process of an independence test can be highly time-consuming. Therefore, this investigation developed a method for estimating the necessary particle number in the Lagrangian model. The results show that the proposed method can estimate the necessary particle number with a reasonable magnitude and thus reduce the effort that is normally required for evaluating different numbers of particles in order to achieve statistically meaningful results. Moreover, the superimposition and time-averaging method was proposed, which can reduce the necessary particle number, and, as a result, the computing cost can be further reduced. Although the traditional Eulerian and Lagrangian models can provide informative results of transient particle transport indoors, they are considerably time-consuming. Thus, this study further developed a new particle model on the basis of a Markov chain frame for quickly predicting transient particle transport indoors. When solving the particle transport equations, the Markov chain model does not require iterations in each time step, and thus it can significantly reduce the computing cost. The validation results show that, in general, the trends in the transient particle concentration distributions predicted by the Markov chain model agreed reasonably well with the experimental data. Furthermore, the Markov chain model produced similar results to those of the Lagrangian and Eulerian models, while the speed of calculation increased by at least 6 times in comparison to the latter two models for the studied case. To further identify a suitable model for indoor transient particle transport simulations, this study systematically compared the Eulerian, Lagrangian, and Markov chain models in terms of performance, computing cost, and robustness. This investigation used four cases, including three cases with experimental data, for the comparison. The comparison shows that all the three models can predict transient particle transport in enclosed environments with a similar accuracy. With the same time step size and grid number, the Markov chain model was the fastest among the three models. Unless super-find grid was used, the Eulerian model was faster than the Lagrangian model. The Eulerian and Lagrangian models were more robust than the Markov chain model, because the Markov chain model was sensitive to the time step size.




Chen, Purdue University.

Subject Area

Architectural|Civil engineering|Mechanical engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server