Date of Graduation


Document Type


Degree Name

Master of Science in Chemical Engineering (MSChE)

Degree Level



Chemical Engineering


Tom O. Spicer, III

Committee Member

Heather Walker

Second Committee Member

Larry Roe


Environmental Dispersion, Gases, Hazard Assessment, Wind Tunnel


The 2015 Jack Rabbit (JR) II field trials were conducted to improve understanding of the denser-than-air dispersion of chlorine in an urban environment. The field study involved five large-scale, outdoor release trials of chlorine in a mock urban environment (MUE) at Dugway Proving Ground (DPG). Various instrumentation was deployed, including JazTM Ultraviolet-Visible (UV-Vis) point sensors capable of measuring gas-phase chlorine concentration throughout the mock urban array.

The Chemical Hazards Research Center (CHRC) was tasked by the Chemical Security Analysis Center (CSAC) with the Mock Urban Wind Tunnel (MUWT) program to study the 2015 JR II Trials at wind tunnel scale. A 50:1 physical model of the 2015 JR II Trials was developed. After the establishment of proper wind profile and source characterization, the physical model was prepared for model concentration measurements at Jaz locations in the scaled MUE.

Seven concentration-measurement studies (Tests A–G) involving either JR II Trial 2 or Trial 4 conditions were conducted. Each test used either a fast-response nondispersive infrared detector (Cambustion NDIR500) or flame ionization detector (Cambustion HFR400 FID) to obtain concentration measurements with carbon dioxide or methane as a chlorine cloud “tracer,” respectively. Two methods for achieving similarity were used: Froude (Tests A, B, and G) and Richardson (Tests C–F) number similarity. Froude similarity required that the source-gas density in the full-scale experiment (e.g., JR II) be identical to the source-gas density in the wind tunnel model. Achieving near source-gas density equivalence required the use of sulfur hexafluoride (a high molecular weight species) in tunnel source-gas mixtures. Richardson similarity relaxed the density-match requirement; wind tunnel source-gas densities (“high-density case” and “low-density case”) were less than full-scale source-gas densities, but a lower wind speed and increased scaled release duration were required to compensate for density deviations.

Tests A–T4NFr and B–T2NFr involved releases of a 1.787 g/L simulant gas mixture under (50:1) JR II Trial 4 and Trial 2 conditions, respectively. Tests C–T2FRi(H) (high-density case) and D–T2FRi(L) (low-density case) utilized Richardson similarity to compare with Test B–T2NFr and investigate the reliability of the Richardson similarity approach for different deviations in the release gas density. Likewise, Tests E–T2FRi(H) (high-density case) and F–T2FRi(L) (low-density case) utilized Richardson similarity to model (50:1) JR II Trial 2 Jaz profiles and investigate the reliability of the Richardson similarity approach. Test G was an extension of physical model capabilities and involved simulated releases of a 1.440 g/L simulant gas mixture under JR II Trial 4 atmospheric conditions.

Results from Tests A–G were transformed to field scale and compared with each other and JR II concentration-time profiles. Comparison studies showed that the physical model was capable of reproducing field-scale (JR II) and wind tunnel (validation) experiments. Similar durations of exposure and cloud times of arrival and departure confirmed that model gas cloud dynamics throughout the physical model were scaled correctly. The validity of Richardson number scaling was verified as well as the method to scale measured concentrations between gases of different molecular weights.