Session: 06-01-01 Computational Mechanics and Design Applications

Submission Number: 176738

Establishment and Verification of Preliminary Methodology for High-Fidelity Ammonia Diffusion Behavior Analysis 

Title: Establishment and Verification of Preliminary Methodology for High-fidelity Ammonia Diffusion Behavior Analysis

Authors: Hyung Yeon-woo*, Seo Young-gyun**, Jeong Jae-ho*
*Department of Mechanical Engineering, Chung-Ang University, Seoul, Republic of Korea
**Korea Research Institute of Ships and Ocean Engineering (KRISO), Daejeon, Republic of Korea
Corresponding author: jaehojeong@cau.ac.kr

Ammonia (NH3) is emerging as a carbon-free marine fuel aligned with the International Maritime Organization’s 2050 ambition. Because NH3 is toxic at ambient conditions, safe deployment on ships requires credible predictions of leak behavior across humidity levels. Full-scale trials in realistic compartments are difficult, which motivates physics-based computational fluid dynamics (CFD) methods that can be compared against controlled experiments. This study develops and validates a preliminary CFD methodology for predicting ammonia dispersion in humid air, focusing on the delay time before initial detection by a sensor and the subsequent concentration rise.

The computational domain reproduces a 0.5 m × 0.5 m × 0.5 m chamber used in companion tests, including a 4° floor gradient and a 23 mm circular release port. The gas mixture comprises air, water vapor, and ammonia. A structured hexahedral mesh of approximately 7.7×10^5 cells is employed, and a laminar model reflects quiescent conditions. Instead of prescribing a velocity inlet, the source boundary is modeled as an evaporative mass-flux condition implemented via a user-defined function (UDF). The mass flux is tied to an interfacial driving force between the liquid surface (30 wt% aqueous ammonia) and the surrounding gas, and is evaluated using local mixture properties. Relative humidity (RH) is introduced by converting RH to the initial vapor composition of the chamber and using it consistently in mixture-density and driving-force calculations.

Model validation uses controlled chamber measurements from KRISO and Chung-Ang University. Two metrics are compared across several RH levels: (1) the delay to first detection at a fixed sensor 0.5 m from the source, and (2) the post-detection concentration history. Geometry, thermophysical properties, and boundary conditions follow the experiment. A single parameter—the mass-transfer coefficient—is calibrated within a narrow, physically reasonable band. Simulations reproduce two key experimental trends: the detection delay increases approximately linearly with increasing RH, and the subsequent concentration rise slows.

The dominant mechanism couples early evaporation with surface thermal response. At high RH, latent cooling drives the liquid-surface temperature downward toward the dew point, promoting condensation on the ammonia solution. The resulting thin condensate layer increases interfacial resistance and suppresses the effective driving force for ammonia transfer into the chamber. This chain of effects delays the onset of detectable gas and moderates concentration growth even after detection. The proposed workflow captures these behaviors without artificial source terms by consistently treating humidity, mixture density, and interfacial mass transfer within the boundary condition.

The contribution of this work is a practical, humidity-aware CFD template for ammonia diffusion under quiescent or weakly ventilated conditions. The template comprises three elements: (1) specification of humidity-consistent gas composition and density; (2) an evaporative mass-flux boundary tied to interfacial driving force; and (3) validation against detection delays and concentration histories. Although demonstrated at chamber scale with laminar flow, the methodology is directly applicable to engineering studies of shipboard compartments and associated ventilation scenarios. Results provide inputs for detector placement and ventilation settings in shipboard compartments.

Keywords: ammonia, CFD, humidity, evaporation, condensation, diffusion, safety, shipboard leakage

Presenting Author: Yeon-Woo Hyung Chung-Ang University

Presenting Author Biography: Yeonwoo Hyung received a B.S. from Gachon University in February 2025 and is currently an M.S. student at Chung-Ang University (March 2025–present). He is a full member of the Korean Society for Fluid Machinery (KSFM).

Authors:

Yeon-Woo Hyung Chung-Ang University
Yeong-Gyun Seo Korea Research Institute of Ships & Ocean Engineering (KRISO)
Jae-Ho Jeong Chung-Ang University