Bayesian estimation of pseudo Langmuir isotherm parameters in adsorption beds under time-varying inlet concentrations

This study proposes a strategy that integrates a Bayesian estimation with a gas absorption CFD framework to reproduce the breakthrough behavior of sulfur mustard (HD) in swatch fabric gas absorption experiments under the time-varying inlet concentration conditions. The target reproduction range of breakthrough curves is defined between 5% and 95% of the maximum outlet concentration. The Linear Driving Force (LDF) kinetics model is employed, with Langmuir isotherm parameters treated as pseudo-parameters. The overall mass-transfer parameter, k_LDF, is calculated using two equations to clarify the role of pseudo-parameters via an evaluation function. Notably, while these equations omit explicit gas film resistance, the pseudo-parameters effectively compensate for this omission, resulting in a more accurate reproduction of the breakthrough curve. As the volume of data used for inference increased, the parameter estimation became more stable, yielding predictions that closely align with the swatch fabric breakthrough experiments. The proposed strategy efficiently predicts breakthrough curves with the target range for swatch fabrics and offers valuable insights for optimizing the design of chemical warfare-protective clothing.