Constrained Multi-Objective Optimization of Membrane Dehumidifier for Air Cycle Machine Icing

Membrane dehumidification is of interest to increase air cycle machine cooling capacity while decreasing the risk of flowpath icing. A key concern of this approach is the sweep requirement of membrane modules, with all prior investigations using product sweep which significantly reduces flow rate. In an air cycle application, reduced flow rate must be avoided to maintain cycle performance. Thus, alternative sweep sources, such as the system exhaust air, must be evaluated. For this study, a custom membrane dehumidifier component model is exercised to explore the membrane dehumidifier module design space via constrained Monte Carlo simulation. From this, the Pareto front maximizing efficiency and minimizing volume is obtained for a range of membrane permeabilities to highlight the benefits of material and manufacturing advancements. The most efficient module is integrated into three air cycle machine subsystem architectures: a baseline system, a membrane-only system, and a combined system. Steady-state results show that the membrane-only system provides the greatest performance improvement, boosting cooling capacity by an average of 68.5% over the baseline. Alongside increased cooling capacity, other saturation and control metrics are also discussed, demonstrating that membrane dehumidification enhances performance and reduces icing risk at low altitudes.