Thermal Management Investigations for Fuel Cell Systems On-Board Commercial Aircraft Hamburg University of Technology

Format:

Conference/Event

Author/Creator:

Vredenborg, Vredenborg, author.

Contributor:

Thielecke, Frank

Conference Name:

SAE 2013 AeroTech Congress & Exhibition (2013-09-24 : Montréal, Canada)

Language:

English

Physical Description:

1 online resource

Place of Publication:

Warrendale, PA SAE International 2013

Summary:

The integration of fuel cell systems as an independent energy source (Auxiliary Power Unit, APU) requires enhanced aircraft cooling architectures. New environmental control systems and systems with an increased cooling demand are investigated in various research projects. Cooling system architectures can be designed which benefit from similar requirements, e.g. by using the same cooling loops. Additionally, an increased cooling demand makes the investigation of alternative heat sinks necessary.For detailed system investigations simulation studies are used. A model library has been created in Dymola/Modelica containing the necessary component models to simulate cooling systems. The used modeling approaches and main model information are presented in this article.In order to understand the basic system behavior a Design of Experiment (DOE) is useful. If only two or three parameters are considered, simulation studies can be performed for each possible parameter combination. Analyzing more parameters with this method is limited by the computational expense. With Latin hypercube sampling methods or low-discrepancy sequences, sampling plans with reduced sample numbers can be created.Analyzing large simulation models and identifying the most influential parameters is a time consuming work. Sensitivity analyses are used to quantify the parameters influence on key performance indicators like mass and energy consumption. The applied sensitivity analyses are variance-based methods. By using sensitivity indices the influence of the input parameter variance on the output variance can be measured and it can be identified whether parameters are essential or unessential. A significant advantage of this variance-based method is the model-free approach

Notes:

Vendor supplied data

Publisher Number:

2013-01-2274

Access Restriction:

Restricted for use by site license