Architecture and System Optimization


During architecture-level design, various options are examined to compare their benefits and drawbacks. Depending on the specifics of the problem, on each architecture option, a number of studies are performed, such as determining the dimensioning operating cases and loads, weight, and  preliminary safety analyses.
In this context, architecture and system optimization are used. Such methods deliver answers to the question for best possible design. In general, this covers the aspects of topology (“Which sub-systems and components are connected in what manner to each other in order to constitute the complete system architecture?”), sizing (“What are the internal and external geometries and dimensions of the relevant elements in the architecture?”), and operational management (“How is the system operated in all relevant operational conditions?”).
Decisions made during early design have an enormous impact on the performance and value of any product. Architecture and system optimization methods are a highly valuable in this regard, in particular for highly integrated systems. More Electric Aircraft are among the most challenging applications. Together with key industrial players, the team Energy Systems does research on generic methods and tools for this type of applications.

Virtual Iron Bird

The aim was to identify, optimize and validate innovative aircraft equipment contributing to the reduction in consumption of non-propulsive power.
A set of “feasible architectures” was considered and compared against a baseline. The main focus was on several more- and fully-electric aircraft architectures having engine-embedded starter/generators, reduced bleed or bleedless environmental control, reduced hydraulic or fully electrical flight control and landing gear actuation. All changes in non-propulsive power demand, weight, reliability, cost etc. were assessed.
For this purpose, the team Energy Systems set up a modeling and simulation environment, the so called Virtual Iron Bird, and exploited it for evaluation and optimization of new systems architectures at aircraft level.

Design Environment New ECS

In bilateral research with a strategic industrial partner, the team Energy Systems is developing a design environment for unconventional air conditioning and cooling systems onboard commercial aircraft (commonly referred to as Environmental Control Systems or Air Management Systems).
The environment contains dedicated modeling and simulation infrastructure as well as design and optimization tools. It provides models for conventional baseline ECS architectures as well as a modular toolkit to assemble unconventional architectures. The models allow evaluating potential ECS architectures in early design with respect to system weight, required block fuel and performance metrics such as electrical power demand or engine bleed air mass flow.
 Notional aircraft-level optimization results with respect to conflicting objectives
Building on the modeling layer, an optimization layer is used to establish best possible unconventional ECS architectures.

Electric Network Architecture Optimization

To improve the efficiency of the design process, an integrated tool based on modeling and simulation is developed by the team Energy Systems. Named as the Electric Network Architecture Design Optimization Tool (ENADOT), it is capable of evaluating architecture designs w.r.t. power behavior, dimensioning loads, weight, rejected heat (cooling demands), safety and reliability. ENADOT is implemented in the object-oriented, physical modeling language Modelica. It employs a steady-state modeling approach (architecture level).
Exemplary minimal path sets automatically deduced from a given topology of an electric power network
Diverse automatic analysis and optimization capabilities are incorporated, so that the different aspects of architecture design are covered in a combined manner.

Generator optimization

In European research projects and supported by industrial partners, the team Energy Systems is developing design and assessment methods for the electrical onboard network. This includes components for electrical generation, distribution and conversion. Trade offs between performance, weight, and latterly power quality have to be made and design choices have to be backed up by qualitative and quantitative measures. Especially the generator design has to face new demands and constraints for an altered power distribution network.
The technology portfolio contains a dedicated design and simulation infrastructure. This includes a design tool chain for an externally excited synchronous generator. Methods are both analytical and simulation based. For the modeling, there was put the emphasis on efficiency via right sized multi-level modeling and improved initialization methods.


URL for this article