The objective of the project is to develop a tool for the inverse design of rotary positive displacement machines (RPDM) (e.g. twin-screw machines, rotary vane machines) with liquid injection on the basis of a chamber model simulation, which is provided to the participants of the Priority Programme for modelling the fluid energy machine within thermodynamic cycles (compressor or expander). Input parameters for the inverse design tool are the requirements from the selected process, i.e. in addition to the type of fluid and the mass flow rates, the fluid conditions at the inlet port, injection nozzle and outlet port of the RPDM (e.g. pressure, temperature, vapour quality). The design parameters to be inversely determined are the machine geometry and the speed. Within the tool the design parameters are optimized with regard to energetic efficiency. The optimization of machine geometry is not only limited to machine size, but also includes specific dimensionless geometric parameters of RPDM (e.g. build-in volume ratio, wrap angle, diameter to length ratio) with particular focus on the fluid dependency. Crucial factors in the design of the RPDM are known to be the correct determination of the fluid-dependent effects, such as the two-phase gap mass flow rates, the heat transfers due to evaporation or condensation, and the throttling effects in the inlet and outlet ports of the machine. The outlined objective is associated with various scientific challenges. On the one hand, it requires an automated chamber model generation for the RPDM. Abstraction of the geometry into capacities (e.g. volumes, machine parts) and their connections for exchange of mass and/or energy is needed. Furthermore, valid models for the exchange of mass and energy between the capacities are required (e.g. multiphase flows through gaps or valves/openings, phase change and heat transfer models). In this respect, one focus of the project will be the development of valid models for the effects of an injected liquid (mass flow rate of the working fluid or an auxiliary fluid, e.g. oil) into the working chambers of the RPDM. Questions about the temporal development of the liquid jet, the time-dependent distribution of the liquid in the rotating working chambers and the mass and heat transfer between gas and liquid phases are taken up and investigated experimentally using a generic test rig and accompanying numerically. With respect to the Carnot battery, the possibility to adjust the outlet temperature of the RPDM to a temperature requirement for the process by injecting a liquid phase - while maintaining an optimal pressure ratio for the machine - appears to be advantageous. 

Professor Dr.-Ing. Andreas Brümmer
Technische Universität Dortmund
Fachgebiet Fluidtechnik

We develop new simulation methods based on deep neural networks that shorten computation times with respect to traditional CFD considerably. In particular, we apply physics informed neural networks for a rough representation of fluid flows in turbo-machinery components operating in Carnot batteries. Such rough solutions are refined by conditional generative adversarial networks (GAN) in order to create realistic fine structure of turbulent flows. in particular, we study the physical and the generalization properties of such deep learning based solutions to fluid flow with respect to changing boundary conditions and geometry. This enables us to rapidly evaluate designs under strongly changing boundary conditions, as they aretypical for the discharging cycle of a Carnot battery. In this way, we provide an invaluable tools for the inverse design of turbocomponents for Carnot batteries. 

Professor Dr. Hanno Gottschalk
Technische Universität Berlin
Fakultät II - Mathematik und Naturwissenschaften
Institut für Mathematik

Professorin Dr. Francesca di Mare
Ruhr-Universität Bochum
Fakultät für Maschinenbau
Lehrstuhl für Thermische Turbomaschinen und Flugtriebwerke