To achieve net-zero greenhouse gas (GHG) emissions by 2050 it is necessary to introduce the use of sustainable fuels, such as biofuels, hydrogen, or Power-to-X (P2X) based on “e-fuels” in everyday life. Therefore, there is a clear need to develop advanced simulation software to assist in this transition, and to identify the technological developments for the new generation of combustors to operate more efficiently and with fewer emissions.
The simulation of turbulent combustion in realistic configurations is extremely demanding in terms of computational resources. A transition towards exascale architectures from High-Performance Computing (HPC) systems is foreseen in the next years, making massive numerical simulations possible. Nevertheless, the Computational Fluid Dynamics (CFD) codes need to optimize the existing algorithms or develop new models to be suitable for this kind of architecture.
CoEC´s work on simulation methodologies aims to enhance and develop mathematical and numerical methodologies enabled by the use of (pre-)exascale systems, with the focus on the applications defined by the Exascale Challenge Demonstrators.
Direct Numerical Simulations have the highest level of details, but they can be realized only for simple system. Nevertheless, they are essential to understand the physics behind the combustion process. Based on this research, innovative models are built so that simulations including less details, namely Large-Eddy Simulations, can be realized for more complex system.
This figure includes the images of a turbulent sooting flame realized at ITV and a combustion chamber by TUD
The project will create the tools to make such simulations possible to be performed. In details, the tasks include:
⦁ development of high-order spatial and time discretization formulations to be able to correctly characterize the propagating flame fronts;
⦁ development of Adaptive Mesh Refinement (AMR) techniques which allow for having a local refinement on the flame reaction zone;
⦁ development of more accurate and optimized dynamically adaptive chemistry models will allow to better predict the pollutant emissions, such as NOx and CO;
⦁ realize innovative solvers for the stiff systems of ODE representing the chemistry that will be efficiently coupled with CFD solver;
⦁ realize robust algorithm to model transport and evaporation of the fuel liquid droplets, as well as formation and transport of particulate matter, such as soot and aerosols
All these tasks are the bricks to build the simulations framework enhancing the accuracy, reliability, and moreover to have the capability to efficiently use the exascale systems.