Hydrogen for zero carbon emission in heavy duty fleet

NAISS 2023/5-304


NAISS Medium Compute

Principal Investigator:

Hesameddin Fatehi


Lunds universitet

Start Date:


End Date:


Primary Classification:

20306: Fluid Mechanics and Acoustics




In heavy-duty commercial vehicles for long-haul application, both from a CO2 equivalent and total cost point of view, hydrogen powertrains represent a rapid measure for achieving CO 2 -free mobility. Hydrogen is considered to be the energy vector of the future and the EU has plans to increase the share of hydrogen in its energy mix. Given the high cost of hydrogen, fuel economy and high efficiency are important aspects in developing hydrogen internal combustion engines (H2 ICE). The focus of this project is on a new hydrogen concept based on direct injection compression ignition (DICI) engine with the potential of delivering high power output with high efficiency. While this concept can significantly increase efficiency, the ignition of hydrogen is a major challenge that needs to be addressed. To understand the fundamental physical aspect of this concept and to address its challenges, we aim to develop Computational Fluid Dynamics (CFD) tools capable of predicting hydrogen combustion in an ICE environment, addressing the special physical and chemical properties of hydrogen including differential diffusion, diffusive-thermal instability, pre-ignition, and multi-mode combustion. The CFD tools will be used to develop and verify zero carbon emission hydrogen engine concepts with high efficiency, a pure hydrogen DICI concept, by replacing the diesel pilot with a hydrogen jet flame. Computational Fluid Dynamics (CFD) simulations are considered as an effective tool in developing new engine concepts and understanding the interplay between different physical and chemical processes inside an internal combustion engine. Two approaches are commonly used to model turbulent flow; Reynolds-averaged Navier-Stokes (RANS) and Large-Eddy Simulation (LES). In the RANS approach, all the turbulence scales are modeled and only the mean cycle (ensemble) averaged flow is resolved. LES, on the other hand, resolves large scale dynamics of flow structures and models flow structures smaller than spatial filter size. LES has much higher accuracy and fewer limitations in representing complexities in the flow field, with higher computational cost. HPC resources are required for conducting LES simulations. Even RANS simulations in the presence of chemical reactions and multi-physics, are dependent on HPC resources. In this project, following studies are planned: - LES model of hydrogen injection and charge preparation, - Modeling of ignition in stratified hydrogen/air mixture, - Modeling of hydrogen combustion and NOx emission, with the goal of Identifying the relation between charge preparation, reaction front propagation and NOx formation. - Engine performance simulation and verification of zero carbon emission engine concept Computational domains are simple geometries for hydrogen injection studies and engine geometries for combustion studies.