Abstract The transition of large-bore internal combustion engines to hydrogen offers a pathway for decarbonizing stationary power generation. However, achieving competitive power densities with port fuel injection requires high charging pressures due to intake air displacement and lean-burn combustion. The combination of high charging requirements and low exhaust gas enthalpies constrains conventional single-stage turbocharging, necessitating alternative concepts such as multi-stage or assisted systems. Because the charging architecture directly influences in-cylinder thermodynamics, this study experimentally quantifies the impacts of exhaust backpressure on a single-cylinder large-bore hydrogen engine by comparing an unrestricted exhaust baseline with configurations emulating the backpressure of single-stage and assisted turbocharging concepts. Results indicate that applying a realistic single-stage turbocharger efficiency decreases the indicated engine efficiency by 3.7 percentage points compared to an unrestricted baseline. Emulating an electrically or mechanically assisted turbocharger setup restricted the stable operating map, reduced the maximum indicated mean effective pressure from 19.7 to 18.7 bar, and caused an indicated efficiency penalty of up to 3.1 percentage points. This highlights the fundamental trade-off: achieving the necessary boost pressure either induces a thermodynamic penalty via high exhaust backpressure or requires additional power to drive the compressor. Regarding emissions, the elevated backpressure substantially reduces nitrogen oxides at medium loads with less lean mixtures compared to the unrestricted baseline due to the increased trapping of residual gas, while causing a slight increase at low loads. At high engine loads, however, both nitrogen oxides and unburned hydrogen emissions remain unaffected by the increased backpressure. These findings demonstrate the necessity of advanced charging architectures, such as multi-stage or externally assisted systems, to optimize the efficiency of stationary hydrogen engines.
Kinematics of independent wheel suspensions are often described by the concept of roll centers. In this article, this concept is critically examined. To overcome the typical simplifications, an enhanced model for a single axle is presented which is able to describe the roll and heave motion without the need of defining a roll center. The focus is on the so-called jacking forces generated by the suspension geometry and constraints. These jacking forces are modeled using the principle of virtual work and enable the description of the geometric load transfer without assuming an arbitrary roll center. Although this transient model is simple compared to complete vehicle multi-body models, effects such as aerodynamic downforce or effects due to camber angle, specifically camber thrust, on the individual suspension system can be considered accordingly. Consequently, characteristics such as ride height changes can easily be determined. Additionally, unequal lateral tire forces on the left and right side of an axle during cornering, and its resulting jacking forces can be taken into account. The proposed model is compared to an existing model, which considers the kinematics of independent suspension systems, and to the classic roll angle calculation, both found in literature. Finally, a validation against a multi-body simulation shows that the proposed model is able to predict roll and heave motions of independent wheel suspension systems in an efficient and physically plausible manner.
Detailed stack models and dynamic balance-of-plant (BoP) models each capture important aspects of proton exchange membrane (PEM) fuel cell systems but are limited when applied independently. While detailed stack models are computationally demanding for system-level investigations, reduced-order system models often lack the physical fidelity required to accurately represent electrochemical processes and their interaction with auxiliary components. To address this gap, this work presents a modular cross-tool co-simulation framework that couples a physics-based PEM fuel cell stack model with a dynamic BoP model. In contrast to existing co-simulation approaches focusing primarily on stack-internal behavior, the framework enables a consistent analysis of the interactions between stack operation and BoP dynamics, linking electrochemical, thermal, and auxiliary losses within a unified system-level environment. The framework is implemented in AVL CRUISE™ M and MATLAB® and evaluated under representative transient operating conditions. Solver synchronization and coupling strategies are assessed with respect to numerical stability, simulation accuracy and computational performance. The results demonstrate stable and accurate co-simulation while achieving a real-time factor of 1.74, making the framework suitable for real-time applications. Furthermore, the integrated loss analysis reveals that the overall system efficiency decreases from 49.6
Abstract Hydrogen closed-cycle engines combine hydrogen combustion with an argon–oxygen working-gas loop, offering a promising path toward high-efficiency, carbon-free, and nitrogen-oxide-free power generation. Previous demonstrations of this Argon Power Cycle (APC) concept have shown high indicated efficiencies, but low power density remained a central barrier to practical application. This paper presents a simulation-driven preliminary study on power density enhancement of a hydrogen closed-cycle CI engine. The study deliberately covers a very wide parameter range; its purpose is concept and process development, so the reported results should be read as trends and potentials. The long-term goal of the underlying project is a large-bore two-stroke closed-cycle engine; the single-cylinder research engine considered here serves as a model application to study the governing effects that must be understood for the design of the large engine. The investigation includes one-dimensional system simulation, four-stroke and two-stroke parameter studies, and three-dimensional CFD combustion modeling. Test-bench measurements provide the experimental basis for boundary and validation data. A 54-case parameter study shows that the volumetric power density ratio E v can be approximately doubled through hydrogen enrichment, compression-ratio adaptation, and elevated system pressure, constrained by mechanical peak-pressure limits; two-stroke conversion adds 60–80% further potential. The validation work identifies the caloric properties of the trapped working gas as a critical parameter: the species composition of the argon–oxygen loop directly determines the effective heat capacity ratio and therefore governs compression temperature, ignition delay, and pressure rise. The 1D system model quantifies the power-density potential and establishes thermodynamic boundary conditions for the subsequent 3D CFD analysis, which is applied to a selected operating point after reduction of the compression ratio to resolve the ignition process and heat release under well-defined boundary conditions. The CFD model, based on detailed gas-phase reactions and a combined H 2 /O 2 –diesel mechanism, demonstrates that combustion can be simulated successfully. An accompanying economic assessment indicates that efficient argon and oxygen recycling is an enabling condition for the concept. The results provide a validated methodological foundation for future scaling to larger-bore engines and further process development.
Abstract Urban traffic areas, such as intersections, significantly escalate the risk of accidents due to the high density and diverse behaviour of road users. Achieving higher automation levels necessitates robust object recognition and tracking capabilities to mitigate accident risk. Modern vehicle architectures utilize multiple sensor systems to ensure robustness and high-quality perception. For camera systems in particular, multi-camera multi-target tracking (MCMTT) with visual re-identification (ReID) is fundamental for understanding object behaviour across different fields of view. This work directly addresses this need by presenting and evaluating a deep learning-based concept for near-field MCMTT using Waymo’s real-world perception dataset [1]. This work enables comprehensive and reliable tracking, focusing on the investigation of concept parameters and highlighting the essential factors for the automated driving task.
