With the combustor burning hydrogen, as well as the strongly coupled fuel and cooling system, the configuration of a hydrogen-fueled aero engine is more complex than that of a conventional aero engine. The performance, and especially the dynamic behavior of a hydrogen-fueled aero engine, need to be fully understood for engine system design and optimization. In this paper, both the transient modeling and performance analysis of hydrogen-fueled engines are presented. Firstly, the models specific to the hydrogen-fueled engine components and systems, including the hydrogen-fueled combustor, the steam injection system, a simplified model for a quick NOx emission assessment, and the heat exchangers, are developed and then integrated to a conventional engine models. The simulations with both Simulink and Speedgoat-based hardware in the loop system are carried out. Secondly, the performance analysis is performed for a typical turbofan engine configuration, CF6, and for the two hydrogen-fueled engine configurations, ENABLEH2 and HySIITE, which are currently under research and development by the European Union and Pratt & Whitney, respectively. At last, the simulation results demonstrate that the developed transient models can effectively reflect the characteristics of hydrogen burning, heat exchanging, and NOx emission for hydrogen-fueled engines. In most cases, the hydrogen-fueled engines show lower specific fuel consumption, lower turbine entry temperature, and less NOx emissions compared with conventional engines. For example, at max thrust state, the advanced hydrogen-fueled engine can reduce the parameters mentioned above by about 68.5%, 3.7%, and 12.7%, respectively (a mean value of two configurations).

The afterburner is essential for military aeroengines, providing rapid thrust increase during critical maneuvers. However, complex failure mechanisms pose challenges for fault diagnosis and troubleshooting. This study investigates a specific aviation engine’s afterburner engagement principle and typical faults, using fault tree analysis and computational fluid dynamics (CFD). We propose a hybrid diagnostic fault architecture combining traditional observer fault diagnosis with fault trees for online fault diagnosis of critical faults. Simulations and test bench verification confirm the feasibility and effectiveness of our approach, offering valuable insights and guidance for addressing afterburner engagement issues in aviation engines.

Due to the substantial disparities in physical attributes between hydrogen fuel and conventional fuels, the development of an efficient controller presents a formidable challenge. In this context, this paper delves into the utilization of a model-based design (MBD) methodology for the purpose of conceiving and enhancing control systems for hydrogen-fueled turbojet engines. The investigation commences by adopting an established physical model of a hydrogen-fueled turbojet engine and subsequently validates its performance through rigorous simulation exercises. Consequently, this research undertakes a systematic deconstruction of the design process into discrete sub-phases, thus facilitating a seamless progression from system requirement analysis to system verification. This approach engenders a concurrent design and optimization of the control system. The ultimate confirmation of the controller’s efficacy and reliability is achieved through exhaustive simulations and Hardware-In-the-Loop testing. The research findings not only serve to augment design efficiency and mitigate design expenditures, but also propose avenues for further performance ameliorations in the realm of hydrogen-fueled turbojet engines. The control system accuracy of MBD is compared with the experimental results, and under high hydrogen fuel flow conditions, the errors reach an extremely low level of 0.1%. This affords a novel design paradigm within the domain of aero-engine control.

Artificial intelligence technology has become a key technology for improving the efficiency and reliability of aircraft engine maintenance in the field of engine fault diagnosis. AI-based fault diagnosis technology uses advanced algorithms such as machine learning, deep learning, and transfer learning algorithms to analyze the large amount of data generated by aircraft engines during operation to achieve early identification and accurate prediction of potential engine faults. This technology can learn fault patterns from complex data to improve the accuracy and speed of diagnosis. By real-time monitoring of an engine’s operating status, this technology can not only remind maintenance personnel in a timely manner to intervene to prevent faults but also optimize maintenance plans and reduce unnecessary maintenance costs and downtime. In addition, with the continuous optimization of algorithms and the enhancement of data processing capabilities, AI-based fault diagnosis technology will become more precise and efficient, which is highly important for improving the operational safety and economy of aircraft engines.

Current data-driven methods for diagnosing gas path faults in aero-engines often rely on large, costly fault sample sets and face challenges related to class imbalances. These include disparities in the quantity of normal and fault data, differences among fault types, and variations in fault severity levels. This paper proposes a novel generative model, TL-GMVAE, which integrates Gaussian mixture models (GMM) with variational auto-encoders (VAE) and incorporates a transfer learning (TL) strategy. Using a large dataset of normal operational data, the VAE learns latent feature mappings of engine behavior and establishes a joint probability distribution across sensor measurements. To account for complexity in engine operating conditions, the GMM is used as the sampling distribution of the VAE. This enhances the model’s ability to represent diverse operating scenarios. The pretrained model is then fine-tuned with a small dataset of gas path fault data, transferring knowledge to fault domains. Each fault-specific TL-GMVAE model serves as an independent generator for synthetic fault samples. The proposed approach is validated using several established classifiers. The impact of different Fault-Normal ratios and imbalances across fault categories on classification accuracy is analyzed. Additionally, the robustness of the method to individual engine variability is evaluated. Results demonstrate that the TL-GMVAE generates high-quality fault samples and significantly improves fault diagnosis accuracy. These findings highlight its potential application in aero-engine health monitoring and fault diagnosis systems.

This chapter reviews digital twin-based simulation modeling methods for aero engines, which integrate physical model-driven and data-driven approaches to achieve high-fidelity simulations of aero engine behavior mechanisms. This method first establishes a performance mechanism model of the target engine and uses bench test and flight data to adaptively refine it using machine learning algorithms to obtain a digital twin model for simulation. This model can reflect the physical characteristics and dynamic behavior of the engine, thereby enabling precise monitoring and prediction of engine performance. In addition, this method combines artificial intelligence algorithms to analyze the modeling bias compensation mechanism, thereby improving the reliability of aero engine simulation and providing strong support for engine performance optimization, performance prediction, and fault diagnosis.