Optimization of digital twin-driven engine ventilation system

Aviation heavy oil piston engines have broad application prospects in general aviation vehicles and drones at small and medium power levels. As a new piston engine, the overhead valve two-stroke aviation heavy oil piston engine has the advantages of low lubricant consumption and sharing manufacturing processes with four-stroke engines. In recent years, it has become a key research and development target for domestic and foreign aero engine teams. However, the new overhead valve structure has brought many new challenges to the design, manufacturing and performance optimization of aviation heavy oil piston engine ventilation systems. On the one hand, the transmission mechanism of the overhead valve ventilation structure has become very complex due to the introduction of multiple new components (such as valves, rocker arms, etc.). The design, manufacturing and assembly of these new components need to meet the requirements of lightweight and safer aviation engines at the same time. The transmission and sealing performance after assembly of the system structure is often unsatisfactory. Relevant manufacturing and process processes urgently need reasonable planning, and real-time data recording and feedback, and iterative optimization of manufacturing process solutions to realize intelligent manufacturing as much as possible. On the other hand, the overhead valve ventilation structure has introduced a number of uncertain system parameters, which increases the degree of freedom of valve timing parameters, thereby increasing the difficulty of multi-objective performance optimization. At present, foreign countries mostly rely on repeated trials and relies on repeated trials for the system parameters, which can consume a lot of time and cost, or establish a virtual engine performance simulation model. However, due to the lack of feedback and correction of real-time test data, the accuracy of the performance model needs to be improved. Therefore, it is very challenging to explore fast and effective optimization methods for the design, manufacturing and performance of overhead valve two-stroke aviation heavy oil piston engine ventilation systems.

In industryIn the context of 4.0 and intelligent manufacturing, digital twin technology establishes multi-dimensional dynamic virtual models of physical entities in a digital way to simulate attributes, behaviors, and performance in real environments. The real-time iterative optimization attributes of the digital twin model provide new technical approaches for the development of air circulation systems for heavy oil piston engines. Based on the deep fusion characteristics of information physics and iterative optimization coordination mechanisms of the real and virtual worlds, rapid and effective optimization of engine ventilation system design, manufacturing and performance will be achieved.

picture1 Digital twin module optimized for air circulation system of heavy oil piston engine

This paper proposes a manufacturing and performance optimization method for overhead valve aviation heavy oil piston engine ventilation system driven by digital twin. Based on the digital twin five-dimensional model, six twin modules of the structure, materials, processes, behavior, performance and environment of physical entities are established to cover the multi-faceted and full process of system optimization, as shown in the figure1 shows. Based on these six twin modules, the specific optimization method is shown in Figure 2. In the virtual space, a virtual engine model containing behavioral and environmental twin modules is established. Further, virtual performance optimization based on DoE iterative optimization and mechanism dynamics verification is carried out smoothly, thereby providing better performance parameters for virtual manufacturing simulation. In virtual manufacturing simulation, virtual geometric modeling, virtual manufacturing processes and virtual assembly processes are implemented one by one. In real physical space, the system manufacturing assembly and performance testing process are carried out under the guidance of the corresponding twin modules of the virtual space. The data related to real manufacturing measurement and performance testing are recorded and analyzed in real time in the digital twin service system and fed back to the corresponding virtual twin modules in real time. Therefore, the twin module that receives real-time data can continuously perform real-time self-update and optimize. Under the guidance of this method, the simulation results based on the twin module are very small and the actual manufacturing assembly and performance errors. The virtual model can perform multi-objective multi-directional optimization instead of the real process, thereby significantly improving the optimization level of the overhead valve two-stroke aviation heavy oil piston engine ventilation system, reducing the number of physical trial production and trials and repeated trials, reducing R&D costs, and shortening the development cycle. The methods proposed in this article are universal and can be promoted and applied in other complex system optimization, which is of great significance to promoting the development of technologies such as intelligent manufacturing, digital twins, and system optimization.

picture2. Optimization method for air ventilation system driven by digital twin