Research on Key Issues in Multi-Scenario Inverse Reconstruction of Heat and Mass Transfer for AI-Driven Advanced Thermal Protection Materials

Submission Deadline: 15 May 2027 View: 25 Submit to Special Issue

Guest Editor(s)

Assist. Prof. Shuyuan Zhao

Email: angel.zsy@126.com

Affiliation: School of Astronautics, Harbin Institute of Technology, Harbin, China

Homepage:

Research Interests: theoretically modeling and simulation of thermal management systems of spacecrafts, synthesis and military application of thermal protection materials, inverse reconstruction of radiation properties for thermal protection materials, multidisciplinary analysis and optimization of aircraft thermo-structural systems, damage analysis and design of composite key components for space application, sustainable resources and energy technology and management

Summary

The technology of inverse reconstruction of heat and mass transfer within thermal protection materials is extensively utilized in the optimal design of thermal management for engine hot-end components and spacecraft thermal protection systems (TPS). It also enhances energy conversion efficiency and system safety in solar thermal power generation system receivers and nuclear reactor cladding materials, as well as facilitates microstructural design, reconstruction, and new material development in advanced functional materials. However, solving inverse problems related to heat and mass transfer in these materials presents three core challenges: a high-dimensional parameter space, strongly nonlinear mapping relationships, and ill-posedness. Traditional iterative optimization algorithms such as gradient descent and genetic algorithms often necessitate substantial computational resources and are susceptible to becoming trapped in local optima or being sensitive to initial values when addressing these issues. The deep integration of artificial intelligence into the inverse reconstruction of heat transfer in thermal protection materials signifies not only an inevitable trend in technological advancement but also serves as a critical catalyst for transforming the discipline from "experience-driven" to "data-and-physics co-driven" paradigms. Nevertheless, the evolution of AI-driven inverse reconstruction methods for heat transfer in thermal protection materials encounters numerous urgent scientific and technical issues that require resolution. These include the acquisition of high-quality experimental data spanning broad temperature ranges and multiple operating conditions, or high-fidelity simulation data, which remains costly and lacks universality. The effective integration of physical prior knowledge, such as radiative transfer equations and material optical laws, into AI models to enhance interpretability and reliability poses a significant challenge. The accurate consideration of multi-scale, multi-physics coupling effects during inverse reconstruction remains a current research hurdle. Addressing these challenges necessitates deep interdisciplinary collaboration among heat transfer, materials science, and computer science. Artificial intelligence solutions for inverse heat and mass transfer problems in thermal protection materials exhibit diverse application scenarios and hold broad prospects across multiple fields.

 
This special issue will focus on, but is not limited to, the following themes: 

1) Advancing Physics-Informed Machine Learning (PIML) by embedding stricter physical laws into AI models to enhance generalization capability and interpretability; 

2) Exploring multi-modal data fusion (e.g., spectral, thermal imaging, structural characterization data) approaches for inverse reconstruction of thermal protection materials to improve solution accuracy; 

3) Developing real-time online solving algorithms for low-latency, high-reliability AI solvers addressing inverse problems in flight vehicles' real-time thermal protection requirements during orbital operations or flight; 

4) Integrating inverse reconstruction with materials genome engineering to enable intelligent design and rapid iteration of thermal protection materials; 

5) Incorporating Digital Twin technology to construct digital twins of thermal protection systems, achieving dynamic interaction and optimization between physical entities and virtual models through inverse problem solutions.

Keywords