A novel transfer matrix-based method for steady-state modeling and analysis of thermal systems

• Published in: Energy (Oxford) 2023-10, Vol.281, p.128280, Article 128280
• Main Authors: Zhao, Tian, Sun, Qing-Han, Li, Xia, Xin, Yong-Lin, Chen, Qun
• Format: Article
• Language: English
• Subjects: Heat current method; Heat exchanger network; Thermal resistance; Thermodynamic system; Transfer matrix
• ISSN: 0360-5442
• DOI: 10.1016/j.energy.2023.128280

Performance analysis and optimization of thermal systems are unprecedentedly crucial in pursuing higher energy efficiency. However, the precise modeling and high-efficiency calculation of modern, complex thermal systems are still challenging. Here we present a transfer matrix-based system modeling method to tackle these challenges. The transfer matrix of heat exchangers is defined to describe the temperature mapping relation between inlet and outlet nodes. Heat-work conversion devices are described by the working fluid temperature variation. The system modeling follows a three-step procedure: (1) building system transfer matrix by aggregating component models; (2) separate given and unknown node temperatures by removing redundant rows and columns in the system transfer matrix; (3) rearrange the reduced system transfer matrix to calculate internal and outlet node temperatures. The constructed system model does not introduce unnecessary simplifications and describes the system’s nonlinearity accurately. The proposed modeling method is first presented and validated on a parallel-series and a multi-loop heat exchanger network. Numerical results with both constant and varying fluid properties confirm the method’s validity. Case studies on three thermodynamic systems further validate the method’s capability and efficiency. The proposed method is finally compared with conventional modeling methods and heat current method to highlight its features and advantages. •Propose a transfer matrix-based modeling method for thermal systems.•Model construction follows a three-step procedure and can be automated.•System model describes nonlinearity and inside the system accurately.•Both first and second law analyses can be performed from the system model obtained.•Two HEN and two power cycles are used to validate the proposed method numerically.