Research On Distribution Parameter Modeling Of Thermo-fluid System Using Object-oriented Method

Fluid flow and heat transfer phenomena are found everywhere in nature. In industry, thermo-fluid systems are widely used and play an very important role. Numerical simulation complements theoretical analysis and experimental method with its advantages, and they together form the basis of research methods for the thermo-fluid discipline. System level simulation technology has become a supporting mean for the design of complex thermo-fluid system. However, the existing modeling and simulation software plats for thermo-fluid system can’t satisfy increasing requirements of reusability, extendibility and flexibility. And these plats also don’t support multi-domain and physical modeling fully. Modelica is capable of multi-domain modeling, non-causal declarative modeling and continuous-discrete hybrid modeling. These characteristics make Modelica becoming the standard of multi-domain modeling language. It is a general trend to apply Modelica to thermo-fluid system simulation. So far, the latest version of Modelica cannot directly represent models containing partial differential equations. And there is no mature two-phase flow model in Modelica. The two aspects strict the application of Modelica to thermo-fluid system heavily. Aiming at solving problems above, we carry out research on distribution parameter modeling of thermo-fluid system using object-oriented method. The key issues are studies in the paper.Firstly, the general control equations of both one-phase and two-phase one dimension fluid flow and heat transfer are mainly derived from the basic mathematical models and physical theory of thermo-fluid system. In consideration of modeling and numeric characteristics using Modelica, the control equations are transformed specially. Thus, The control equations are obtained, which are fitting for establishing models using Modelica.Secondly, according to the requirements of distribution parameter modeling, PDEs are discretized and transformed into differential algebraic equations using finite volume method, which Modelica can represent directly. The general method of PDE discretization for Modelica is proposed. This makes Modelica capable of representing PDEs to some degree. The results comparing our models with Fluent models show the proposed discretizing and transforming method is valid.Thirdly, the theory and characteristic of modeling and simulation using Modelica are analyzed in detail. In order to meet the requirements of modeling thermo-fluid system, object-oriented method, system decomposing principle, interface designing mechanism, model complexity level, non-causal equation representation and model hierarchical structure are investigated. The means of using inheritance, aggregation, replacement, redeclaration and class parameter in model are also explored. The distribution parameter modeling method of thermo-fluid system based on Modelica is generalized to guide the process of model design and establishment.Next, with the guidance of distribution parameter modeling method, a set of two-fluid model architecture and its base models, which have high level of reusability, extendibility and flexibility, are designed and built based on Modelica. On the basis of architecture and base models, a preliminary two-fluid model is implemented. A kind of evaporator is modeled and simulated, common knowledge of two-phase flow in evaporator are obtained.Finally, a type of gas-pressurized liquid propellant engine system is modeled. Simulation and analysis are performed to study dynamic transients and obtain useful design experience of liquid propellant engine. Performance of alternative design schemes can be predicted and verified through simulations. The application of distribution parameter modeling in real thermo-fluid system design process is presented. The results show that achievements in this paper have practical application values, and provide an efficient mean for design and simulation of liquid propellant engine system.