TRANSIENT MODELING OF HEAT EXCHANGERS

Three approaches were used to simulate the transient condensing, or evaporating, refrigerant flow inside a single tube heat exchanger. They are the lumped parameter analysis, the unsteady one-dimensional flow analysis and the quasi-steady one-dimensional analysis. The tank model is a single lump analysis in which the refrigerant properties inside the heat exchanger are uniform. This model was solved by a fourth-order Runge-Kutta method, and predicted qualitatively the transient condensing or evaporating flow. In the unsteady one-dimensional flow analysis, the flow is single phase in part of the heat exchanger, and two-phase in the rest. In the two phase region, a homogeneous flow is assumed. The details of a transient evaporating flow was solved by four finite difference methods: the method of characteristics, Lax\u27s method, MacCormack\u27s method, and a semi-implicit method. Due to the stability criterion, the time step for the one-dimensional flow model is very small. A large amount of computer computational time is required for each transient condition. Unless higher speed computers than are now available are developed, the unsteady one-dimensional model cannot be recommended. The quasi-steady one-dimensional analysis assumes that the time constant of the flow inside the tube is much smaller than the time change of the inlet and/or outlet conditions. This allows the problem to be solved as a series of steady state flows. Each steady state flow was solved by a fourth-order Runge-Kutta method. The time step can be varied, depending on how rapidly the inlet and/or outlet conditions are varying. This results in a large saving of computer computational time. However, initial conditions of the heat exchanger and the thermal capacity of the tube wall are not considered in this approach. Both conditions can be corrected. It is concluded that, with the speed of the present computers, the lumped parameter analysis is the best choice in the simulation of the transient response of heat exchangers. This method has a drawback in that it lacks any spatial information. The quasi-steady model has the potential of allowing some transient spatial information and should be investigated further