The Use of Convolution Integrals in Determining the Dynamic Response of Heat Exchangers Having Random Inputs

Partial differential equations to describe the behaviour of a heat exchanger, in which both flow rates and temperatures very simultaneously, are derived. The equations are then expressed in finite difference form using a Crank-Nicolson implicit method. A numerical model based on these equations is developed and validated both theoretically and by experimental measurements.
A technique for analysing a linear system to obtain the impulse response functions is evolved, and its suitability is demonstrated using a simulated system. This technique, is then applied to 16384 sets of outlet temperatures calculated by the numerical model in order to obtain the impulse response functions of heat exchangers.
A model of a heat exchanger based on convoluting the impulse response functions with the inputs is developed and the accuracy of the model is studied for ten different tests.
Finally a method of scaling the impulse response functions for different inlet conditions is given and the overall model validated against experimental measurements.