The aim of the dynamic model is to predict the transient pressure
on the shell-side due to a partial or full-bore tube rupture. This
requires the solution of the primary ordinary differential equation
(ODE), Eq. (8), along with the hyperbolic partial differential equations
related to the spatial and temporal variation of pressures and
flows in the upstream and downstream connecting piping systems,
Equations 9e11. Equation (8) is solved with a 4th order Runge-
Kutta method, although a simple explicit finite difference scheme
could be adequate, while Equations 9e11 are solved by the Method
of Characteristics described in Botros and van Hardveld (2012).
The above set of equations including the primary ODE (Eq. (8)),
the algebraic compatibility equations resulting from the method of
characteristics, and other auxiliary equations resulting from the
release rate and adiabatic flash of the effluent mixture into the
shell-side of the exchanger, are solved simultaneously in a
marching-in-time numerical scheme from an initial condition just
prior to tube rupture. The time step for the solution of these
equations is selected to comply with that required by the method of
characteristics for the solution of the transient flow equations according
to Courant et al. (1926) stability criterion.