Encyclopaedia Index

TITLE : SFT analysis of Heat Exchangers


    BY               : D.B. SPALDING and S.V.ZHUBRIN
    FOR              : Technical-discussion meeting
    DATE             : March, 1998
    PHOENICS Version : 3.1

PROBLEM :

Despite their great industrial importance, heat exchangers are still designed and analysed by techniques which are inferior to those of Computational Fluid Dynamics in several respects, including:-

SOLUTION :

The presentation will illustrate how it is possible to use a single computer program to calculate from the partial-differential equations governing relevant SFT ( Solid, Fluid and Thermal) processes the distributions of:-

GENERAL IDEA :

The method adopted is a three-domain one. This implies that the computational grid covers the shell-and-tube space three times: once for shell-side simulation; once for the tube-side simulation; and once for metal simulation.

In the metal, it is displacements and dilatations which are computed rather than velocities and pressures.

All governing equations, formulated differently for each domain, are then assembled and solved simultaneously in one computational space.

The method is aimed to develop HEXAGON, a PHOENICS based, Heat EXchAnGer simulatiON, package .

IMPLEMENTATION :

All multi-domain settings and inter-domain links have been made by PLANT.

RESULTS :

What follows are the series of PHOTON plots illustrating the calculation results.

The fluid movements, temperature distributions and stresses in materials will be shown for a model heat exchanger.

The following picture depicts its lay-out.

The velocity vectors show the movement of tube fluid in the header, in the U-tube array and zigzag flow in baffled shell side.

Temperature distribution reflect the conjugate nature of the transfer processes which results in the followings:

CONCLUSION :

The study has provided the first practical example of the Solid-Fluid-Thermal-in-single-operation analysis put forward recently by the first author.

HEXAGON multi-domain technique has been demonstrated as being practicable and simple.

It is readily extensible to 3D, unsteady, two-phase etc.

The technique is capable of easy implementation of recent advances in Multi-Fluid Model which have been recently shown by the second author to be important for predicting thermal performance of heat exchangers.

One can consider the HEXAGON idea as opening the benefits of SFT computer simulation to all thermal engineers who need the complete solution of their problems by single computer code.