COMPARISON OF THE 10 RESOLUTION METHODS F OR ENERGETIC FLUIDS NETWORKS

The ten methods compared are: the recurrent method (successive applica tions of groups of lengths in parallel or serial assemblies); the node s-pressure-method (the pressure and/or the fluid exchanges are fixed f or each node); the eight circuits-methods (based upon the combination of the two Kirchhoff laws, two types of linearization and two methods of resolution). The comparison criterions selected for this study are, for each of the above-mentionned methods: the type(s) of network(s), among the nine possible, to which it can be applied;the constraints it induces in defining the components of the network; the conditions tha t it requires in initializing and memorizing; its computer compatiblit y : memory capacity and time of treatment; the physical fields to whic h it can be applied : hydraulics, thermics and energetics; the calcula tion purposes it can fulfil : initial dimensioning, study of variable speed and continuous processes. One of the major results of this compa rison is that none of the methods analysed is optimum for all the crit erions. However, studies have showed that it is possible to elaborate a single very simple calculation method for the eight circuits-methods . It can be done by using simultaneously codification principles and b oolean variables as exponents of physical properties. This will be of great utility in ulterior applications as the software will at any mom ent have the possibility to choose the method best adapted to the calc ulations.