Numerical Optimization of Thermo-Compressors Based on Non-Dimensional Geometrical Parameters

Abstract

Thermo-compressors are essential parts in desalination units and using the high pressure motive steam to compress large amounts of low pressure steam without any moving parts or blades. One of the current challenging issues in thermo-compressor design is the correct size selection for different dimensions to achieve the maximum entraining capability at a given compression ratio. The conventional design methods are based on varying shape parameters such as lengths, angles and diameters of internal parts and select a particular geometry which yields the best performance under given boundary conditions. The major characteristics of thermo-compressors which have to be compromised during such design procedures are “compression ratio” and “entrainment ratio”. In the current study, a new method of size selection for different geometrical parameters of thermo-compressors will be introduced. The basis of this method is to define two new non-dimensional parameters in terms of geometrical parameters. The effect of varying these parameters on both characteristics of thermo-compressors has been studied numerically via a CFD method and subsequently, a practical correlation has been developed to express the relationship between the characteristics and the geometrical parameters. The expression obtained from numerous simulations can be applied to evaluate the performance of a given geometry prior to manufacturing. The numerical simulation has been performed on more than 320 different models with different geometries according to the finite-volume steady-state method. The structured meshes were generated within the computational domain. In addition, several mesh concentrations have been studied to develop a mesh-independent model. Moreover, a modified k-epsilon method has been adopted to model the turbulent flow. Steam was taken as the working fluid and it was assumed a real gas with a nonlinear relationship between pressure and density. All governing equations have been discretized with a second-order upwind scheme and solved by implicit method. The numerical results verified with experimental measurements obtained from a real life model of thermo-compressor. The outcomes of applying the validated method on several models enable us to develop the practical characteristic curves based on non-dimensional parameters which further leading to achieve optimized shape selection for thermo-compressors.