An Evaluation of Atomization Models for Dense Sprays N.B.H. Abdelkarim 1 , A.R. Masri 2 , S.S. Ibrahim 1 and G. Wigley 1 1 Department of Aeronautical and Automotive Engineering, Loughborough University, LE11 3TU, UK 2 School of Aerospace, Mechanical and Mechatronic Engineering, The University of Sydney, NSW 2006, Australia Abstract Calculations of a transient atomization process are presented, which simulates fuel injection of sprays in gasoline direct injection engines. Only non-reacting sprays are considered with the focus on the atomization process. The FIRE code, developed by AVL, is used as the platform to test three different atomization models: (i) Taylor Analogy Breakup (TAB) model; (ii) surface wave instability (WAVE) model; and the more recent (iii) FIPA (Fractionnement Induit Par Acceleration) model. Comparisons of calculations with experimental data reveal significant discrepancies regardless of the atomization model used. It is acknowledged that, in this study, only the standard model constants are adopted and that may be further optimised to improve the calculations. However, the fact remains that all the atomization models start with an initial distribution of spherical droplets at the injector tip. An assumption that is not supported by recent measurements which show that fluid elements rather than spherical droplets dominate this early zone. 1. Introduction Modelling of fuel spray combustion is a highly complex process, which involves the phenomena of the disintegration of fuel sheets and ligaments into droplets which are poorly understood. Breakup and atomization in the formation of dense sprays is a typical example of such processes where semi-empirical models are used and where there is a severe shortage of relevant and reliable data. This near nozzle region of the spray is obviously critical as it determines the nature of the droplet evaporation, coalescence and combustion processes, which take place further downstream in the fully developed spray. There are a number of excellent reviews and monographs on sprays, [1-2] which highlight the lack of understanding, the crude modelling of the atomization process and the need for improvements. There are a number of models for spray breakup and atomization described in the literature and some of these are implemented in commercial codes used by the spray community, particularly in the automotive industry [3-5]. There are also significant efforts made by various groups to generate new data for dense spray atomization and to enhance the current models and validate them against these data [6-8]. Most of the models used in commercial codes make the assumption that atomization involves the break up of large spherical droplets into smaller ones [9]. None of these models have yet been adopted