Emission control efficiency and limited fuel consumption penalty and are the main design factors driving the development of engine-after-treatment exhaust systems according to both ACEA/DOE targets and continental regulations. The particulate-filter is certainly a critical technology to this aim as usually presents very high pm reduction efficiencies (even more than 90% on a mass basis depending on soot loading) leading however to a back pressure increase and eventually to an appreciable fuel consumption penalty. Nevertheless, it is in general discussion that health hazard related to particulate depends primarily on total number of emitted particles rather than on mass.The partial-flow-filter has been recently developed presenting lower reduction efficiencies on a mass basis but also a reduced penalty on fuel consumption. As a selective capability of this filter in capturing the smaller particles has also been experimentally observed, technological and scientific interest in the development of design and analysis tools dedicated to this technology has developed. However, traditional 1D and 2D models cannot be directly applied to study this technology, as the coupling between fluid-dynamics and filtration must be properly taken into account by exactly modeling single channel geometry as it is.Thus, in order to represent filtration effects by varying diameter class in partial-flow-filters, a 3D model has been coupled to a phenomenological filtration submodel taking into account Brownian diffusion and interception phenomena within an Eulerian transport framework for different particle classes depending on diameter. Results have been compared to experimental data and state that the model is able to represent the main occurring fluid-dynamic and filtration phenomena by varying mass flow rate and particle diameter.