The design of compact and efficient Diesel Oxidation Catalysts (DOC) is primarily important to comply with emission regulations not increasing engine fuel consumption at the same time.
To design DOCs, Sherwood number correlations are typically used to calculate mass transfer by varying operating conditions in terms of catalyst volume, active area and mass flow rate. To that aim, Sherwood number trend over channel length has been extensively studied during last decades. However, Sherwood number correlations are highly dependent on channel geometry, and on the possible presence of special structures (such as blades, fins or bumps). These modifications, which characterize the latest developments in substrate technology, allow to improve mass transfer performance and require a special characterization.
In this paper, a joint experimental/3D-numerical approach is used to study the role of special structures, and namely of LS (Longitudinal Structure) blades, on mass transfer mechanism inside a DOC. It is proved that the generation of unsteady/turbulent flow structures due to the inclusion of blades is essential in LS substrate, and leads already at Re in the order of 700 to an underestimation of the performance by using a steady state laminar 3D model. Re dependent Sherwood correlations are then proposed for standard and LS substrates, to calculate mass transfer between the bulk and the wall of a given channel by varying operating conditions (channel length, channel density, mass flow rate) taking also into account unsteady/turbulent effects.