Fuel assemblies in liquid-metal cooled reactors often utilize a hexagonal array of pins with wire spacers helically wrapped around each pin. This classical rod bundle geometry is defined by six parameters: the number of rods, rod and wire diameters, pitch, wire lead and edge pitch. Since the 1960s many pressure loss tests have been performed in many countries, covering wide ranges of these parameters. Several authors also measured the flow velocities in the subchannels, i.e. the so-called flow-split. Making use of this database, the Upgraded Cheng and Todreas Detailed (UCTD) model described the friction losses in each subchannel individually and derives the bulk friction factor and flow split. UCTD predicts most friction data within 10% and reduces smoothly to bare bundle results if the wire diameter is set to zero. In this work, some features of the UCTD model are revised in order to address newly identified user needs relevant for uses in the context of licensing. In particular, UCTD consistently over-predicts the available data for flow velocity in the edge sub-channels. This is addressed by incorporating an additional physical mechanism of momentum exchange due to flow mixing. The new model, here called Pacio-Chen-Todreas detailed model (PCTD), smoothly reduces to UCTD if the mixing coefficients are set to zero. A best fitting set of empirical parameters is obtained by calibration using an extensive experimental database. PCTD can predict the bulk friction factor slightly better than UCTD, even with fewer non-zero empirical parameters. Moreover, PCTD provides an accurate prediction of the edge flow split. This feature is important for calculating scenarios, which cannot be compared directly with experiments, as in the context of licensing.
|Title of host publication
|NURETH-19 - 19th International Topical Meeting on Nuclear Reactor Thermal Hydraulics
|Subtitle of host publication
|Brussels, Belgium, March 6 – 11, 2022
|American Nuclear Society
|Number of pages
|Published - 6 Mar 2022