@misc{9233020,
  abstract     = {{High-pressure homogenisers (HPHs) are used for a variety of food products to create emul-sions. It is interesting both from a sustainable and economical perspective to make HPHs more energy efficient due to their high energy demand and low efficiency. Previous research has shown that drop deformation is affected by the whole HPH valve geometry, including the inlet chamber, and it is hypothesised that droplets can be primed for breakup. 

This project aims to evaluate how well two computational fluid dynamics (CFD) methods predict viscous drop deformation in the inlet chamber of a lab-scale HPH valve model. The first method involves using mathematical models based on velocity gradient data from 1D discrete phase method (DPM) CFD simulations. The second method includes a 2D Volume of Fluid (VoF) CFD simulation where the spatial resolution of the droplets is considered. Both methods were compared to experimental observations with respect to degree of defor-mation, location of deformation and to the dimensionless Capillary number. 

Drop release events were recorded with a high-speed camera on an experimental HPH valve model with a gap height of 0.75 mm for three different volumetric flow rates. The record-ings show that the droplets start deforming at two gap lengths before the gap inlet and reach their maximum deformation within one gap length after the gap inlet. Lower volumetric flow rates give larger droplets and more deformation. A trend of increasing deformation with increasing Capillary number is shown for all three data series. However, each data set is separated from each other in the degree of formation at similar Capillary numbers. This suggests that the Capillary number does not capture the drop deformation on its own in this HPH valve setup. 

For the 1D DPM method, two out of the three mathematical models predict the deformation well with regards to the degree of deformation. However, the position of the deformation maxima differed slightly from the experimental maxima for both models. This can be ex-plained by the 1D method considering the droplets as point objects, in contrast to the 2D VoF method which account for spatial resolution and predicts the location of the defor-mation maxima well. However, the 2D VoF simulations require an effective interfacial ten-sion 16 times higher than the real interfacial tension to yield similar degrees of deformation as the empirical data. Similar to the empirical data, the 2D VoF simulations show increasing degree of deformation with 
increasing Capillary number. However, the 2D VoF method does not reproduce a clear sepa-ration in deformation magnitude between volumetric flow rates as experimentally observed. 

These results show that both CFD methods predict drop deformation well. The 1D DPM method is advantageous since it requires significantly less computational power.}},
  author       = {{Wernersson, Linus and Chalfork, Hannes}},
  language     = {{eng}},
  note         = {{Student Paper}},
  title        = {{Predicting drop deformation - CFD-based modelling of high-pressure homogeniser valve inlet chamber}},
  year         = {{2026}},
}