At the 43rd UIT International Heat Transfer Conference in Turin, Nicolò Varallo presented the project’s first measurements of the drag coefficient of single bubbles rising in a water–ionic liquid mixture.
Preliminary assessment of the drag coefficient of single bubbles rising in a water–ionic liquid mixture
N. Varallo, D. E. Aissaoui, A. Ferrario, H. Hessenkemper, G. Besagni
43rd UIT International Heat Transfer Conference, Turin, 22–24 June 2026.
The contribution addressed the bubble-scale question on which the project’s closure relations depend: how much does an ionic liquid change the drag on a single rising bubble? The system studied was a water–[BMIM][BF₄] mixture at 10 wt% ionic liquid, compared throughout against pure water. [BMIM][BF₄] is a physical absorbent of relatively low viscosity and a well-established benchmark for CO2 studies, which makes it a natural starting point before the more viscous, chemically active solvents are approached.
The measurement
Bubbles were recorded with a high-speed camera and the images processed by background subtraction and thresholding to extract the bubble contour, and from it the aspect ratio, the position, the trajectory and the rise velocity. The drag coefficient then follows from the balance between buoyancy and drag.
A part of the work was devoted to the release mechanism itself, which turns out not to be a detail. Freely injected bubbles reached terminal velocities of 0.232–0.261 m/s against 0.164–0.224 m/s for bubbles released from a dumping cup, consistently across every diameter investigated, and the two mechanisms produced visibly different trajectories — free injection giving fewer oscillations, the dumping cup a more pronounced oscillatory motion. The initial conditions leave their mark on the whole ascent.
What came out of it
The drag coefficient increases with bubble size and is systematically higher in the water–[BMIM][BF₄] mixture than in pure water, over the entire diameter range. Bubbles in the mixture also deform more, showing lower aspect ratios and non-spherical shapes; viscosity and surface tension both play a part.
The comparison with existing drag correlations was the point of the discussion: the measured coefficients exceed the predictions across the whole range, so the correlations systematically underestimate the drag. Interfacial contamination and effects specific to the ionic liquid are not captured by relations calibrated on clean systems — which is precisely why single-bubble dynamics have to be measured before reliable closure relations and predictive models for industrial-scale bubble columns can be written.
Paper (PDF, 1.1 MB) · Slides (PDF, 1.4 MB) · The UIT 2026 page