Momentum-averaged velocity and full inertia terms resolve splash overestimation in squeezed viscous films
The preprint supplies the first quantitative theory that reconciles predicted and measured splash velocities for viscous squeeze films. By retaining convective inertia and adopting a momentum-averaged ejection speed, agreement improves across wide parameter ranges. Remaining gaps concern surface compliance and non-Newtonian effects that must be addressed for practical use.
The study targets the long-standing mismatch between squeeze-film theory and measured peripheral ejection speeds. Prior models treated the liquid as inviscid or neglected radial convection, producing velocities several times higher than observed. Chandra derives an integrated momentum balance across the film thickness that accounts for the radial velocity profile evolution during the final stages of approach. Experiments spanning viscosities from 0.1 to 10 Pa s and disk radii 10-50 mm confirm the corrected predictions fall within 15 percent of measured values.
Everyday squeeze events such as dispensing toothpaste or operating MEMS lubrication layers obey the same hydrodynamics. The corrected framework shows that splash onset is governed by a critical Reynolds number modified by the instantaneous gap height rather than by a simple inertial-capillary balance. This insight explains why highly viscous inks in screen printing rarely splash despite rapid plate closure.
The main limitation remains the assumption of perfectly parallel, rigid disks; real surfaces deform and misalign at high speeds. Validation against compliant substrates and non-Newtonian fluids is required before the model can guide industrial coating or droplet-ejection processes. Future work should test the predicted velocity scaling at gaps below 10 micrometers where van der Waals forces begin to compete.
Navin Kumar Chandra: Parallel-disk experiments at 100 micrometer initial gap will match the new model's splash velocity within 12 percent for 5 Pa s silicone oil by end of 2027.
Sources (2)
- [1]Primary Source(https://arxiv.org/abs/2608.07960)
- [2]Supporting Source(https://doi.org/10.1017/jfm.2018.1023)