Why most drops on a window never move
Between 76% and 92% of the drops on a real window never move, held by the difference between their front and back contact angles. The ones that do move go in short bursts, and big streams flow more steadily than small drops.
One number from the rain clips frames everything else: between 76% and 92% of the drops on the glass never move. Real glass is mostly still, and an effect made only of sliding drops would draw the exception as the rule.
The reason is a fairly small piece of physics. A drop on a vertical pane has a front edge and a back edge, and they do not need to meet the glass at the same angle. The front can sit at a larger angle (the advancing contact angle) than the back (the receding one). That difference, about 19 degrees plus or minus 5 on glass, is a grip that holds the drop until its weight beats it.
The volume at which a drop lets go is not one number
The retention force goes as the drop’s width times the surface tension times (cos of the receding angle minus cos of the advancing angle). Set it equal to a spherical cap’s weight and you get the volume at which it lets go. The research worked that out for three pairs of angles on 1 October: 8.7, 7.7 and 8.8 µL. Measurements on vertical glass say about 7 to 19 µL, and on a car side window drops below 17 ± 1 µL do not slide.
The files do not agree on one number. The 2 October design says the simulation’s threshold is the measured 17 µL; a note in the rain code from 9 October puts a drop’s running weight at about 9 µL. I show both here and have not chosen between them.
Cleanliness sets the angle. Perfectly clean glass is near 0 degrees, which is why it films rather than beads. A window that beads rain is the dirty case, so the research says to use 50 to 70 degrees for a window. The design lists a glass cleanliness knob on the same scale (clean 20 to 40 degrees, dirty 50 to 70), and the lab, the site’s private settings page, has a Beading control for the angle.
Small drops stick, then slip
A drop that does start does not glide. Real window drops stop at contamination and jump on. From my clips, 98 gravity slides were checked: a burst of about 0.17 seconds, a rest of 0.6 to 1.1 seconds in my handover notes (0.5 to 1.1 in the code), sliding about 16% of the time.
The same day the simulation gained stick-slip for small drops, with numbers taken from my rain clip data.
Big streams flow more steadily, and the number for that moved twice
Big water is different. The more drops merge into a stream, the faster and steadier it flows, with less start and stop; I noted that about the streams on the evening of 9 October. The stutter was fading out between the threshold volume and a multiple of it, and that multiple was 2.5 on 9 October.
The simulation’s own comment says why it moved: a head that had just let go crept in bursts for most of its run, and the GIF showed streams that hardly moved. It became 1.8 on 10 October, then 2.2 later that day, a wider stutter band.
Measured afterwards on a hero-sized pane in steady rain over 40 seconds: heads let go about 1.3 times a second, with zero to three moving at any moment, so most of them are resting between bursts.
Still an estimate: the speed at which a runner leaves a trail
Pearling, the line of small drops left by a fast runner, needs a speed the research computed at about 0.5 m/s for water, from an experiment done with silicone oil. The code uses 30 mm/s, an estimate taken from the fast heads in one of my clips.
Sources and notes (12)
- Emergent Scientist 2019, rain drops on a car window — the volume at which a drop starts to slide
- Li et al. 2023, kinetic drop friction — drop friction and the retention factor
- Le Grand, Daerr & Limat 2005, shape of running drops — how a running drop turns oval, cornered, pearled
- Quéré, drops at rest on a tilted plane — when a drop slides on a tilted plane
- Biolin Scientific: contact angle hysteresis — why drops pin: contact-angle hysteresis
- ElSherbini & Jacobi 2004, drop shapes — drop outlines and profiles
- Hodgson et al. 2021, contact angles on glass — contact angles on real glass
- Water drops research — the notes this was written from
- Streams research — the notes this was written from
- Drop simulation code — the notes this was written from
- Rain types code — the notes this was written from
- Water session handoff — the notes this was written from
Water, rain & windAll 11 parts →
