Coalescence of viscous blisters under an elastic sheet
W-E. Khatla, L. Duchemin, A. Eddi, E. Reyssat · 2026
Read Coalescence of viscous blisters under an elastic sheet by W-E. Khatla, L. Duchemin, A. Eddi, E. Reyssat (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-17731.txt.
Introduction
Khatla, Duchemin, Eddi, and Reyssat watch two viscous blisters merge under an elastic sheet. A synthetic schlieren setup tracks the full thickness field as the blisters join. The team zooms in on how the neck between them grows. A one-dimensional lubrication model then predicts that growth with scaling laws. Bending of the sheet drives the early motion, they find. Full nonlinear runs match the measured neck shapes well.
Catalog record
| CREATOR | W-E. Khatla, L. Duchemin, A. Eddi, E. Reyssat |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 25072 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.17731
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Coalescence of viscous blisters under an elastic sheet W-E. Khatla, L. Duchemin, A. Eddi, E. Reyssat Abstract We study the coalescence of identical viscous blisters beneath an elastic sheet both experimentally and numerically. Using a time-resolved synthetic schlieren technique, we measure the evolution of the thickness field of the merging blisters and more specifically the dynamics of the coalescence region. To explain this dynamics, we develop a one-dimensional model based on the lubrication approximation, from which we derive scalings to predict the growth of the coalescence neck. We also numerically solve the full non-linear equation to assess the theory and compare to experiments. Our model illustrates that, at short times, the dynamics of coalescence is mainly controlled by the bending of the elastic sheet, leading to a relationship between the coalescence speed and the radius of curvature of the interface at the coalescence neck. I Introduction The spreading of a viscous fluid on a solid substrate occurs at a diversity of scales in laboratory, industrial or natural phenomena. Droplets for inkjet printing technologies are typically 10 µm in radius and spread within timescales as low as a millisecond ( Lohse2022 ) . Dermo-epidermal suction blisters are submillimetric liquid pockets, form within hours, and are used as a diagnosis tool for some bullous diseases ( Willsteed1991 ) . At the geological scale, the flow of lava results in the formation of a number of characteristic elongated structures whose final shape depends on the interplay of gravity, viscous stresses and strongly time-dependent rheology resulting from cooling and crusting processes ( Griffiths2000 ) . Among them are laccoliths, shallow structures with a typical thickness of order 100 m and width spanning typically 10 km, which form in 100 to 1000 years ( Bunger2011 ; Michaut2011 ; Mathieu2008 ) . One key ingredient in the dynamics of these systems is the boundary condition at the surface of the liquid. While the spreading dynamics of droplets under the effect of surface tension has been studied extensively ( Bonn2009 ) , liquid is often covered by a skin or a crust whose solid character strongly influences the shape and dynamics of the underlying fluid. It has for instance been shown that the growth of blisters under quasi two-dimensional thin sheets gives rises to wrinkling instabilities characteristic of slender membranes under compression ( Pandey2025 ) . The displacement of a high viscosity liquid by a lesser viscous fluid in a Hele-Shaw geometry exhibits the classical Saffman and Taylor instability which is strongly affected when confinement is provided by a flexible membrane ( Pihler2012 ) . A number of recent experimental and theoretical studies describe such quasi-parallel viscous flows through model systems that couple fluid flow in the lubrication approximation to the mechanics of the covering elastic or plastic crust ( Lister2013 ; Hewitt2015 ; Peng2020 ; Saeter2023 ; Ball2023 ) . Besides spreading, coalescence is another generic and practically important event in the life of a sessile drop. It has been explored in many configurations and regimes ( Eggers2025 ) . Specifically, the growth of the height and width of the liquid bridge forming in the coalescence of wetting viscous droplets on a solid have been shown to follow scaling laws in time ( Ristenpart2006 ; Hernandez2012 ; Hack2020 ; Kaneelil2022 ) . In the present paper, we focus on the onset of coalescence of two liquid blisters covered by an elastic membrane. A recent work combines experiments, lubrication theory and numerical simulations to address the coalescence dynamics of two identical drops underneath a thin elastic sheet Saeter2024 . We propose to revisit and discuss the results of this letter through experimental results done in a different parameter range, scaling models and numerical computations. We first present the experimental setup and geometric quantities measured. We then derive a scaling theory that links the upward velocity of the coalescing neck to its local radius of curvature. Finally, we present an idealized numerical study that confirms and discusses the short-time self-similar solution predicted by ( Saeter2024 ) . II Experiments Figure 1: Sketch of the experimental setup. Sunflower oil is fed under an elastic membrane (thickness d ) through two holes (diameter \phi\simeq 1 mm) made in an underlying transparent plate, forming two identical liquid blisters of typical thickness h\sim 0.1-1 mm separated by a distance L_{gap}\sim 3 cm. A checkerboard pattern is placed under the plate and imaged from the top through the whole setup. Apparent deformations of the pattern enable one to reconstruct the liquid thickness field h(x,t) . Two identical liquid blisters are formed by injecting sunflower oil (viscosity \mu=5\times 10^{-2} Pa.s) between a planar PMMA plate and an elastic sheet (figure 1 ). The blisters are fed at constant flow rate Q=0.05-12 µL.s -1 through the PMMA substrate using two channels of diameter \phi=0.7 mm separated by a distance L_{gap}\in[1.5,3,6] cm. Prior to the injection, the substrate is cleaned using ethanol, dried, and covered with the elastic sheet. The latter is made of PDMS (Silex Superclear ) with Young’s modulus E\simeq 1.6 MPa and Poisson ratio \nu\simeq 0.5 domino2020artificial ; Ledoudic2025 . We control the bending stiffness B=Ed^{3}/12(1-\nu^{2}) by using membranes with different thicknesses d\in[0.2,0.8,1.6] mm. Injection is stopped before the blisters get into contact. The total injected volume V is then fixed and spreading goes on spontaneously beneath the elastic sheet until both liquid pockets meet and start coalescing. The values of V and Q are used to control the geometry of the membrane at the onset of coalescence. The dynamic evolution of the thickness h(x,t) of the liquid layer is extracted using a digital schlieren technique, i.e. from the apparent deformations of a checkerboard pattern placed under the PMMA substrate and imaged from above through the different layers of the setup. A Basler camera with 2048\times 2048 pixels is placed at a distance H_{c}\simeq 40 cm above the PMMA plate. The field of view is a square region of 10\times 10 cm 2 , corresponding to a horizontal spatial resolution around 50 µm/pixel. Images are acquired at 10 to 20 frames/second. The apparent displacement field of the reference pattern is fed into a Fast Checkerboard Demodulation (FCD) algorithm which results in the reconstruction of the deformed sheet height field h(x,y,t) ( Moisy2009 ; Wildeman ) . Calibration prefactors have been checked by integrating the measured height field during the constant flow rate injection phase of the experiment and ensuring that the resulting volume indeed grows linearly in time: V=Qt . The checkerboard tiles are 0.5 mm accross, which yields a vertical resolution of the method estimated at 10 µm. We therefore obtain a complete cartography of the blisters height in time. An example of this reconstructed field is shown in figure 2 a at the early times of the coalescence processes. In this particular example, V=0.5 mL, L_{gap}=3 cm, d=0.2 mm. First, we identify the peaks of the blisters and analyze a uni-dimensional cut through the (xz) plane that contains these two characteristic points. We thus measure the temporal evolution of the profile heights in-between these peaks during the total duration of the experiment. An overview of this procedure is presented in figure 2 b where we can clearly notice the edges of