Models constraining models: Analogue modelling to assess theoretical uncertainties

Our research group is focused on understanding the movement of magma in the upper crust. Such movement is commonly in the form of a dyke, a sheet of magma that migrates through the crust by continually cracking the crust at the upper end.

A key force allowing dykes to propagate is the buoyancy of the magma, in which relatively-less-dense magma pushes up through the denser crust. More specifically, we are examining how a dike slows down and stops propagating (arrests) and how a dyke accelerates in the vicinity of the surface. This helps us to understand the conditions that can lead to the stalling of a dyke or the occurrence of a volcanic eruption.

Simple “quasi-static” experiments are to be performed, in which a dyke is subjected to a static load and allowed it to evolve into a new static state. Previous experiments have used gravity as the driving force. However, as gravity cannot be turned on and off (thankfully!), we will instead apply a compressive force to one end of the dyke, compressing it and causing it to propagate.

Experimental Set up. You can see the dyke propagating through the gelatin, and the resultant stress shadow emanating from the dyke tip. (Source: Stephen Pansino)


To remove gravity as a factor in propagation, neutrally buoyant liquids will be used to make the dyke. This can be done by introducing a density gradient to the host gelatin and allowing the dyke to propagate laterally at its preferred depth. In order to perform a lateral propagation, a new experimental apparatus will be set up allowing continual examination the conditions for dyke arrest and acceleration while simultaneously benchmarking numerical codes.A block of gelatin with many small dikes filled with air. The gelatin was prepared with a layer of sugar solution, filled with small bubbles at the bottom of the tank. Bacteria began to consume the sugar solution, and their waste gases caused the bubbles to grow. (Source: Stephen Pansino)

A photo of the experiment conducted in the team's new laboratory. (Source: Stephen Pansino)

In one experiment, a tank is filled with gelatin, which acts similar to the Earth’s crust. Inside the gelatin, a crack filled with oil grows upward towards the surface, which represents magma moving through the crust. A camera in front of the tank tracks what happens inside. An array of cameras positioned above the tank is used to make a 3D model of the surface. As the crack moves up, it begins to affect the surface, which can be measured with these models.

A set of schematics and accompanying photos of experiments. Magma chambers (at the left the red circles; on the right the black balloons) can increase or decrease the pressure on the surrounding gelatin (which represents the Earth’s crust) when they inflate or deflate. Cracks in the crust filled with magma (a.k.a. dikes), can ‘feel’ this pressure and change direction or shape accordingly. In general high pressure tends to attract dikes and low pressure repels them. (Source: Stephen Pansino)



Funding Sources: 

  • Earth Observatory of Singapore
  • ENSTA ParisTech

Project Years: 


Meetings & Abstracts: 

Pansino, S, & Taisne, B., Can a dike “feel” a free surface?, Abstract V43B-3109 presented at 2015 Fall Meeting, AGU, San Francisco California 14-18 Dec.

EOS Team: 

Principal Investigator



Adel Emadzadeh, Asian School of the Environment, Nanyang Technological University

Allan Derrien, Institut de Physique du Globe de Paris (IPGP)


Kim Pham, ENSTA ParisTech