Interview with Institut Polytechnique UnilaSalle

This blog article presents the work of two groups: one composed of Anaïs Farrenc and Eléa Gaudefroy on the study of the movement of crystals in magma, and one composed of Jean-Marie Vient studying the nest of Torquigener albomaculosus (or puffer fish) that has physical properties influencing the flow of the current.

Hello Eléa, Anais and Jean-Marie, before starting this interview, could you please briefly introduce yourself and your education background?

[E] Hello, my name is Eléa and I’m 21 years old. After having passed my Scientific Baccalaureate, I completed a preparatory class in BCPST (Biology, Chemistry, Physics and Earth Sciences) and then joined UniLaSalle in 2018 in Earth Sciences and Environment.

[A] Hello, my name is Anaïs, I am 22 years old and my school career is the same as that of Eléa.

[JM] Hello, I’m Jean-Marie, currently in the fifth and final year of the engineering program in Earth Science and Environment at UniLaSalle. I take a deepening course in energy resources. As part of my studies, I have already worked on numerical simulation projects applied to the identification of crystal movement variations (Feldspath plagioclase) according to their shape in a magma (basaltic).

Can you please describe the project you worked on and what motivated you to do it?

[E-A] Our project consists in the realization of an interactive tool allowing the understanding of different parameters defining the flow of a magma. Clearly, it is a question of creating a model of magmatic flow in which it is possible to modify parameters such as density or viscosity and to see with the help of a simulation what are the effects of these parameters on the flow of magma. The purpose of this project is to provide students with an educational and interactive tool to understand the role of physics on the flow of a magma.

This project was started last year by three other UniLaSalle students. What motivated us to take it back is that it brings together Earth Science and Computer Science. In the preparatory class we had the opportunity to learn how to code in python, which aroused our curiosity about what was possible thanks to computer science. In addition, this project gave us the opportunity to work on a 3D modeling software in order to reproduce the flow of a magma thanks to the physical and geological knowledge. The fact that the software used, Abaqus, was not initially designed to model simulations on fluid mechanics made the success of this project more uncertain and an interesting challenge.

[JM] This project takes root following the publication of articles proposing several hypotheses concerning the nest of a fish observed off the coast of Japan. The complex structure built by this fish is unique in the animal kingdom. The main hypothesis is that this nest provides egg protection at its center by reducing currents through the structure. Several questions have arisen about the element conditioning the ability of the nest to protect or not the current. It could be the characteristics of the materials used by the fish (granulometry or grazing) or the shape of the nest itself. We decided to focus on studying the influence of nest shape on these currents. A first team of students modeled the nest in 3D with a parametrized circular pattern on the 3DEXPERIENCE platform.Thanks to my experience on Abaqus acquired in the realization of numerical simulation as part of the simulation of magma, I was motivated to continue in this way. The marine domain is an environment that fascinates me and the prospect of analog simulation on the field also motivates me. Finally, I like to be in different disciplines (geology, computer science, biology). So I decided with other students to continue the project. First, on our free time, then this year alone as a full-year project.

Before starting this project, would you have thought that numerical simulation, especially Finite Elements Method, could help you understand the phenomenon?

[E-A] Before we started working with the software it was hard to imagine that numerical simulations could help to understand such a complex phenomenon. However, the more we worked on the software, the more information we learned about the phenomenon became clear. Simulations and their development require focusing on the essentials, which is sometimes difficult to do when the subject is complex.

[JM] Numerical simulation is not an integral part of our program. So this is a discipline that does not necessarily come to mind when it comes to addressing a problem of this type. However, when writing the final dissertation of magmatic project last year, it became clear that numerical simulations are relevant for naturalistic use. The ability to solve this science opens up new fields in geology, as illustrated by this work. Biomimicry is a science still young, in full development. It is about using and understanding natural structures and processes in order to adapt them for human use. And it is by experimenting with different methods that it will be possible to make rapid progress. Naturally, numerical simulation is a valuable aid in order to be able to demonstrate the hypothesis that the current is well modified by the structure of the nest.

Please describe in a few sentences your simulations: the methods used, the physics, etc.

[E-A] Our simulations are based on the use of both Abaqus and 3DEXPERIENCE software.

Remember that we had to model a magma as close to reality as possible in order to obtain a teaching tool. We decided to model the flow of this magma in a dyke. A dyke is a fracture in the rock in which the magma circulates rapidly.

We first used 3DEXPERIENCE to model crystals naturally present in the magma and we integrated these crystals in our models realized with Abaqus software. It is directly in Abaqus/CAE that we have modeled a fluid representing magma thanks to the Coupled Eulerian Lagrangian technique. For this to be consistent we have assigned to the fluid the properties of a natural magma (such as viscosity, rise velocity, density, crystal shapes, presence of gas bubble …) and used the Equation Of State model.

We then sought through simulations to reproduce the natural behavior of a magma. For example, by simulating a flow of fluid in a cavity, it is necessary to be able to observe the effect of the friction forces as well as the crystal / crystal and bubble / crystal collisions to check for inconsistencies.

Overall we worked step by step. Each parameter has been studied independently until a consistent result is obtained. They have been studied separately to be sure not to confuse the effects of the parameters. Once all the parameters have been configured, they have been added to the model.

[JM] The first step of the project was to import the nest model from 3DEXPERIENCE platform into Abaqus. We used Abaqus/CAE choosing to use the Coupled Eulerian Lagrangian technique.

We placed the nest in the center of a block representing the water line. This has been characterized by the property of water in terms of viscosity. The nest has been defined as a rigid body at first. Then we used the Volume Fraction Tool to define initial material location into the Eulerian mesh domain.

The simulations were calculated over a step time of few seconds. The output observed during the results are the acceleration and the flow velocity, and this, in inlet, above and outlet of nest.

Have you faced technical challenges? Any limitations?

[E-A] In terms of technical challenge we faced a lack of information, indeed very little work has been done on the magma flow simulation. We have also been limited by time because the appropriation of specialized software like Abaqus and its various functions is not immediate. We had to understand the different parameters and master them so that the model is as close as possible to reality. The actual difference in scale between the dyke (metric to kilometer) and the crystals / bubbles (millimetric) makes the model more complex: it is difficult to make a dyke on a real scale because on a global vision the bubbles and crystals would not have been visible and it would therefore not be possible to compare the differences in trajectory as a function of the position in the dyke. Moreover on this scale we would have had to make much longer simulations to see the movements in the magma but this is not relevant for our need.

[JM] Several limits have been highlighted by this first series of simulation. In particular, the fact that it was difficult to arrive quickly at a result close to the natural environment. The methodology was questioned as was the relevance of continuing to use Abaqus in this type of context.

That’s why the solution to use XFlow was chosen. First simulations gave encouraging results. These could quickly provide evidence to conclude on some of the assumptions.

From an academic point of view, this project was realized until then on personal time, which did not allow to have a continuous work. As the work is now done as a full-time project, it will be easier to apply a suitable work methodology.

Numerical simulation, and finite elements in particular, are not part of your training curriculum. How did you learn?

[E-A] In our BCPST preparatory classroom studies, we discussed some numerical simulation concepts. An important documentary research was necessary to enable us to acquire the complementary skills essential to carry out this project. We also used the information gathered by the previous group that worked on the subject as well as the technical instructions provided by Dassault Systèmes. However, what has helped us most to progress in the understanding of the software was to be able to exchange with Yannick Margani (Dassault Systèmes).

[JM] It is true that numerical simulation is a theme that is not very thorough in our curriculum. In Geology, analog simulation is often preferred. However, with digital development in this sector, numerical simulation is likely to become more important in the future.

For my part, my training in numerical simulation began during my 3rd year (equivalent 1st year of a post-preparation engineering cycle). Motivated by the realization of my thesis it was carried out on a model privileging experimentation and empiricism. It has of course been mandatory to go through several phases of reading and personal research courses on the subject. I was helped on the mathematical part by the specialized teachers on campus, who offered me several books to understand the basics of finite elements and mathematical simulation in general.

Then, once the first simulations on Abaqus were put in place, we had an important follow-up of Michèle Alexandre, our reference at Dassault Systèmes. We submitted our problems to her and applied her corrections and advice. In the same way, Mrs. Alexandre has put at our disposal 3DS Academy workflows corresponding to our needs, and training us on the essential tools for our project.

Similarly, during the last school year, our simulations were followed by Yannick Margani, who replaces Mrs Alexandre. His support has given us a great deal of insight into the use of Abaqus in a naturalistic context. And it is by following his advice that we chose to continue with the use of XFlow.

What about project continuation?

[E-A] The project is expanding and refining from year to year, another group will take over in 2019. Each working group is free to explore any ideas to increase the accuracy of the model. The ultimate goal is to obtain a realistic model to become a teaching aid on the flow of magma.

The main axis of improvement identified that the next pair will work on will be the rise of gas bubbles and their interaction with crystals that are currently not natural enough.

[JM] Of course, many areas of development are possible. At first, the study on the structure is to be completed. The simulations on XFlow are only in their infancy and all our hypotheses have to be proved. This represents an extremely challenging challenge for a year project.

Then, applications can be envisaged in the research sector, of course. Many analog simulations are needed to characterize all aspects of the nest. Beyond the structure, it is also interesting to determine if the materials used by the pufferfish induce different characteristics for the nests. Finally, the structure could possibly be optimized to improve its contributions and to propose a configuration favorable to better performances.

Applications can also be associated with the industry sector. Protection of marine structures or coastlines is one of many applications imaginable. Problems of coastal erosion are more and more encountered following the large and sometimes uncontrolled urbanization. Climate change is also a source of premature coastal erosion. Solutions of this type would be much less intrusive and would protect the integrity of the environment. The ongoing energy transition offers many solutions in the marine environment such as bottom turbines, or depth bases. Again it is possible to imagine structures built and protected by a solution inspired by the nest of the puffer fish.


Eléa, Anaïs and Jean-Marie participated in the Project Of The Year 2018 contest with their respective projects. You can find their full projects in 3DS Academy website: