One would think turbulence is like chaos: unpredictable movements going every direction, changes in speed every second, erratic trajectories, big swirling structures superimposed to tiny vortices, and never twice the same pattern. From a non-specialist’s perspective, turbulence looks like a giant mess.
Yet still, when taking a closer look, it is not exactly the case. But for researchers and engineers, understanding and predicting the behaviour of turbulent flows remains a tricky challenge: the mechanisms are complex, the equations governing the dynamics are non-linear and their solution can only be approximated, and the scales at which turbulence develops are wide in both space and time, with strong couplings between them – according to the famous butterfly effect, the fluttering of a single butterfly’s wing can affect the weather in another place, kilometres away. This coupling of scales is one of the problems that researchers try to circumvent to better predict the behaviour of turbulent flows.
Today, the simulations involve the 3D computation of billions of points in order to calculate the fluid’s velocity from the smallest to the biggest scales, as imposed by the non-linear (butterfly) effect. Such calculations require huge processing time and resources, and even so, complex realistic configurations are still out of range of our biggest supercomputers.
Guillaume Balarac has done remarkable work in this area, especially thanks to a number of collaborations with colleagues specialising in experimental fluid mechanics. His idea was to better understand the way that the small scales influence the larger scales, and find a way of expressing this influence in numerical models without calculating every single point of the flow. In other words, the strategy consisted in finding more simple terms and equations that could be used to report the overall contribution of the small scales to the flow at larger scales, without computing all of it. The challenge lies in the reduction of computation times and resources necessary to simulate complex turbulent flows, without losing the accuracy of the predictions.
His results apply to the study and the enhancement of systems where turbulent fluids are involved, such as hydro or wind turbines, rotating machines, atomisation devices and injectors - to cite some of his most recent work.
On the 26th of November 2019, he was awarded the IMT* “Espoir” prize by the French Academy of Sciences, for his exceptional scientific contribution to the field.
Guillaume Balarac is a researcher at the Laboratoire des Ecoulements Géophysiques et Industriels in the “Modelling and simulation of turbulence” group.
*Institut Mines Télécom