The transport of particles by turbulent flow is quite common in nature and industry. Some examples are the transport of sand by wind in deserts and by the water in rivers and oceans, the transport of sand in canals and dredging system and the pneumatic conveying of grains in the food industry. In horizontal flow, this transport can occur as a moving layer of grains, known as a bed load. Under certain conditions, the granular bed can be deformed, leading to the formation of ripples and dunes. Among the dunes formed by this type of interaction, the barchan dune has been the object of studies by scientists and physicists in general. Barchans are formed in areas with predominantly unidirectional fluid flow and limited amount of granular material (Bagnold (1941)). We can find barchan dunes on several environments, such as, deserts, rivers, bottom of oceans and also on the Martian surface. In addition, they belong to dune fields, where they can influence and be influenced by other dunes with different sizes and velocities (Bacik et al., 2020; Assis and Franklin, 2020; Assis and Franklin, 2021). So far, experimental studies made use of monodisperse particles to understand the interaction between dunes. However, dunes in nature consist of polydisperse grains. How the polydispersity within the dune may affect the interaction and how are the motion of particles with different sizes within the bed are questions that remain unanswered. Therefore, the problem is still open. This project proposes an experimental study of dune-dune interaction with polydisperse grains and particle segregation during this interaction. Because the problem is complex, we propose to inquire into two-dimensional dunes, which behave roughly as a central slice of barchan dunes. The experiments will be carried out in a two-dimensional circular channel where 12 equispaced paddles drive water flow by paddling the free surface at pre-determined speeds. Piles consisting of bidisperse mixtures of grains will be placed on the bottom of the flume and afterwards will be deformed by the fluid flow into dunes. To track the bedforms we shall use two cameras: the first one will corotate with the bottom wall and provide information about local sediment transport; the second one will record the long-time evolution of the morphology. The morphodynamics will be tracked by image processing. This work will have two parts: the first one, described in this proposal, will be conducted at the University of Cambridge and will take around four months to be executed. It will focus on the effect of varying bidisperse mixtures of grains to investigate the dynamics of single and pairs of dunes. The second one, to be submitted after the conclusion of the first part, will be carried out at the University of Colorado - Boulder, and will take 8 months. The possibilities for follow-up work include further laboratory work on dune interactions, numerical modelling work on particle motion during dune interaction, and calibrating field work at the Great Sand Dunes National Park. In both parts, the student will be supervised by Prof. Nathalie Vriend. Professor Vriend's research focuses on the geomechanics of granular materials, wave propagation in particulate materials, segregation of an avalanching granular material and migration and interaction of dunes. Other than showing some mechanisms of segregation, layer formation, and their influence on barchan-barchan interactions, this work will help us to understand how different structures are formed in a dune, generating the strata usually employed to deduce the ancient history of barchan fields on Earth (Alvarez et al., Phys. Fluids, 2021).
News published in Agência FAPESP Newsletter about the scholarship: