Dataset for Turbulent Flow in Binary Barchan Interaction: PIV_Single_Dune

Published: 21 November 2023| Version 1 | DOI: 10.17632/hyrfnpf8hc.1


The following folders contain data obtained from two-dimensional Particle Image Velocimetry (2D-PIV) experiments for each region where measurements were conducted. Due to the domain size and experimental constraints, the domain containing the barchan dune interacting with a turbulent water flow was divided into three sections (parts). Along the folders, these portions of the domain are referred to as part_1, part_2, and part_3. Cross-correlated PIV data, in vc.7 format, obtained from the commercial software DaVis, are made available. However, due to size restrictions of this data repository, only 10% of the total of 2,000 temporal velocity fields are included. Post-processed data, saved as Matlab matrices, are also made available, corresponding to the total of the 2,000 temporal velocity fields. Additionally, Matlab codes are provided to generate the matrices from the data on DaVis and to plot pertinent results. Note that this folder contains only data from experiments on single dunes. In each subfolder is provided an explanation of its contenent.


Steps to reproduce

1. Using the file named "DaVis2MatLab.m" contained in the folder "/Data_Matlab_format," read the files from the folder "/Single Dune_PIV_vc7." Adjust the folder names to read and save the matrices depending on the user's needs. 2. In the folder "MatLab Codes," look for a file named "Mean_velocity_two_dunes" and change the variable "part" depending on the part number or section of the dune to be read. Before running "Mean_velocity_two_dunes," the user should perform an average in the third dimension on the matrices, depending on how many time fields the user wants to incorporate in their analysis. 3. In "Mean_velocity_two_dunes," replace the number (1 to 3) of the variable 'part' according to the region being processed. 4. Verify in "Mean_velocity_two_dunes" the number of zeros in both rows and columns; this depends on the size of the matrix named "part" obtained in Step 2. The number of zeros (null values of velocity) depends on the size of the matrices. For this purpose, look at the variable Vx and determine how many columns exist. 5. In "Mean_velocity_two_dunes," verify the variable "limits"; based on the variable named "Vxb1nozero" obtained after running the functions named "killzero" and "formatocampo," determine the number of columns containing zero values. Check the zeros in rows and columns; find the number of zeros in rows counted from the maximum row. The goal is to match this pattern of zeros between parts 1, 2, and 3. 6. Plots of the mean velocity fields and profiles of the velocity components are obtained.


Universidade Estadual de Campinas


Dune, Granular Material, Turbulent Flow, Particle Image Velocimetry


Fundação de Amparo à Pesquisa do Estado de São Paulo

2020/15624-3; 2018/14981-7