aPriori Documentation
  • 👋Welcome to aPriori
  • Getting started
    • What is aPriori?
    • Installation
    • Quickstart
  • Fundamentals and usage
    • aPriori Fundamentals
      • Data Formatting
      • Cut a 3D scalar
      • Filter a 3D scalar field
      • Initialize a DNS field
      • Data visualization
      • Cut a DNS field
      • Filter a DNS field
    • Machine Learning Tutorials
      • Data-Driven Closure for Turbulence-Chemistry interaction
      • Dynamic Data-Driven Smagorinky Closure for LES
  • API guide
    • Field3D
      • Field3D.build_attributes_list
      • Field3D.check_valid_attribute
      • Field3D.compute_chemical_timescale
      • Field3D.compute_kinetic_energy
      • Field3D.compute_mixing_timescale
      • Field3D.compute_residual_kinetic_energy
      • Field3D.compute_residual_dissipation_rate
      • Field3D.compute_reaction_rates
      • Field3D.compute_reaction_rates_batch
      • Field3D.compute_strain_rate
      • Field3D.compute_tau_r
      • Field3D.compute_velocity_module
      • Field3D.cut
      • Field3D.filter_favre
      • Field3D.filter
      • Field3D.find_path
      • Field3D.plot_x_midplane
      • Field3D.plot_y_midplane
      • Field3D.plot_z_midplane
      • Field3D.print_attributes
      • Field3D.update
    • Scalar3D
      • Scalar3D.is_light_mode
      • Scalar3D.reshape_3d
      • Scalar3D.reshape_column
      • Scalar3D.reshape_line
      • Scalar3D.cut
      • Scalar3D.filter_gauss
      • Scalar3D.plot_x_midplane
      • Scalar3D.plot_y_midplane
      • Scalar3D.plot_z_midplane
    • Mesh3D
  • BIBLIOGRAPHY
Powered by GitBook
On this page
  1. Fundamentals and usage
  2. aPriori Fundamentals

Filter a DNS field

Exercise 07

Filtering a DNS field to resemble a LES field is a really common practice in aPriori validation. This however can be tricky due to all the complexities of Direct Numerical Simulation data, their storage and their dimension. Moreover, once filtering the data, there is the problem of downsampling them. The downsample is useful because, once the field is filtered, the gradients must be computed on the actual LES grid, that hence downsamples the data keeping a smaller portion of all the dataset.

With the utilities that are in this library, the derivatives are computed taking into account the fact that once filtered, the neighboring points of each cell are translated. This allows to easily perform operations on the filtered field.

PreviousCut a DNS fieldNextMachine Learning Tutorials

Last updated 1 year ago