Targeted CT: Predictive Digital Rock Physics Models Without Segmentation

Project Cover Image


In order to understand measurements from geophysical surveys, it is important to study rock samples in a laboratory setting. Digital rock physics (DRP) is a non invasive way estimate these same properties using numerical models. We have introduced a workflow to estimate rock properties, and this repository accompanies a manuscript where we explain the method. Here, we include our CT scans of four Berea Sandstone plugs which are used in the manuscript by Goldfarb, Ikeda, Ketcham, Prodanović Tisato (in prep). Recorded in the manuscript, we predict sample densities within 47 kg/cm3 of laboratory measurements, and porosities are estimated within 1.4 porosity percentage/units of the laboratory measurement. We name our method "Targeted"", or "segmentation-less". With it, we can quickly and accurately estimate rock properties. This includes large samples, very small samples or oddly shaped samples, and rare samples. Processing samples with this technique is less arbitrary than "segmentation" (a common digital rock physics processing method), and is less cumbersome than some physical laboratory testing. DOI: 10.17612/9X96-AC88


Usage Information



  • Masa Prodanovic (The University of Texas at Austin)
  • Richard Ketcham (The University of Texas at Austin)
  • Nicola Tisato (The University of Texas at Austin)
  • Ken Ikeda (The University of Texas at Austin)


Feb. 13, 2020


ODC-BY 1.0

Digital Object Identifier


Data Citation

Download Project

The downloadable archive contains all project data; the size of the archive file for this project is 1.39 GB.