Measure CT pore distance to the part surface in Dragonfly
Find enclosed pores in CT data, measure their distance to the part surface, distinguish scan boundaries from real surfaces, and export a pore measurement table.
A pore's size alone does not describe its position within a part. Pores to Surface Analysis combines solid segmentation, enclosed-cavity detection and distance measurements in one Dragonfly workflow. It can help inspect near-surface pores in castings and other scanned parts.
What the workflow calls a pore
The solid is the segmented material. Background connected to the volume boundary is treated as outside air; a background component enclosed by solid is treated as a pore. A cavity open to the outside is therefore not counted as an enclosed pore.
The largest solid component is retained by default. Check that this matches your specimen: isolated noise or multiple separate parts can otherwise affect the segmentation and surfaces being measured.
Check calibration and the field of view
Use a published grayscale Channel with calibrated voxel spacing. The tool requires physical spacing because distances without calibration have no reliable physical meaning. It supports a 3D volume or a single 2D slice.
Where a specimen extends beyond the scan, the reconstruction boundary is a cut, not a physical part surface. Choose whether to include that boundary in the measurement. The output identifies whether each pore's nearest surface is the part or a field-of-view cut.
From CT image to a pore table
- In Source, choose the image and timestep, read the image, and check dimensions, intensity range and voxel size.
- In Segment, choose the material phase and threshold. Otsu provides a starting point; review the solid, pore and outside preview before accepting the segmentation.
- In Distance, decide whether the reconstruction boundary counts as a surface and measure the distances.
- In Publish, create the pore MultiROI and optional solid/pore ROIs. Individual preparation steps compute before publishing; Execute All Steps includes publication.
- In Analysis, choose a distance threshold to flag near-surface pores, optionally colour those pores in the MultiROI, and export the table as CSV.
Two distance conventions
Surface distance measures between the pore surface and the part surface using voxel-face centres. Centre distance measures between pore voxel centres and outside voxel centres. These are different discrete measurements, commonly separated by about a voxel step; always report which one you used.
Face-centre sampling represents flat parallel surfaces well but can overestimate distance on a staircase-like oblique surface. Threshold choice and voxel resolution also affect the interpretation. Near-surface screening is an image measurement, not by itself a mechanical acceptance criterion.
What you can export
The table includes pore label, surface distance and centre distance in micrometres, equivalent diameter, pore volume, voxel count, centroid coordinates and nearest-surface type. CSV export adds the below-threshold flag. The equivalent diameter is that of an equal-volume sphere, or an equal-area disc for 2D data.