Insert pore space instead of a sphere/grain?

[akkendrick]

I'm sorry if I missed this in the documentation, but is there an easy way currently in the code to insert a hole in a grain? I think I could do this with insert shape but there might be an easier way. My goal is to leverage porespy to systematically add a secondary/much smaller porosity to an existing segmented image for reference.

[jgostick]

First of all, I'm loving that this is a 'discussion' rather than an 'issue'. I will attempt to answer.

It sound like you have a tomography image of large features (i.e. rock grains) and you wan to add holes to the grains, or something like that.

You have a few options:

1. You can generate a image with much smaller features than your grains, then multiplying the images together...being sure to 'swap' phases as needed, so that you set void to solid (e.g., im = ~im), then multiply, then revert back. Below is the basic idea and result:

import porespy as ps
import matplotlib.pyplot as plt

im = ps.generators.blobs([400, 400])
im2 = ps.generators.blobs(shape=[400, 400], blobiness=3)
im3 = im * im2
fig, ax = plt.subplots(1, 3)
ax[0].imshow(im)
ax[1].imshow(im2)
ax[2].imshow(im3)

2. You could use RSA, which takes an input image (your solid grains, with solid set to 0) and inserts spheres into the 0 regions until a specified volume fraction is reached.

im = ps.generators.blobs([400, 400])
im2 = ps.generators.RSA(im=im, radius=5)
fig, ax = plt.subplots(1, 2)
ax[0].imshow(im)
ax[1].imshow(im2)

[akkendrick]

Yes that is exactly right thanks! It looks like the first approach should definitely work, I just had to sort some boolean/int issues on my end.

[jgostick]

glad to help. be sure to mark my answer as correct :-) so your question shows up as solved for future generations.