I am currently writing a python script to automate a structure made of cubes forming a 3D grid. The cubes are randomly placed and there are two types of cubes.

It is a single part but I plan to assign different regions to the two types of cubes. Therefore, I first need to select the necessary cubes. To do this I need to create a region. Since I know the positions of the centres of the cubes, I can find the necessary cubes by getting a list of the faces in the structure, computing their normals and then subtracting or adding half of a cube dimension (1.25 / 2) to find the centre. Then I check if this centre is in the list of positions. If it is, then I know the face belongs to that region.

The Abaqus Python API Docs say that '.pointOn' gives the centroid of a face. However, when I look at results, I have centroids at positions such as 0.4166. This doesn't make sense to me as a centroid of a square is half its length, so the smallest possible position for x,y,z should be 0.625.

Here is the code that handles that detection:

soft_cube_faces = []
    for face in part.faces:
        normal_direction = face.getNormal(face.pointOn[0])
        centre_position = (
            face.pointOn[0][0] + (-1 * normal_direction[0] * single_cube_length / 2),
            face.pointOn[0][1] + (-1 * normal_direction[1] * single_cube_length / 2),
            face.pointOn[0][2] + (-1 * normal_direction[2] * single_cube_length / 2)
        )
        if centre_position in soft_cube_positions:
            soft_cube_faces.append(face)

Any help would be greatly appreciated. Thank you :)

I tried the code above and I expected to get central positions of cubes.