Cura shows up the printed surface in yellow, and the support material in red. The red support material is visible through the yellow around the hatches, which means that the surface is too thin. If I print it, the top of the turret would be wafer thin, or it wouldn't print at all.
An interesting discovery is that the underside of the hatches has an enclosure, this would only be of any use on a plastic model. I am wondering whether WOT get there tanks from model kits, simply by 3D scanning the sprue parts, and then assembling the parts in a CAD system. This would explain how you are then able to extract the parts separately. The moulding would prevent a 3D scanner from seeing the underside of the hatches, so it would guess at the thickness of the material behind it, so the underside of the hatches is wafer thin.
The 2mm wall approach will fail if the original 3D scanned part is less than 2mm, which to my untrained eye is what has happened.
I will see if I can master the basics of Blender, and add material in the transparent parts.
....I reverted to Tim's STL file. I will do a test print of just the top of the turret to see what Cura does with the hatch area. I suspect it will either be very thin or non-existent, but I can cut that out and recreate the hatch area in Openscad. I did try sculpting the entire turret, but as I suspected it is a nightmare shape to replicate, so Tim's will have to be made to work one way or the other. I scaled it down 'very' slightly to get it to match the upper hull. The scanned turret is probably the correct size, but I had to make some very minor adjustments to my upper hull to clear the HL gearboxes, so there is a knock on effect with the turret size, and proportions.
