This post is a small digression about a modeling technique that you may find useful.
There is a detail on the bottom surfaces of the SBD center wing: an opening, made partially in the cover of the fuselage belly (Figure 72‑1):
The difficult part of this detail is its flange, stamped in the fuselage cover. I just have two photos of this element, both of average resolution. On both of them you can see a typical circular recession, made around the opening in the belly cover. In fact, such a feature is quite common in the sheet metal design (you can see plenty of such stamped flanges in various places inside your car). This is a minor detail, too small for any serious modeling, but too large for recreating it with the textures.
I had an idea of shaping this recession using so-called displacement modifier. (I used it for a certain purposes in my previous model). It displaces mesh faces along given direction, on the distance determined by the color intensity of assigned texture. (That’s why I waited with this detail for the texturing phase). The displacement modifier requires plenty of small mesh faces. I thought that I will generate them by increasing the number of mesh subdivisions in the Subdivision Surface modifier assigned to this cover. Preparing for this, I split the mesh of bottom fuselage in the middle. This operation created two objects, representing the forward and rear part of the Dauntless “bomb bay”. I was going to increase the subdivision level in the rear part, which contains the flange.
However, after initial trials I went to the conclusion that the displacement modifier is not optimal solution for such a circular shape with rounded edges. It would require relatively high subdivision level, to obtain this shape with appropriate precision. (It would generate hundreds thousands of additional elementary faces). Too much troubles for such a small detail. Thus I decided to find another method that requires less resources.
Finally I modeled it using a technique that resembles me methods used by dentists. First I cut out in the belly cover circular area around the flange (Figure 72‑2):
To not complicate the mesh of this cover, I did it dynamically, using additional Boolean modifier and an auxiliary cone (the latter as the “cutting tool”).
Then I formed around the opening a small ring of faces, and extruded them, creating the basic shape of the flange (Figure 72‑3):
In the next step, I trimmed the extent of this mesh faces using the Boolean modifier and the same auxiliary cone that I used for the belly cover. Then I fitted external edges of this flange to the edges of the belly cover (Figure 72‑4):
Note that, thanks to the Boolean modifiers, I only had to fit these edges along the normal direction of the joined surfaces. It required less work. To further facilitate this task, I assigned a contrasting red color to the rim of the belly cover.
Finally I mapped this small detail on the general UV map (Figure 72‑5a, b):
The UV map of this patch is a simple projection from the vertical view. So far it looks good – there are no visible seams between the patch and the belly cover (Figure 72‑5c).
Figure 72‑6 shows the final result on the rendered picture:
You cannot recognize here that this fuselage cover is created from two separated objects – it looks like a single one. This is the effect I wanted to achieve.
Of course, this method of using shared Boolean “tool” for trimming both involved objects is useful for modeling single features stamped in a sheet metal. Modeling of more than two or three such objects would require too much work. (Fortunately, they do not occur too often).
You can examine the details of this mesh in this source *.blend file (this the same file that I attached to the previous post).