In previous post I have enhanced the bump map texture outcome, using two different images. This is the continuation on this bailiwick.

Have you e'er noticed that the classic stressed peel of a real aircraft is non ideally smooth? Information technology is more visible in the areas where the skin is thinner, especially on an old, "weary" aircraft (Figure seventy‑1):

0070-01
Figure 70-ane Deformations of the stressed skin (in two different aircraft)

The wing on the left (Figure 70‑1a) belongs to a SBD-four (BuNo 10518) from Yanks Air Museum in Chino. This wing was recovered separately from Guadalcanal (circa 1980), and restored a few years later. This shipping is in flyable condition (registered as N4864J), just has not flown since its restoration.

The  wing on the correct (Figure lxx‑1b) belongs to a SBD-5 (BuNo 28536) from Planes of Fame, also in Chino. This fly was also recovered from Guadalcanal, in the same time as for BuNo 10518. This aircraft was restored, registered as N670AM, and made its first flight in 1987. Since that time it has been flying during diverse air shows.

I assume that the pare of the SBDs that were flying in 1940-44 resembled the skin of the wing from Figure lxx‑1a). Note that the leading border and the central panels take no visible deformation. (Nonetheless, their skin nonetheless could deform a little in the flight). This is considering they were created from relatively thick (0.032") sheet metal. The buckling of the pare is more visible on the panel behind the rear spar, because it was fabricated from a thinner (0.025") sheet.

Information technology is quite easy to obtain this effect using textures (Figure seventy‑ii):

0070-02
Figure 70-2 Buckling of the model skin

To do it, I re-used the contents of the Rivets layers from the source Inkscape image. Notwithstanding, before I did it, I drew boosted, thick gray lines below the rivet seams. I placed these lines on a split layer, named Shadows (Figure seventy‑3):

0070-03
Figure 70-three Additional chemical element of the source prototype: boosted lines under rivet seams

One time this was done, I could compose the terminal texture epitome using these lines and clones of the Rivets sublayers (Figure seventy‑4):

0070-04
Figure 70-4 Composing the epitome of the second crash-land map (nor_blur.png)

Starting time I altered the color of the white Rivets:Dome elements, using a simple SVG filter that blackens everything. Then I blurred this composition, using another SVG filter: cascading Gaussian mistiness. (For details of this solution, run into "Virtual Aeroplane" guide, chapter about Inkscape, section titled "Using filters").

Finally, to decrease the influence of this texture on the forward part of the wing, I covered it with a slope-filled shape (Figure 70‑5):

0070-05
Effigy seventy-5 Additional adjustments of the terminal texture paradigm

Every bit you have noticed, in this limerick I re-used contents of the Rivets layers, using their clones. Using such clones in the final texture image allows me to easily modify contents of these pictures in the future. When yous alter any element in the source layer, Inkscape immediately updates all its clones. Thus I rearranged the construction of the SVG file (see the layers pane in Figure lxx‑v). I grouped all the source layers ( Rivets , Panels , Covers , Bolts , etc.) into a layer grouping named Source . And then I created another layer group, named Result . Each of its sublayers contains the limerick of ane final texture image ( Holes , Nor-Details , Nor-Blur ). Their contents is equanimous from clones of the Source sublayers, with altered opacity and (sometimes) practical various SVG filters. (Run across the source  Inkscape file)

When I am working on such a cartoon, I am drawing new elements (or modifying existing ones) on the Source sublayers. Then from time to time I export the concluding images generated past the Result sublayers to the raster files, used past Blender (holes.png, nor_details.png, nor_blur.png).

In the process of creating textures, the nigh troublesome areas are those forth seams, especially when such a seam contains a corner. Some time ago I tried to avoid breaking the peel panel edge along such a UV seam (see this posts, Effigy 67‑3). Now I can come across that this was a bad idea (Figure seventy‑6):

0070-06
Figure 70-6 Deformation of the rivets line near the UV seam

The rivets in the line that runs along the UV seam are skewed. They too have dissimilar sizes. All of this has occurred considering of the high shape distortion of the lesser faces that belong to the big wing fillet.

I placed the pocket-sized role of the fuselage inside the UV seam at the center fly. This fragment is undistorted. The remaining triangle (marked in orange in the effigy below) is an area where the mesh faces mapped onto UV surface have high baloney (Figure 70‑7a):

0070-07
Effigy lxx-seven Altering the UV map of the mesh fuselage

After some deliberations, I decided that it is much easier to bring together the few rivet lines that run beyond an UV seam, than to amend these skewed rivets produced by the electric current UV mapping. (Well, every bit you can see, the "improvement" of the seam line that I made some time ago was a bad idea). Thus I had to shift the UV seams to the outer edges, and "glue" some additional mesh faces to the heart wing (Figure seventy‑7b). This time I took care to minimize deformation of the faces that remained outside the mesh seam.

Effigy 70‑8a) shows, that I was able to precisely match the rivet lines beyond this new seam. It was not equally difficult equally I thought. Figure 70‑8b) shows the UV map of this surface area and the original epitome of the panel seams and rivet lines:

0070-08
Effigy lxx-eight Final UV mapping and drawing of the skin details

Notation that this fourth dimension merely small number of rivets occur in the highly deformed area. On the other hand, because the degree of deformation is lower than in the previous case, these rivets are not ideal, but await "acceptable", at least.

Figure 70‑9 shows both bump map images, that I mix to obtain the texture of the technical details:

0070-09
Figure seventy-9 Texture images – current state

At this moment, I filled with appropriate details all the common surfaces, and the elements belonging to the SBD-3. Equally you can also come across, I already drew some asymmetric elements on these textures. Withal, before I map them, I have to apply the Mirror modifiers to the appropriate meshes of my model. I volition do in the adjacent mail service. (I delayed this operation as long as I could, because presence of the Mirror modifiers allowed me easily alternate the model shape. (I had to modify its left side only. Blender took intendance on updating of the right side). All the same, after then many months of various checks I tin can only hope that the shape of this model "seasoned" plenty, and so I will not have to modify information technology in the hereafter).

Figure 70‑10 shows my model. (To make the result of the bump textures more visible, I significantly increased their intensity):

0070-10
Effigy 70-10 Bump maps on the model (increased effect intensity)

Strangely enough, I obtained such an intensity increase past setting control nodes of these 2 textures to negative values: Moderate:Range = -ane (nor_blur.png) and Range From Min:Min = -three (nor_details.png).

Actually, the textures of this model are symmetric, which ways that there are many missing/incorrect details on the fuselage correct side. In the next post I volition introduce disproportion to these meshes.

In this source *.alloy file y'all can evaluate yourself the electric current version of the model, and here is the source Inkscape file of its textures.