I'm modelling a run-of-the-river power station which involves modelling several canals and intakes.

besides right now I'm trying to put some retention walls in a terrain in a way like this:

enter image description here

So, at one side of the wall the terrain has some altitude, and in the other side, a different one

This has come to be a little troublesome because I think using boolean modifiers (to produce the intersections) is a little messy and tricky in Blender, and I don't need a watertight or non-manifold model (I'm not running any simulations over it) I just want it to look right.

I was thinking, that if I think of the terrain as a sheet of paper, I would make cuts and displace up and down one part of the paper, and place a "wall" in between the cuts...

What would be a clean equivalent to this in Blender?


1 Answer 1


Like many (most?) things in Blender there are multiple solutions to any problem, and this challenge is no different. The decision on which method to use will largely depend upon underlying geometry, you need, what you plan to use the project for, what future use you might want to make of the geometry later, and at what stage you are at in the development of the project.

I've actually modeled bits of terrain where I was modeling actual parts of the US. The technique I like best so far is to start with a "circle" with four vertices, subdivide the edges, so that there is one vertex at each point where a contour line crosses the edge. I then extrude from vertices (or sometimes pairs of vertices, looking elevation to the vertices on the opposite edge. Then I fill the faces. In this method, the retaining wall poses no particular issue; the land above is at one elevation, the land below is at another. Depending upon the nature and construction of the retaining wall, I would either construct it as part of the terrain, or model it later with another object. Whether or not I would merge the retaining wall with the terrain would depend upon a number of factors. I can easily imagine instances where I might combine the retaining wall into the same object as the terrain, but generally would keep it separate.

Ive uploaded a simplified explanation of my general workflow here, along with a ~.blend file that illustrates the various steps. The first step is illustrated in layer 2, the second in layer 3, &c.

In my workflow, I would be unlikely to use a boolean modifier on the terrain, because the workflow I use would simply bypass any places where I would a boolean modifier.

  • $\begingroup$ Exactly, i would like to avoid boolean as much as possible, and model the walls and canals as separate objects. how would you do that to get an "integrated" result that looks good? $\endgroup$
    – invicente
    May 21, 2014 at 17:39
  • 1
    $\begingroup$ Well, sometimes I model something twice, that is, if I'm not satisfied with the wall one time, I'll think about the matter some more, and do it again differently. In the specific case of the retaining wall, at lest ones that will be seen close up, I do things you couldn't do in the real world, like run the terrain "into" the wall, thus making sure there is no gap between the two. Similarly, I set most buildings I'm modeling into the terrain, as well, (unless, I'm modeling one that in real life does not fit into the terrain. like some here in rural Texas, built in "pier and beam style. $\endgroup$
    – brasshat
    May 21, 2014 at 19:40
  • $\begingroup$ i appreciate your help but i think i don't fully understand your methods. Could you possibly post some screenshots to clarify? Specially in the circle method, I didn't get to see the relation with the terrain elevations. thank you! $\endgroup$
    – invicente
    May 21, 2014 at 22:21
  • $\begingroup$ It may take me a day or two, but I'll post a blend file that illustrates. $\endgroup$
    – brasshat
    May 22, 2014 at 7:31
  • $\begingroup$ invicente, I edited a description of my methods into my original answer. When looking at the blend file, be aware that the object that looks like a hexagon, is really a circle with only 6 vertices. I decided 6 vertices is enough to demonstrate the method, and I was too lazy to do more. $\endgroup$
    – brasshat
    May 23, 2014 at 9:06

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .