Merge: topology error

I am constructing the geometry of an object, which is made from three connected parts with different materials. Basically this is a simple task. To avoid unnecessary tying procedures I prefer to merge the adjacent surfaces. This is perfectly working for three surface pairs, but the fourth pair does not merge because of an announced topology error.This is confusing insofar as both surfaces are equivalent to a rectangle and have the same size.

I have produced the surface on the inner cylinder by imprinting the lower surface of the small rectangular part in the outer hollow cylinder on the inner (full) cylinder. The small rectangular surfaces merge, but the bigger surfaces do not.

Edit: Surfaces 24 and 29 do not merge.

Maybe I am too blind to see an obvious error in my procedure? Help is needed!

Thanks in advance, best,
hp

MergeTest.cub5 (85.8 KB)

Hi @hpp,

when you draw the vertices for volume 1 and 2 you will see that they are not overlapping and the surfaces 24 and 29 don’t share the same topology.

So you need to imprint one more time before you can merge.

imprint vol all
merge vol all

Should fix this.

Hi @Norbert_Hofbauer,

THX for your support and the advice. Indeed, those two additional commands do their job. Although - I do not get conforming meshes. The solution is much simpler: I have to get rid of the extra vertices in Volume 2. This can be done by combining the two bounding curves on each side to one (“composite”). Then it looks right.

I can understand how the vertices 3 and 4 have been generated. But where have the corresponding points on Volume 1 gone? Is this an automatic feature, caused by the subtraction of the little brick? I can accept this, of course, but one should know about it … it looks bizarre, at least sort of.

Best,
hp

I am not quite sure in which step you are missing vertices on volume 1.

I am afraid of being too unprecise. My somewhat nasty problem is the vertex marked red on your screenshot. Where does it com from? When I generate cylinders with Coreform, they have only two bounding curves, one on each side, that’s to say there are no vertices on the circular contours. Looking at the screenshot I suspected that the circular curves always have one vertex (with y=0). From this I concluded that such vertices must have existed on the outer cylinder regardless of subtracting the small brick, which I called - misleadingly - disappearing vertices. There might never have been a vertex on the bounding circle of a cylinder, ok, but then there is the problem of a suddenly emerging vertex on the inner cylinder.

Maybe I did something completely wrong, but I remain confused.

Best,
hp

If you create a cylinder and take a look at the Model Tree you will see that the curves that represent the circles all have at least one vertex. Every curve needs their vertices as every surface needs their curves.

You can draw all vertices with

draw vertex all color red add

Thanks again! I can follow your arguments. But that confirms my first idea that the “original vertices”, positioned at 3 o’clock, of the outer (hollow) cylinder have disappeared (eliminated?) in favor of the new vertices generated by the boolean subtraction of the small brick. I did not expect such an automatism.

Have a nice weekend,
hp