Question about topology


Hi. If you look at the attached picture, you can see a few curves. I want A1 and A2 to touch B1 and as you can see they fall just a little short.

I formed A1 and A2 by doing a tween between the outer verticals curves. This is the front view (BTW).

In the side view, A1 and A2, are straight up and down, as they are in real-life. (Curves represent stringers for a nose section on a fuselage.)

Attaching the 3DM file too.

Question is this: How can I make these curves touch (easily) and retain their proper vertical straightness (on side view) AND have their proper curvature as view from the front.

(By “easy”, I mean is there something like “connect” or “match” that I am not doing right… I like how “matchcrv” works, but it only works when dealing with ends of curves. I am not sure about “Connect”. It seems like I need something like matchcrv(edge)? But then, I would need to have a surface already drawn.

I run into this issue a lot. I’d be grateful for any help. Thank you.Q20200801.3dm (133.5 KB)

Thank you.

You can use the Extend command for this.

1 Like

Another way is to use Sweep2 to create a surface (using B1 and B2 as rails and the outer verrtical curves as profiles), then use ExtractIsoCurve on that surface to get your A1 and A2 curves.

Hello - if you are certain about the ‘target’ curve, then I would mark the Int points in top:

Move the two end + 1 or +2 control points of the curves to these locations:

Then adjust the points to make the curvature graph look nice.
Incidentally, just poking around what you have, these curves could match one another a lot better and get you a better surface:


1 Like

Hi. Thank you for each of your responses. Each of these are good ideas; I am hoping to find something that is more deterministic Let me explain. Let me start with more screenshots:

Here is the front:

Here is the side:

And here is the perspective:

(If you are wondering why did I layout my topology like this? Because, I have heard or read that it is best to layout topology, as it is constructed in read-life. As example, B1 and B2 represents how that specific panel bends and warps. It has the weird curvature at the very beginning - points 1 and 2. [it’s an A330 btw].)

Anyway let me respond to each:
Daniel Piker: Your suggestion is what I was looking for, but curve A1 (or A2) will not snap to B1 as I Extend. I have set my snap options to “Near”, hoping that it would snap to B1 as I move the end point vertically. However, it does not recognize B1 as a boundary.


Menno Deij:
I have tried this approach, and it does work. Problem is: I want to get away from creating intermediate surfaces or curves as much as practical.

Pascal Golay: I have tried this approach too. And this works too. Problem is: I can never know for certain whether my point lands on one line or the other. Here is example:

Here, I am in Top view and I drop a point at Apparent Intersection of B1 and A2.

In this example, the point falls onto A2, which is what I do not want.

(I need the point to fall on B1, so that I can snap the end of A2 to the point on B1).

Also, I know what you are saying about the points on the red curves not being placed well nor consistent. Plus, they result in bad surface. I was wondering about that too, and have additional questions about it. Can we tackle that subject separately?

Basically, is there something like Connect, but Connect will latch onto B1 AND the vertical curve retains vertical “ness” (they have to remain vertical,because they represent stringers…and I need them vertical, so that I can compare and contrast curvature from photos and line drawings). I know that A2 will lose some of its curvature when I “connect”, but I can fix the curvature, but I need topology to remain constant.

Am I explaining it well/correctly? (I am a computer software developer - not Engineer) ? Is it possible? Thank you for reading and any thoughts or ideas?

Is there a script that might help out? Am I just not using “Connect” correctly?

Again, thank you for your time and reading all of this.

Also, here is updated 3DM: Q20200801A.3dm (55.1 KB)