Finally, some Giles G-200 progress. Finite Trailing Edge Thickness Airfoils in Rhino3D
Now that we’ve got some projects cleared off, I’ve finally had time to return back to the Giles G-200. It’s amazing to me how when you come back to a project after working on some truly difficult stuff, suddenly the things that were so hard on that earlier project are now much easier. Such is the case with the wheel pants and gear leg fairings on the Giles. David Lednicer was kind enough to suggest an airfoil for the gear leg fairings, which I redrew in Rhino using the method I described on cosine spacing your rebuilt curves. I decided to keep the gear leg as NURBS surface, as really it’s a fairly straight forward loft, and what I’ve found is that sometimes it’s nice to NOT make the entire model one T-Spline. Even so, I redrew it with 32 edges on each side (based on that power of 2 thinking I outlined), so that when I went to pare down the topology where it fares into the wheel pant, it would happen in a nice, smooth and predictable manner. I made the gear leg fairing slightly longer than it needed to be, so that I could split it using this trick I showed you here on matching T-Splines surfaces to NURBS surfaces. I also redid how I creased the trailing edge crease fade. Instead of using tsCrease with its’ limited flexibility on crease fadeout and problems of fading around star points, I decided instead to simply bunch 3 points together and manually point edit those to get the look I wanted. I then used my little trick of extracting isocurves out of my T-Spline surface and using BlendCrv to make some reference geometry to guide the curve of the trailing edge fade. I got the trailing edge isocurve as close as I could by hand, then used tsMatch to finalize it, which put it within 0.007″ of the target geometry, or something crazy like that. Some pics to hopefully make sense of all this:
Current version of the gear leg fairing and wheel pants.
One of the trickier bits in going from a conceptual model to an actual production surface is the fact that you’re going to have a finite thickness for your trailing edges, instead of an infinitely thin, perfectly sharp trailing edge. It’s much easier to make blends based on that perfectly sharp geometry, but if you go to actually make parts from that surface, you sometimes end up with a “smooshed out” trailing edge, since suddenly you are trying to take real world thicknesses and make them infinitely thin, which just won’t work. So, I figured I would share how I manage my trailing edge thicknesses in Rhino3D. Take the gear leg fairing as an example – this part is a loft made from 5 cross sections, and it tapers. Here’s what the input geometry looks like at one of those 5 cross sections:
You can see I’ve got a centerline that is in the direction of flight, a cross section of the landing gear, and the airfoil shape. The airfoil was scaled from the original redrawn airfoil, which was done with a very small but finite thickness to the trailing edge. Basically, when you first draw the airfoil with cosine spaced points, you should NOT close it off at the back – keep it open with a finite gap between the control points at the back, even if it’s just a gap of 0.002. Now, when you go and scale that airfoil to the particular cross section, the thickness of the trailing edge will change as well. So, you need to adjust it to whatever your ACTUAL thickness will be. In this case, the fairing will be made up of 3 layers of 282 carbon, which will come out to around 0.007″ per layer, when layed up under full vacuum. So, each side of this airfoil section needs to be deflected out from the centerline by 0.021″. What you want is to have that deflection be smoothly fared into the airfoil shape, and that’s what I’m going to show you. So, the first question to ask is, where do you want to start making this change? I would say, for something small like this gear leg fairing, it’s probably best to start the change at the thickest point of the airfoil. For something the size of a wing, it would probably make more sense to do this farther back, but for this, let’s go with the thickest point. How to find that? Easy, run BoundingBox, and then run Curve->Curve From Objects->Intersection. You will get points at the thickest part of the airfoil:
I repeat, draw that line off the CONTROL POINT, not the point that was created with the intersection. The point that we created with the intersection was just so we could easily discern which control point is closest to the thickest part of the airfoil. We are going to scale the control points, so everything we do needs to be in relation to the control points, not the curve. Draw an ortho line from the aft centerline that is HALF the total thickness for the trailing edge, since we will be applying this to each side of the airfoil. In the graphic below this line is blue. Then draw an ortho line that meets the first line we drew perpendicularly. This line below is shown in green:
Now, run Scale1D. On one side of the airfoil, grab all the control points on the back of the section, starting with the one that you drew the first line off. The picture below should tell the rest of the story:
Now, repeat this on the other side of the airfoil. This will very smoothly deform your airfoil, only in the thickness, so that it perfectly matches whatever target trailing edge thickness you desire. The deformation at the thickest part is zero, since the scaling is based off of that point, and it gets more and more pronounced as you approach the trailing edge. This keeps you from getting any kind of sharp kinks in your CurvatureGraph of your airfoil:
Really, this takes me less than one minute per airfoil section, so even with a complicated loft with multiple sections, the amount of time it takes to really make that trailing edge perfect is quite small.