Sanding the plug of the cabin

Besides optimizing the control mixer and working on a second prototype for load testing, I’ve been working sporadically on the plug of the cabin. The rough shape is there, but many necessary and essential steps follow.

The video on making a molded fuselage by The Arnold Company  is very inspiring and shows very well how much work there is behind a good plug:

I follow essentially the same steps, but decided to cover the surface with veener instead of glass fabrics. It’s going to be the same procedure as last time for the center section’s plug. The reason is that veneer smoothens somewhat and filling the surface should be easier. Enough abachi veneer is ordered and should arrive soon.

Before covering with veneer, the XPS core needs to be sanded. Finishing a plug is a multiscale process, in the sense that you start with a rough shape and in each following step you increase the quality of shape and surface. At the current step it’s not about achieving the best surface quality, but rather about getting the right shape without kinks in curvature. Surface quality is achieved later by sanding the veneer and filling.


This picture remebers me to Ed Sheeran’s song Shape of You:

„Girl, you know I want your love
Your love was handmade for somebody like me
Come on now, follow my lead
I may be crazy, don’t mind me
Say, boy, let’s not talk too much
Grab on my waist and put that body on me
Come on now, follow my lead
Come, come on now, follow my lead

I’m in love with the shape of you
We push and pull like a magnet do
Although my heart is falling too
I’m in love with your body…“






Prototype of a control mixer

Having a cardan joint for the control stick, a mixer for the elevons was the next consequent step. There are several points to consider when designing a mixer:

  • proper deployment angles of at least ±20°
  • correct direction: „up“ and „right“ should deploy the right control surface „up“
  • comparable elevator and aileron angles (good starting point for a maiden flight)
  • negligible differentiation between up-down and left-right (differentiation should be set near the control surface up)

After a search in the internet, and in particular RC-Groups which had some posting on mixers including Fauvel- and Mitchell-wings, I came with following mixing principle (looking from top to bottom, up is front and top is aft of airplane):


The movement of the control stick (top center) is translated to rotation of two levers (bottom left and right). The angle, distance between control stick and levers determines the kinematics. The behavior is non-linear, as a circular motion (lever roation) is coupled to a prescribed distance between control stick and joint at lever. Choosing the right layout is not obvious, but can be done using, for example, a CAD program by playing around. Just in case you are wondering, the angle between control stick and levers is pretty near but not equal to 45°. As I wrote, it’s not obvious…

So much on the theory. Last time we made the cardan joint for the control stick. This time we finished a prototype for the mixer. We started by looking for suitable parts in our workshop: a couple of standard radial bearings and some Igus GmbH plastic spherical bearings:


The CNC-machining of AW-7075 aluminum worked like a charm! The complete mixer works as expected and looks like this:


The proper mixing is exemplified in the following video. Note the symmetric rotation of the levers on „up“ and „down“ and the asymmetric rotation on „left“ and „right“:


Cardan joint works

After some problems with the spindle, we managed to finish the prototype of the cardan joint. The thread of the tool holder of th spindle was damaged and we had to get a one… Anyway, we learned a lot regarding the production of such a complex part: It needed more than 16 steps and some didn’t work out as expected. The result is more than good enough for playing around and testing a mixer:


Here’s a video of it at work:


Next step is making a prototype of the mixer. The material for the levers is ready and waiting!


Back again: Machining a cardan joint!

The controls of a Horten are slightly different than those of a usual configuration. The ailerons are used for both pitch and roll. A consequence of this is that control inputs have to be mixed. There are several mixers around, but the one I like most is very simple:


The movement of the control stick is converted by levers into either symmetric (pitch) or asymmetric (roll) movement of the control surfaces.

In usual airplanes the control stick is supported such that pulling/pushing acts directly on the elevator, while moving to the left/right pushes rods that are connected to the airlerons. This presupposes that the center of rotation of the stick is above the pushrods of the ailerons, which is easy to realize. See for example the ULF2 build log of Matthias Zahn:


For Schneewittchen’s mixer, the situation is slightly different. The connection between the pushrods and the stick is above the bearing. It has to have a small footprint to keep everything compact and small. So the callenge is to have a small cardan joint at the bottom of the stick. I designed a prototype for testing:



Today we started to produce the first piece of the joint:



Last foam pieces on the plug

Two complex shaped foam pieces were missing on the plug. We used our new CNC machine to mill these:

Though the machine makes most of the work, a good sketch is always needed. And this isn’t easy. This is even more evident when  three dimensional pieces are milled. We used a mixture between QCad and FreeCAD to design the parts (both open source programs).

The piece at the front of the cabin was straightforwad to draw:




The piece at the aft has a very complex shape: It has on the top a cone shaped section, which is merged to something that looks more like a trailing edge of a wing. Drawing a good transition between these two quite different shapes is challenging. We, thus, decided to keep the piece simple but still providing the basic shape. We will sand it to its final shape:



Time consuming details

„Kleinvieh macht auch Mist!“

This little nice german proverb—which means translated: „Flock makes also muck!“— describes very well the last weeks. Though the CNC machine technically runs since a couple of weeks, not until today most important details were solved. We had to take care of having a proper workholding. This included a tooling table with many drillings and face milling the surface. It is ready up to some changes in the cooling system and vacuum cleaner:


We had to learn a lot about milling. Without experience it is difficult to have a feeling for a good feed rate and turning speed of the spindle. And in case you are wondering: Yes, we broke a couple of tools while testing. At least nobody got hurt. I guess we found now a good combination for wood. Look at those nice chips:


It is ready to work for Schneewittchen! There are a couple of styrofoam pieces and templates for the cabin’s plug waiting to be machined…


CNC milling machine runs

It took us roughyl three to four weeks to build the kit of our new CNC machine. Not only the mechanics took effort to get together and aligned, but also assembling the control unit and wiring the drives was quite some work… Anyway, though we moved the machine a couple of times before to set it up, today it had its first „mission“:

From now on, we can focus again on Schneewittchen. The cabin plug is waiting…


Meccano for big boys

Our new toy is finally here. A couple of hours before we picked „The Kit“ up, we realized that we should paint the room:

We’ve built a couple of large LEGO® technic sets with our kids, but this is by far our largest kit ever. This is the result of one and a half day of work:

We’ll need a couple of days more to finish it…


Foam here, foam there, foam everywhere

The tamples we got a couples of days ago have been very valuable. We started to cut the core for the plug of the cabin:


We had some problems at the beginning, because the approach we decided to use didn’t work out as expected. Originally, we though to cut the foam perpendicularly around the template. Heavy waves was the result:


hence, we decided to use the tamples on both sides. This has as a consequence that a tiny bit of extra material is cut when the shape is spherical. This tiny error is acceptable, as the plug core will have to be slightly adapted to our taste anyway.

The plug looks everyday more and more like a cabin:

We started a small video series on how the mold is made: