More details

I wanted to know more!


Now, lets look at some more details.

Pilots position

In all Hx-versions he is lying on his stomach. Might look weird, but is already very common in hanggliding. The HXb has a slightly different pilot position than the hanggliders. The pilot is a bit more upright (30° instead of nearly totally flat). It gives less strain in your neck during long flights. And your upwards view is 30° further upwards. I already have a complete page about proned pilots. Go see that page to better understand the advantages and misleading prejudgments of this system.

Later more about the system of fixing the pilot into the glider.

Picture of the Horten HXa, while the pilot is "wearing" the glider.

He is standing with his back at 30°. (Picture by Reimar Horten, published in book "Nurflugel", permission to use by Peter F. Selinger)

I deleted the right part of the wing to better let see the pilots position when standing on the ground.

Fernando Sairez mails to my question:
[Me]: "Does the pilot have enough sideways view during flight?"
[Fernando]: "The pilot eyes are about 4 centimeters (1 1/2 inch) above the upper wing covering surface (his head in the small canopy), so he can see anything around him and above him. To the sides, his vision is limited to about 15 degrees downward. Think the view is satisfactory, but don't know for sure for I'm not a pilot."
[Me]: "I see two bubbles in the glazing of the canopy. Am I right to guess that the top one is for the pilot during running and the front bubble while he is lying down?"
[Fernando]: "No, the HXb was designed to have the pilot's head in the canopy in both positions: when running and when laying in semi-prone position. The bubble allows some limited downward looking. Actually there is a dummy occupying the pilot position in the HXb in the museum, that is misleading."

Steering (first findings)

Fernando Sairez mails to my question:
[Me]: "The mixer of the controls. How did they do it in the HXb?
[Fernando]: "These I didn't analyze totally, but think it has some resemblance to a very very simplified version of the controls in the H-IV. There are 4 pushrods: two form a "Vee" in the center, two run parallel to the spars outwards: when you want to "pitch up", you pull on the Vee, it "opens" (widens its angle) pushing outwards the pushrods, and the elevons go up; to bank, simply move the "Vee" side wards."

I still have the opinion that the mixer used in several Horten design is still the easiest to make elevon-mixer. I hope to show its working in a few 3Ds. (All drawings below are the same size. So ... if you want you can place them into a small GIF-animation. I didn't use the animation, because it is not possible to print all the positions of a animation.)

The stick has two push-rods.

They are placed at a angle. The V-pushrods are connected to the elevons (see note).

Note: the elevons were not truly connected to these push-rods. There were pushrods between the V-push-rods and the elevons. In this picture you can see that the V-push-rod is connected to another push-rod which will go to the elevon. The one at the bottom is the V-push-rod.

Neutral position.


Stick fully forwards. Both elevons are pulled in the down-position.

Stick fully backwards. Both elevons are pushed in the up-position. Question to the readers: Is this the maximum up position? If yes, look at the last two drawings. Here the stick is also at the most backward place AND ... the stick is at its most sideways place too. The elevon that went up due to the pulling of the stick is now even higher due to the sideways sliding of the stick. Is it here above the maximum position? If yes, does it not stall this wing-area in this position?

Stick fully backwards and stick to most right position. The right elevon is pushed higher and the left elevon is pulled deeper.

Stick fully backwards and stick to most left position. The right elevon is pulled deeper and the left elevon is pushed higher.

Be sure to look a bit further on this page for more details about the mixer.
If you know any easier system, pleeeeaaase, tell me.

Skid or legs / Fixing the pilot

The Horten HX's had several versions of landing gear. The HXa started as foot-launchable without skid. But to speed up the tests, due to not enough wind, they decided to use car-tows and placed a skid under the glider. The later models (b and c) were only intended as foot-launchable.

Not possible? Just look at the SWIFT. It can be landed on the feet of its pilot. Heinz Scheidbauer, Reimar Hortens testpilot,  remarked that the pilot has to be more skilled to perform such a landing. But ... I think that a few small skids (just like the SWIFT has), can solve that problem. Just, let it belly-land with some protection. The HXb and c will need a bit longer skid-legs as the SWIFT, because the knees of the pilot are in the danger-zone.

I got a reply from Fernando Sairez to my question:
[Me]:"How did they make the pilot run? There is some text about a plate that is being rotated. a plate where the pilot lies on. I hope to understand how this works. Details of hinges and locking devise and so.
[Fernando]:  "In the HXb (different from HXa) there is a steel tube hexagon that rotates about two lateral pins; the pilot enters in it and fixes himself by belts. One of the pins can be plunged in, to allow fast bail out."

He mailed also:
[Me]>You saw that there was no skid. Euh ... how was landing planned? On feet? Belly-landing?
[Fernando]: "On feet, because the HXb was designed to have a smaller wing loading than that of the HXa. The HXb had a design landing speed of 35 Km/h; with a head wind of 15 Km/h, the pilot would have to run at 20 Km/h (5.5 m/s) that is not difficult to do. (But you need the wind!!).
[Me]>In one of the articles about the Horten on (translations of Argentina articles) there is a sideview of a glider with the pilot running. He is attached to a plate. Is that the configuration of the HXb pilot? If yes ... I am right to think that the pilot has to PULL himself up to enter? He is not hinged at the hips. I fear problems using this system. What is your opinion?
[Fernando]: "No, the HXb uses a different scheme than that you mention; the pilot is hinged at the hips. In fact, the HXb is like a plane you "wear" rather than "enter in". The HXa was more similar to the system you mention, although the pilot had to pull with small force.

In your site "Nest of Dragons" in the section "Prone pilots", there is a drawing of the HXb pilot position , which is CORRECT. The dummy pilot that is in the museum has a full prone position, that is INCORRECT, they should bend its legs up and rotate the trunk up."


I nearly forgot this detail. Al Bowers did help me remember: "The H Xb didn't have any dihedral in it.  It was built flat.  The H Xc did have dihedral."

Type of elevon

Elevons with Frise function. "Huh", you ask. Just read the first part to better understand what a Frise is. If you do know, just jump to the quote of Al Bowers.

At the "WIKIPEDIA, free encyclopedia" (, you can read at the word "aileron": "An unwanted side-effect of aileron operation is adverse yaw - a yawing moment in the opposite direction to the turn generated by the ailerons. In other words, using the ailerons to roll an aircraft to the right would produce a yawing motion to the left. The yaw occurs because the down-going aileron will, as well as increasing lift, increase drag. Conversely, the wing with the upgoing aileron will see a reduction in drag as well as the intended reduction in lift.
Adverse yaw can be countered with the aircraft's rudder (a co-ordinated turn), but can also be reduced with clever design. If the upgoing aileron moves further upwards than the downgoing aileron moves down, it will create extra profile drag on that wing and try to yaw the aircraft into the turn. This set-up is known as "differential aileron". Another solution is to use a "Frise aileron", where the up going aileron also projects a section downwards below the wing, again increasing drag on the inside of the turn."

Al Bowers wrote me: "The H Xa had the Frise elevons. The H Xa also did not have a bell shaped lift distribution because of it's very small size and the need to foot launch (limited launch speed).

The H Xb used conventional elevons (no Frise nose) with the fully developed bell shaped span load. As an experiment it should have had no adverse yaw through most of the envelope.

The H Xc also had the fully developed bell shaped span load and also had conventional elevons. I also discovered that Horten intended to use a 3:1 ratio between roll and pitch in his control system, with a
small idler arm at the 17th rib. By doing that, he could use a simple push-pull rod and a little bellcrank to the elevon for his whole control system. The result is simple and elegant.

IMHO, Reimar was brilliant in so many ways. If only he had written more of this down so reverse engineering everything wouldn't be so hard. I'm sure there are many nuances and subtle things I am not
aware of that he had complete design details for..."

Fernando mails to my question:
[Me]: "In some text they mention that the elevons have some downwards dragbrake system. If the elevons goes up some sort of part is coming under the underside of the wing and it serves as brake. Any details
are welcome."
[Fernando]: "I think you mention some declarations of Rogelio Bartolini, speaking about the HXa (L'alita). These ailerons are "Frise" type. But the HXb has a different design, the elevons are like "doors", with rectangular leading edge and great "gap". Perhaps the big gap acts as a rudder-brake, because when the elevons go up, the gap widens."

Steering (Pitch - Roll ratio)

Now here comes a interesting thing. Al Bowers mentioned me that the roll-pitch ratio of the Horten HXc was intended to be 3:1 . Well, that must be the answer to some question I had. When I saw the pictures of the stick and the push-rods, I noticed that they were not at a angle of 45° from the longitudinal center-line. I did expect that when I first saw the V-pushrod-system in some picture (forgot where I did see it). It did make sense, that 45° angle. It would give a 1:1 ratio between roll and pitch.But ... I guess it is more a angle of 60° from longitudinal center-line that is being used in the HXb. Now I am wondering ... what would the ratio-result be of such a guess. 3:1 by any change?

What I expected to see (45°)

What I thought to see (60°)

OK, I didn't draw here the pushrod which is connected between the elevons and the V-tubes.

That tube can be seen here. (V-tube is the one near. Not draw tube is the one further away).


Ok, you found out that I learned a bit more about the mixer and the pushrods. Currently I think it is like this. V-pushrods angled to longitudinal axis at 70° and pushrods continuing to the bellcrank which pushes/pulls the elevon. The blue or red boxes are just sliding-guides or rotation point. Just to make visible that the green pushrod is sliding. Once I have more knowledge of how they really really look I will make a better drawing.

I did make a small EXCELL-file to calculate the roll-pitch ratio. I was not able to calculate the situation as in the HXb, because the angle of the last pushrod (the one not drawn above) is not used in the calculations.

The result so far is (I used a stick distance of 30cm (about 1foot) and a V-pushrod of 1 m (about 3 feet)):

at 45° I get a roll-pitch ratio of 1,058:1

at 60° the ratio is 1,792:1 (not the 3:1 ratio that I did hope).

I get 3:1 at angle 71,17° . Did I guess the angle wrong? I probably did. 11° is not that large to misjudge on sight. Anyway, I do think now that the 3:1 ratio was already built into the Horten Hxb.


A mail by Al Bowers:
"On your site, you ask the question about the wing fences on the H Xa. There were no fences on the H Xb, yet the H Xc had fences again. Andy Kesckes captured a comment I made on them back in 1998:

"A comment was made about the small vertical surfaces on the H Xa. Al noted that they were used as stall fences because of some control surface problems. They do contribute a lot of drag but were necessary to improve the effectiveness of the elevons. He did get rid of them on the H Xb and the drawings of the H Xc also don't show any fences. Phil Burgers noted that the Urubu (H Xvc) that is being restored in Argentina was used as a prototype tester of these fences for eventual use in the I Ae 38 cargo plane to have a positive yawing moment."

The fences on the H Xc drawing and as shown on the restored H XVc (I Ae 41 Urubu) in Argentina has the actual fences that were to be used as the experiment on the H Xc. the eventual goal was to put them
on the I Ae 38 crago plane (and they were put on the aircraft when it was flown, please see the pictures in Horten/Selinger/Scott "Nurflugel" book or on Doug Bullard's site). But in the case of the H Xa, they were necessary to prevent stall control problems. I might be able to dig that up from one of Argentine articles I have from Roberto Tacchi or Rogelio Bartolini..."

But ... in the note of Reimar Horten we read: "The model was made without having vertical fins, but Dr Nickel (of the Horten group) says that without them the airplane would suffer losses of command due to the skid (yaw) component. Horten considered that due to the low speed in which it would fly, the fin would be necessary. In later experiments, first flights were made by Bartolini, take off the fins later and this flight made by Tacchi. It was observed that the machine did not demonstrate any alteration in flight."

So...they were removed from the HXa and the flight-behavior didn't change.

Construction of D-tube

I found a old mail of Al Bowers in reply to my question of the nurflugel-yahoo-group. Just read it.

[Me] Question 2: how was the D-tube of the H X b constructed? I read about a multi layer technique. But how did they actually do it? Starting for a negative mould? Starting from a (slender) positieve mould?

[Al] Looking at the images of the IAe 38, I would assume that it was not a complete mold. When I was thinking about building the H Xc, I planned on doing something very similar. I had planned on cutting the leading edge contour of the D-tube out of 3/4 inch (about 20mm) plywood, and epoxying them to the concrete floor at the correct stations. After building one wing, I would break the plywood loose from the floor, and glue them onto the floor for the other wing. Then the entire wing could be built in the fixture with the correct twist (EXTREMELY IMPORTANT!).

Washout location

I saw this email at the nurflugel-Yahoo-group. It was new info to me, so it will probably be good info to you if you are planning to make a RC-model or design a full-scale airplane.
"[Adam Till]: Just curious if anyone knows how washout should be "properly" added to a wing. I don't mean how much should be added (that I understand fine), but rather how the wing should be twisted. Do you rotate the tip section about the 1/4 chord point (typically how I've done it in the past)? Can you rotate the tip section about the LE instead without changing the effect very much? Does it really matter?

[Al Bowers]: Just a thought: you should twist it about the hinge-line of the control surfaces. The hinge-line becomes a straight line, which simplifies the hinge and the seal of the surfaces. This way, you maximize the seal
(or at least minimize the complexity of the hinges and seals) for the control surface. Leaking pressure from the bottom of the wing to the top of the wing (reducing aero efficiency and control power
efficiency) is usually a bad idea..."

Reimar Horten always did it that way. So it MUST be a good idea...

Al Bowers"