Considering Starship's atmospheric entry, belly dive and "bellyflop", latest iteration displays four aerodynamic surfaces, two aft fins and two canard fins. Soon, this configuration will be put to a 15km altitude test, hopefully with SN8.

Reading this "How hard will it be to actuate the fins during reentry?", I wondered why attitude control had to be done the way it had, having flaps hinged along longitudinal axis of Starship.

The proposition below also goes with one degree of freedom, yet a different one (slide along longitudinal axis) which suppresses the need to fight using brute actuating power, against the aerodynamic forces encountered during belly dive.

The two flaps illustrated below have a constant fixed dihedral angle, which provides inherent roll stability, the same way a shuttlecock stabilizes itself.

Aerodynamic loads are actually withstood by linear rails and bearings, two per flap, compressive rail on the upper surface of the flap, tensile rail on the lower surface of the flap.

With one fixed attitude axis (roll), the two remaining axes are enough to steer Starship anywhere it has to during belly dive.

Pitch attitude is controlled by sliding both flaps in the same direction along longitudinal axis, shifting center of pressure relative to center of mass. Shortly before landing, it pitches Starship up during "bellyflop".

Yaw is controlled by sliding both flaps in opposite direction, inducing yawing moment around center of mass, the same way a paperclip helicopter spins.

The power required to actuate these sliding flaps is quite low and should allow weight savings. (battery size, actuator size) Going from four to two aerodynamic surfaces also reduces the number of moving parts. Nosecone is freed from canard fins, which adds payload configurations and deployment options.

The sliding rails are where tubings and wiring already pass, becoming a fairing for them.

Question is, what are the issues with this design alternative, is it viable?

enter image description here

enter image description here

enter image description here
(paperclip helicopter)

  • 1
    $\begingroup$ The biggest immediate concern is that mixing pitch and yaw could induce roll which is only counteracted by the dihedral, but this might not be an issue if you design and balance it right. $\endgroup$
    – Dragongeek
    Nov 4, 2020 at 15:57
  • 2
    $\begingroup$ I doubt confident in the hypersonic stability of any design would be enough for a manned reentry without dedicated roll authority. $\endgroup$ Nov 4, 2020 at 19:02
  • 2
    $\begingroup$ Fantastic concept though, and you've clearly put effort into mocking it up. Nice question. $\endgroup$ Nov 4, 2020 at 19:03
  • 2
    $\begingroup$ @AntonHengst, Apollo capsules were missing two attitude axis authority during reentry, yaw and pitch, yet were able to virtually build these attitudes using only roll to offset center of mass and lift vector. thanks for nice comment! $\endgroup$
    – user721108
    Nov 4, 2020 at 19:37
  • 2
    $\begingroup$ I have the same concerns about sealing the slots and about the amount of volume the slot-based mechanisms take up as I did back then. $\endgroup$ Nov 5, 2020 at 1:53

3 Answers 3


The sliding rails in and of itself could be a major issue. Some reasons include:

  1. There is a HUGE amount of stress placed on these wings. Keep in mind that they can effectively land the vehicle on these wings without any issue because they have to manage that much power.
  2. The design would require some kind of rails most likely, which would need to work at high temperatures and would be exposed. I'm kind of thinking it would be similar to a roller coaster, but straight.
  3. By your design the surfaces would have to be moved quite far up, even to the nosecone portion. It seems like it would be challenging.
  4. The lack of roll control might lead to some severe complications.

Still, it's an interesting design overall, and perhaps there is room for such a design in the future.

Bottom line, it would make the design more complicated and likely wouldn't help anything unfortunately.


The engineering required for the rails would be insane. They would need to be able to move quickly and with a high degree of precision while exposed to immense temperatures and a lot of vibration. This would make them very difficult (if not impossible) to produce and very difficult to maintain. On top of that:

  1. There is no control over the dihedral and thus drag.
  2. Pitch stability is poor to non existent.
  3. The rate of pitching is far lower (this is important because Starship does a flip and burn near landing).
  4. There is less redundancy (Starship may be able to maintain control on three surfaces).
  5. There are fabrication issues (these rails would need to run alongside both the CH4 and O2 tanks).

I Don't Think That is Feasible.

The problems with the design as I see it:

  1. The sliding wings require the entire structure to be reinforced to handle various wing positions. Also, the linear actuators have to be able to rapidly move the heavy wings/flaps back and forth while they are under tremendous load, both structural and thermal. This does not sound lighter than some flap actuators.

  2. Dynamic stability will be a real problem. How fast do you expect the flaps to move? Without a tail or canard, pitch stability will be low to non-existent, requiring the flaps to move back and forth very quickly to keep the shiny side up.

To keep an unstable vehicle going in the right direction requires a high-speed control loop that uses feedback to constantly adjust flight control surfaces to counteract any direction changes of the vehicle. On fly-by-wire unstable planes this can happen dozens of times per second. The X-29 unstable test plane required 40 adjustments per second to the control surfaces keep keep the plane flying. And if you have even a small delay in your control loop, you get increasing oscillation and failure.

So, you would need to move heavy flaps under stress back and forth many times per second. Overcoming the inertia of those flaps that quickly would require high power and heavy actuators if it could be done at all.

  1. Safety. If those linear actuators fail in any way, or even stick temporarily, you will have a loss of vehicle. These are complex moving parts under high stress and extremely high heat.

The stability issue is key, though. I have a hard time believing you could make that design stable. You could try building a scale RC model and see if you can make it fly.

But in the end, I don't think this design saves you anything - you are replacing a simple canard surface with a heavy, complex sliding rail system that moves the much larger flaps. Remember, it's not just the weight of the flaps, but the weight of the carry structure in the ship that transmits the flap loads to the vehicle. If the flaps can be in many places, the structure of the ship will have to be beefier.

  • $\begingroup$ In point 2.: "existant" -> "existent". $\endgroup$
    – user47149
    May 25, 2022 at 21:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.