Many websites consistently claim that the first stage of the enormous Sea Dragon rocket was intended to be reusable. While recovering a giant booster from sea after a splashdown may seem tricky, it does not break my suspension of disbelief - I am willing to believe that the terminal velocity of a huge, but (mostly) empty tank is low enough to achieve this feat.

But yesterday I stumbled upon the Vintage Space blogpost Sea Dragon is the Biggest Rocket We Ever Dreamed Of. Out of a lot of interesting design details, one seems ridiculous for me - second stage reuse! Behold:

Second Stage recovery would be similar, though its return to Earth would begin with small retrorockets in the nose since it would be coming home from orbit. After atmospheric reentry, a skirt similar to the one on the first stage would slow it to the desired impact speed [...]

Is it even remotely possible? I mean, the second (and last) stage should have (if I understand correctly) a velocity pretty close to the orbital velocity. How is it supposed to survive reentry, then? Especially when retrorockets are described as "small"? (at least "small" by Sea Dragon standards)

  • $\begingroup$ new SciShow Space video Meet the Sea Dragon: The Biggest Rocket Ever Designed $\endgroup$
    – uhoh
    Commented Jul 13, 2019 at 2:34
  • $\begingroup$ No idea if it's actually feasible for Sea Dragon S2, but one route to deal with reentry heating is just to be a heat sink -- have enough mass in thermally conductive metal that you just absorb the entire heat pulse of reentry without raising the temperature to melting. Since Sea Dragon had unusually thick tank walls, that may have been the idea for second stage recovery. $\endgroup$ Commented Jul 13, 2019 at 2:47

1 Answer 1


The linked blogpost in the question is from a series of Vintage Space posts in Discover Magazine. Amy Shira Teitel also has an extensive series of informative and enjoyable videos on her Vintage Space YouTube channel.

The video Sea Dragon is the Massive Rocket of Spaceflight Dreams is worth watching for a helpful breakdown of the Sea Dragon proposal. I've transcribed her discussion of the recovery of both the first and second stages between 06:07 to 07:19. The last paragraph indicates that Aerojet asserts that the proposed Sea Dragon implementation would be a series of solvable engineering challenges.

The first and second stages were both recovered at first just using just simple aerodynamic braking as they fell through the upper atmosphere, but the active stage of recovery would be using an aerodynamic deceleration device.

This decelerator was a reliable, and very easy system. It was a large conical flare 300 feet in diameter, that could be pressurized with the same pressurizing gas as was used in the pressure-fed engines for each stage. This also ensured that both stages would hit the water at the correct nose-down orientation, a way that would minimize the forces and stresses on the body such that it could be refurbished easily and reused, as opposed to building a new one. After the first and second stages both splashed down, they would be recovered and towed back to the assembly lagoon. There they would be refurbished, and any parts that need replacing could be replaced. But at the end of the day, most of the rocket could be just refurbished, mated together again in that lagoon, and used again on another Sea Dragon launch.

As Aerojet noted in its 1963 report, the sea dragon definitely represented a massive leap in technology, but it wasn’t one that was completely impossible. Every challenge with Sea dragon was actually just an engineering challenge, and engineering challenges could be solved.

The report NASA-CR-52817 Sea Dragon Concept. Volume 1: Summary (Aerojet-General Corp.) N88-71080m 28-Jan-1963 discusses recovery of both stages in section III-C, and the second stage in particular on pages III-c-19 to 21.

Additional discussion of recovery development can be found throughout this report dated a few weeks later.

The second-stage auxiliary engines provide orbital injection thrust for the payload, and the second stage will go into orbit with the payload. During ascent into orbit and after the second stage main engine has ceased firing, the expandable nozzle skirt will be separated and ejected from the stage. After attaining orbital condition, the payload and expended second stage will be separated. When the second stage has reached the desired position, small retro rockets in the nose will be fired to eject the vehicle out of orbit. A velocity impulse of 480 ft/sec will be required to give an initial re-entry angle of 2.5 ° from the local horizontal.

Aerodynamic stabilization and deceleration of the stage will be attained with an inflatable flare similar to the one used for first stage recovery. The size of the flare will be smaller than the flare used on first stage. The available tank pressure is lower, 50 psi, however, the recovered weight is also lower, 1.2 x 10E6 lb. This 50 psi tank pressure, available for axial load reactions,corresponds to a terminal velocity of 210 ft/sec for which a flare diameter of 240 ft is required.

Aerodynamic heating for second stage recovery will be more severe than for first stage. Use of flare material with higher thermal resistance_ such as Rene 41 mesh and ablative coatings will be necessary.

enter image description here

enter image description here

  • 3
    $\begingroup$ That picture you added (and that I originally had added) is the first stage. Look at the station numbers and compare to the station numbers in the big cutaway drawing on page 50 of the pdf. I was fooled by the pointy nose and squatty dimensions too. the 2nd stage is actually bigger than the first, because it's LH2 fueled. $\endgroup$ Commented Jan 12, 2018 at 3:03
  • 1
    $\begingroup$ @OrganicMarble as was I also, thanks! OK I'm done, please feel free to edit this further or post supplemental answer. $\endgroup$
    – uhoh
    Commented Jan 12, 2018 at 3:07
  • 1
    $\begingroup$ There is a nice write-up in Sea Dragons & Skycycles: The life and rockets of Bob Truax $\endgroup$
    – uhoh
    Commented Jan 12, 2018 at 3:10
  • 3
    $\begingroup$ Thanks for all that information! Though I am still not fully convinced if aerobraking from orbit without any shield is a good idea, now I recall that Star Raker spaceplane was also supposed to perform reentry without Space Shuttle-style tiles thanks to the "lower wing loading". But given the problems of actually built spacecraft, this still sounds like a magic for me. $\endgroup$
    – Ijon
    Commented Jan 12, 2018 at 19:07

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.