In all rockets that launch satellites to orbit, the satellite is always placed at the top.

Why are payloads always placed at the top of rockets?

  • 6
    $\begingroup$ For a moment, you got me stumped :) Actually, payload can be placed at the side, so your premise is false. One thing I won't do is place the sat UNDER the engines. Deep fried sats are bad for one's career. $\endgroup$ Aug 20, 2013 at 15:25
  • 2
    $\begingroup$ Space Shuttle is launched to the side, but other than that... $\endgroup$
    – PearsonArtPhoto
    Aug 20, 2013 at 15:31
  • 1
    $\begingroup$ @PearsonArtPhoto - STS' payload is in the bay, above the SSMEs :) You are clearly thinking about Buran/Energia. $\endgroup$ Aug 20, 2013 at 15:36
  • 4
    $\begingroup$ On the other hand, stratospheric balloons always have the payload hanging below the balloon ;-) $\endgroup$
    – gerrit
    Aug 20, 2013 at 15:36
  • 3
    $\begingroup$ And with manned spacecraft, there is often an abort rocket atop the craft. $\endgroup$ Aug 20, 2013 at 15:41

2 Answers 2


First of all, let's look at what the limitations are for placing objects in a rocket.

  1. The object must be aerodynamic, such that it does not influence the launch characteristics.
  2. You need to easily be able to get the objects separated from the rocket.
  3. The thrust must happen at the bottom, for maximum efficacy, straight down, and the center of gravity should align with the thrust/center of drag vectors..
  4. You don't want to overheat the satellite.
  5. Every launching system to date has multiple stages.

Okay, so given all of that, what are your options, aside from a satellite launch at the top, as you mentioned? There are two things of note which I will include.

  1. The final booster could be made in to a satellite. This is rarely done, but was done with the Explorer 1, the first US satellite. The reason this is rarely done is that for the most part, the satellite and rocket are built independently, and thus, the final rocket state can't contribute to the satellite much. In addition, this still fits the satellite being the top of the stack.
  2. The Space Shuttle was launched from the side. The Space Shuttle was meant to be aerodynamic, and could take the buffeting of the ride, but it would be difficult for a non-aerodynamic object like most satellites to take the force. It lost some efficiency by its design, although it was not a significant component of loss.
  3. Abort rockets are often placed at the top of the stack for manned spaceflight missions.

The bottom line is, putting the satellite at the top is convenient to dampen vibrations, reduce aerodynamic turbulence, improve performance, reduce heating of the spacecraft, etc. It doesn't have to be done that way, but it sure makes the design a lot easier, and there aren't any compelling cases that I'm aware of to do anything differently.

  • 9
    $\begingroup$ I might add that you want the CG of the stack to be along the thrust vector when all of your engines are pointing in the same direction. If it isn't, you experience what is called "cosine losses." The shuttle, for example, experiences large cosine losses. This comes from the fact that the SSMEs (at the base of the Orbiter) must point at the CG and some of the benefit of their thrust is lost because of this. $\endgroup$
    – Erik
    Aug 20, 2013 at 17:38
  • 4
    $\begingroup$ Note that Robert Goddard's early rockets had the engine at the top, for better passive stability. $\endgroup$
    – Phil Perry
    Jun 12, 2014 at 17:59
  • 10
    $\begingroup$ Goddard thought that would make for a more stable rocket, but he was wrong. en.wikipedia.org/wiki/Pendulum_rocket_fallacy $\endgroup$ Jan 20, 2015 at 0:37
  • $\begingroup$ A new comment to an old question: "The thrust must happen at the bottom". Not strictly true. The DARPA ALASA program was considering a "puller" configuration with the engines at top of the first stage. In fact the engines would be shared by first and second stage (A plumbing nightmare) $\endgroup$
    – Carlos N
    Jun 10, 2018 at 9:43
  • $\begingroup$ There are other ways it could be done, but the highest efficiency is to have the thrust at the bottom. $\endgroup$
    – PearsonArtPhoto
    Jun 10, 2018 at 9:46

Behold a (proposed) counterexample, the Martin Marietta proposal for the Space Shuttle External Tank Aft Cargo Carrier! (The image is credited to Martin Marietta; I found it on the Spaceflight History blog.) Illustration of a space shuttle mounted to its external tank, without boosters. Two cargo volumes are highlighted in blue: the normal shuttle payload bay and a mostly-cylindrical extension of the external tank that protrudes behind the shuttle engines

I haven't found a full version of the conference paper, "External Tank Aft Cargo Carrier" for all the cons, but David S.F. Portee's blog post One Space Shuttle, Two Cargo Volumes: Martin Marietta's Aft Cargo Carrier (1982) sums up some issues (which by my reading, come from the Marietta studies, and not from editorializing by the blogger):

The ACC's position adjacent to the Orbiter's three Space Shuttle Main Engines (SSMEs) and between the powerful twin Solid-Rocket Boosters (SRBs) meant that payloads it carried would be subjected to more heating and acoustic pounding than would those in the Orbiter payload bay. Martin Marietta proposed an ACC "environmental protection system" made up of 707 pounds of thermal insulation and a 2989-pound "acoustical barrier."


Martin Marietta assumed that, with planned Shuttle performance upgrades, an Orbiter would be able to boost 36.9 tons of payload into a 160-nautical-mile-high orbit inclined 28.5° relative to Earth's equator. An empty ACC would add 8.3 tons to the Shuttle's mass at liftoff, potentially reducing the payload mass the Orbiter and ACC could inject into orbit. If the ACC remained attached until SSME cutoff, then the payload mass the Orbiter and ACC could place into orbit would total 28.7 tons.

To my reading it would've been quite complicated to use in practice. Fairing staging was proposed to save mass after SRB burnout, but the external tank isn't destined for orbit so payloads end up needing kick-stages, etc. Still some neat thinking and when I saw this pop up as a "related question" to something else I thought it would be fun to share.


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.