Today it was revealed that SpaceX has a deal with NASA to send a Red Dragon lander to Mars, launched on a Falcon Heavy, in 2018. The plan is that it will use supersonic retro-propulsion during the final phase of descent to slow the vehicle from about Mach 4 to 0. SpaceX's reusable first stage is the only vehicle to have ever attempted the maneuver. There have been 2 successes, and several of the failures might be survivable by a blunt body lander that can roll. But this isn't Mars, and those weren't landers.

Edit: This question is mostly about supersonic retro-propulsion. I think the crux of it is does a Falcon 9 first stage re-entering on Earth that way provide adequate data to confidently predict the same process for a blunt-body lander on Mars will succeed.

Several people have pointed out that SpaceX wouldn't do it if there wasn't a reasonable chance of success, however I think it is quite possible they would send it even if they felt it would probably crash, given that they have a large-scale, long-term business plan to settle Mars. They need the data to continue to develop the vehicle. I don't like to move the goalposts when there are already answers, but I'm changing the title to 'Is it probable', as it seems the choice of 'reasonably likely' was too open and has led to very general answers.

This paper by JPL staff seems pretty upbeat about supersonic retro-propulsion:

From 2013 to 2015, through a partnership with SpaceX, NASA received its first insight into the performance of a flight-qualified propulsion system operated into an opposing supersonic freestream. These efforts focused on analysis of SpaceX first stage recovery flight data. To return this launch stage safely to Earth, operation of its propulsion system in the supersonic regime at the right altitudes on Earth to yield Mars-relevant conditions is required. To date, SpaceX has performed SRP maneuvers during recovery operations of seven Falcon 9 first-stage systems. NASA personnel have independently reviewed these data sets. Multiple flights are in the specific Mach and dynamic pressure regime required by the present Mars EDL system. While the SpaceX first stage is not Mars-like in configuration, no showstoppers have been identified for this technology.

SRP computational, ground-based, and flight data have demonstrated that aerodynamic force and moment modeling uncertainty in the SRP phase is low for steady state, Mars-relevant conditions. Some uncertainty remains for SRP operation during startup and transition to steady-state operation, but performance uncertainty during this small time period may be mitigated by robust control system design. Combined with ground-based test data, the SpaceX flight data bounds the range of SRP thrust coefficients needed for human Mars EDL. Taken in total, these computational, ground-based, and flight test efforts significantly reduce the SRP flight system development risks for Mars EDL.

Emphasis mine. 'No showstoppers' doesn't ring with confidence. And they say data sets have been 'Mars-relevant', which isn't to say they are the same as what a vehicle entering Mars' atmosphere would experience. Based on the kind of risk assessment done for missions like this, is it genuinely probable that Red Dragon will successfully land on the first try?

  • $\begingroup$ You're asking a yes/no question to a reasonably complicated scenario. Are you looking for a binary answer, or is it more like "What are the major challenges?" $\endgroup$
    – uhoh
    Apr 28, 2016 at 3:16
  • $\begingroup$ If they didn't think there was a reasonable chance of success they wouldn't do it! $\endgroup$
    – GdD
    Apr 28, 2016 at 8:06
  • $\begingroup$ Reasonably. That's a broad term, and the final part of landing on Mars is easier than the final part of landing on a barge (though there are far more earlier steps all of which can go wrong). I'd say the chance is greater than 10% but worse than 90%, so not hopeless, but definitely not on par with manned launch requirements. $\endgroup$
    – SF.
    Apr 28, 2016 at 10:07
  • $\begingroup$ @uhoh It is a yes/no answer, but it would be the explanation of why that properly answers the question. The major challenges are similar to that of the Falcon 9 first stage powered reentry, which i've asked about already. $\endgroup$
    – kim holder
    Apr 28, 2016 at 13:52
  • 1
    $\begingroup$ in July 2017, SpaceX announced they were cancelling the capability to use propulsive landing for Dragon 2 due to issues with getting it certified by NASA. This also meant canceling Red Dragon. spaceflightnow.com/2017/07/19/… $\endgroup$
    – Hobbes
    Dec 19, 2017 at 13:34

3 Answers 3


If the question is really "Is it reasonably likely the Red Dragon will successfully land on Mars on the first try?" I'd say 'probably yes' because they wouldn't accept that level of exposure and cost if it wasn't (reasonably likely).

If SpaceX then says it is, then I'd just shorten that to 'yes'.

But if they say "it's probably not going to work the first time" the way they did about landing on ships ("barges") at sea at the beginning of those tests, then I'd say 'probably not'.

So basically, I'd just defer to my best understanding of what their assessment of risk is. They seem to be doing a pretty good job of carrying on NASA's tradition of methodical and exhaustive risk assessment.

  • 1
    $\begingroup$ Further more redundancy is probably planed in case of failure youtu.be/if4Ka-tstJ8?t=18m9s . Within the 8 months windows SpaceX will send an other one, with what every went wrong fixed. $\endgroup$
    – y2k
    Apr 30, 2016 at 13:26
  • $\begingroup$ @w2k good to know! Since ( I'm assuming / hoping ) copious amounts of data will have been collected even in failure, it's not likely to be a 100% failure. And based on the buzz generated by the first few attempts at landing at sea, it wouldn't even be a 100% PR failure. A little like the Rocky movies, but with Matt Damon instead of Sylvester Stalone $\endgroup$
    – uhoh
    Apr 30, 2016 at 13:32

SpaceX has been taking a very methodical approach to development. They started with a single engine vehicle (Falcon-1) and before they even had a successful first launch upgraded the engine. (Merlin-1A to Merlin-1B). Then they upgraded to a 9 engine version. Several inline upgrades along the way (1.0 to 1.1 to 1.1 Full Thrust, with engines going from 1C to 1D versions). Then going to a three core version with 27 engines (Falcon Heavy).

For their capsule, they have been upgrading with information learned from each flight. Their manned capsule, the basis for the Red Dragon will have been flying by 2018 if all goes according to plan, landing humans on Earth to and from the ISS.

They have been experimenting on flights with hypersonic reentry of the first stage. It took a number of tries, but they have now landed at least twice (I personally would count the Jason-3 attempt as a landing, since it landed, but one leg did not lock which caused the topple and explosion, but that baby landed! But will stick with 2 at this time for proprieties sake.)

They have data from Dragon cargo reentries. They have data from Falcon 9 first stage reentries. They will have data on Dragon V2 reentries by the time they get to 2018.

They have access to the data NASA has on the atmospheric makeup of Mars from its previous landing missions.

They have demonstrated that they can deliver, and they seem to get more and more experience along the way.

There are no certainties, but they do look like they have taken the proper approach to the problem. Small steps, one after another, always learning something.


First of all, let's take a look at what SpaceX should have done prior to this point in time with the Dragon 2. The first uncrewed test is expected for May 2017. Before that, the Max Q abort test will have a limited test of landing using the propulsion (The primary will be a parachute, but some testing can be made of the landing from the landing propulsion system). These tests will be sufficient that NASA will agree to send men on the spacecraft, which seems that they have confidence in the system.

Okay, so how does that translate to Mars? The Viking landers were moving at about 250 m/s when the parachutes deployed. It is estimated that the Dragon 2 has about 1200 m/s of delta v. If that holds, there is an excess of capacity to make the landing happen. The spacecraft is rated for a 9 month stay on the ISS.

NASA Ames has done a study of the "Red Dragon" that it believes it would be possible to achieve. There is still a fair amount of uncertainty, of course, but it seems possible.

As to what the odds of success, I suspect the challenges will be with the Falcon Heavy, and not with the Dragon. The cost of such a mission will be just under $200 million. At that price tag, I am confident that SpaceX will only attempt it if they have a reasonable chance of success. So I say if they launch, they will have at least a 50% chance of success, more likely in the 80% range.

EDIT One thought I read from an article online. SpaceX has more experience firing retograde supersonic rockets than any other entity on Earth. Every flight since Falcon 9 v1.1 has tested a supersonic retrograde thrust in the high atmosphere of the Earth, which is very similar to Mars. Bottom line is, they have proven that can work, it will also probably work on Mars.

  • $\begingroup$ SpaceX has an entire business plan based on settlement of Mars. It seems to me they would be willing to send the vehicle purely as a test, for the sake of development. It has the delta v, sure, but the question is if it can maintain control. $\endgroup$
    – kim holder
    Apr 27, 2016 at 22:52
  • 1
    $\begingroup$ It will be man rated for similar landings on Earth, I suspect it'll be fine or at least close to it on Mars... $\endgroup$
    – PearsonArtPhoto
    Apr 27, 2016 at 23:07
  • 1
    $\begingroup$ Note that the Viking landers were in a broad, oblate aeroshell which greatly assisted deceleration; Dragon's cross-sectional density may be a lot different, so more of that 1200 m/s may be needed. $\endgroup$ Apr 28, 2016 at 22:57

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.