SpaceX’s Grasshopper’s Small Hop is a Big Step

An artist’s concept of the Grasshopper in action. Image Credit: SpaceX

Last week, SpaceX took the first step in its quest to build a reusable launch vehicle. The goal is to create a rocket that can autonomously return to the launch site and land vertically on dry ground. Refurbishing and reusing a rocket stage is much more cost efficient than rebuilding a new stage for each launch. The prototype vehicle that demonstrated the technology behind this idea is called Grasshopper, and it’s made its maiden voyage. 

The 106 foot tall Grasshopper, so called because of its spindly insect-like legs, is a vertical takeoff and landing test vehicle (VTVL). It’s made of a modified Falcon 9 first stage with a Merlin 1D engine. The Merlin 1D, which burns kerosene and liquid oxygen and can generate up to 147,000 pounds of thrust, is an upgraded version of the Merlin 1C that generate 78,000 pounds of thrust at sea level.

SpaceX’s Grasshopper during the test on September 21. Photo Credit: SpaceX

Last Friday’s test (September 21) was a successful demonstration flight SpaceX’s test facility in McGregor, Texas. But it wasn’t exactly a simulated landing from stage cutoff after launch. It was a small hop. The vehicle rose just six feet off the ground before landing gently on its legs. Still, it’s a step in the right direction and a critical milestone in the quest to develop a fully reusable booster system. Having a reusable first stage will also complement the reusable Dragon capsule spacecraft, ultimately making the whole SpaceX system more cost effective.

SpaceX plans to follow last week’s short hop with more small hops in the coming months. The next tests will be conservative, though larger than the six foot hop. The first tests will reach between 200 and 240 feet and last less than 1 minute. From there, Grasshopper’s flights will get progressively more daring. The vehicle will get as high as 1,000 feet. At that height, it will demonstrate its ability to hover before making a soft landing. Later tests will likely see the Grasshopper climb as high as 11,500 feet on a 160 second flight before making the same smooth vertical landing.

Meanwhile, SpaceX is continues to work on qualifying the Merlin 1D engine for flight. Right now, full-duration firing tests are pending and currently scheduled for November. If these tests are successful, the Merlin 1D will power a Falcon 9 into orbit in March or April 2013.

A SRB splashes down after a space shuttle launch. Photo Credit: NASA

The only rocket stages that have been successfully recovered and reused after launch were the space shuttle’s solid rocket boosters (SRBs). These two stages on either side of the main fuel tank provided more than 80 percent of the thrust needed to kick the space shuttle into orbit. They fell away two minutes after launch once their fuel was exhausted and splashed down in the ocean. The casings were successfully reused on subsequent missions to cut overall program costs.

If SpaceX does manage to turn the Grasshopper into a fully autonomous vertical landing rocket stage, this, combined with the reusable Dragon capsule, could see a major change in the landscape of spaceflight. Cheaper missions could mean more launches.


  1. Hi Amy,

    Good article, but if I might a couple of points:

    – ”Refurbishing and reusing a rocket stage is much more cost efficient than rebuilding a new stage for each launch.” For that to be true there has to be a market for a pretty large number of launches. For instance, work done in the 1990’s on the Delta Clipper SSTO design indicated that the breakeven point would be about 50 flights per year and that was for an SSTO (which – if practical – would be much simpler to turn around than a multistage vehicle). SO there would need to be a sizeable increase in the current market for that to work out.

    – The Space X plan seems to require the first stage to do a retrograde maneuver which is going to require considerably extra Delta-V (which requires more fuel) in addition to the extra fuel required to perform the vertial landings. It would be interesting to know what payload this reusable Falcon 9 could deliver to orbit.

    Like I said a good article and I am not trying to be argumentative only making a couple of what I hope are salient points.

    • This is a further discussion regarding the Delta Clipper after my comment below.

      The Delta Clipper assumed from the start that their business model was based on a reusable vehicle. This meant that they based their profitability estimates on how many times they would have to fly to be profitable – hence the 50 flights per year estimate.

      SpaceX has started with pricing for a completely disposable rocket, and they claim they are profitable charging $54M/flight for the Falcon 9, and $128M/flight for Falcon Heavy. So the payback model for them is quite different than the Delta Clipper.

      As an example using completely fictitious numbers:

      – They only reuse the Falcon 9 1st stage
      – The 1st stage normally costs $10M to build (a pure guess)
      – The reusability add-ons cost $5M
      – The refurbishment of the 1st stage costs $5M

      For this example SpaceX would break even after their third flight (refurbishment happens after every flight), and save money at the same price point from the 3rd flight and on.

      Swap in whatever numbers you think each cost element should be, but I think this is pretty close to the formula SpaceX is anticipating. And if this is close to their cost formula, then you can see how it has the capability to dramatically lower launch costs pretty quickly once made operational.

      • Yes Ron, but the idea of this discussion was supposed to be to get some numbers that were not “completely fictitious”

        • Hi Joe, I see the conversation you started is about economic models, not pricing, and I was pointing out that the Delta Clipper is not the right economic model to use.

          The Delta Clipper had to MAKE money using reusability, whereas SpaceX plans to SAVE money using reusability. As I mentioned (and everyone knows), SpaceX already throws away the entire rocket on every flight, and customers are already OK with that economic model, so reusing a 1st stage without significant additional costs allows them to SAVE money.

  2. Falcon heavy was to deliver–what–40-53 tons to LEO, so one would think that–with reusability factored in–you would have payloads scaled down to EELV type payloads.

      • Considering that an FH is essentially three F9’s with cross-feeding fuel lines, I imagine the capability to land an F9 would find its way to the FH as well.

        • You bring up a good point on the cross feeding.

          But, then you are expecting the three “first stage” engines (that have been acting in tandem) to become standalone entity’s that will perform simultaneous retrograde maneuvers and then closely planned vertical landings in a confine space in Florida.

          Then less than an hour and a half later the second stage would come back for a similar landing. That kind of choreography would have to be performed once a week to reach a reasonable breakeven point for reusability to be advantageous.

          Please do not misunderstand me; I am not saying that a reusable Falcon Heavy might not be workable only that it would be better to establish the workability of a reusable Falcon 9 (which would be far simpler) before moving on to a Falcon Heavy.

          I know Jim Hillhouse has a background in this sort of thing (I understand the principles but as far as detailed calculations are concerned I know just about enough to be dangerous), so Jim here is a request; will you look at what a reusable Falcon 9 could do for payload to LEO and then we can talk about Falcon Heavy.

          • I would imagine that for Falcon Heavy that SpaceX would start with reusing the two booster sections, since they drop off far earlier than the core. It may turn out that the core, especially with the cross-fed capability, will not be recoverable, or at least not back to it’s launch site.

            In regards to your earlier point about break-even point, I’d be interested to see the math behind the Delta Clipper numbers.

            For SpaceX, since they are already throwing away the 1st stage, as long as the cost for building a reusable 1st stage is not too much higher percentage-wise of a non-recoverable 1st stage, I would think they would be making back their costs after the 2nd reuse. Even if they have to swap out engines for inspection and minor refurbishment (and eventual reuse), the economics of it compared to throwing away the same hardware seems pretty straightforward.

          • My comment in response to Joe.

            First, some numbers:
            Falcon 9 mass @ lift-off – 480 mT (Falcon 9 v 1.1)
            Falcon 9 LEO payload – 13.15 mT (Merlin 1-D)
            Payload MF – 0.027
            Dragon dry mass – 4.2 mT
            Dragon max payload – 6 mT (up) 3 mT (down)

            General Points:

              1) We need to stop talking about Falcon Heavy. It’s a paper rocket at this point. This is about Falcon 9.

              2) The new Merlin engines, with a 50% boost in performance, will help with the delta-V’s.

              3) Sizable fuel margins that currently exist will be required to remain or even increased because we are talking about returning for a landing at KSC, a populated area.

              4) SpaceX will have to re-engineer the Falcon 9 first stage since it was not designed for any of the loads associated with atmospheric re-entry nor of a vertical landing. This will not be a trivial re-engineering either and will add to the current dry-mass of the Falcon 9 first-stage at cutoff. Re-engineering will have its own gotchas.

              5) Further increasing the Falcon 9 first stage dry mass will be the re-entry systems themselves. Think thick heat shield since a good deal of mass will have its re-entry energy concentrated on a small surface area.

              6) The guidance and controls issues will be very interesting.

            Here’s the info needed from someone:

              1) Dry mass of modified Falcon 9 first-stage?

              2) Down-range distance from KSC at sep?

              3) Delta-V at first-stage cutoff?

              4) Mass-ratio or launch mass and remaining mass at first-stage cutoff?

            A guesstimate can be formulated if we take a patched approach. First, the problem is broken-down into three problems:

            1) Lift-off to first-stage sep.
            2) Sep to re-entry.
            3) Landing.

            Key will be finding the needed fuel masses for all three in balance with some payload mass.

            Lift-off To S-1 Sep:
            Knowing the delta-V at first-stage cutoff and sep, the mass-ratio can be calculated from:

            MR1 = e (delta-Vsep / (g * Isp))

            Sept to Re-entry:
            First, the returning Falcon 9 S-1 will have to make up for the 910 m/s imparted by the rotation of the Earth at KSC, delta-Ver. And every single bit of delta-V imparted to the system at the point of first-stage cut-off must be zeroed-out. That means, the delta-Vret will be,

            delta-Vret = delta-Vsep + delta-Ver

            There will be (hopefully) a material delta-V associated with drag due to re-entry. So delta-Vret becomes,

            delta-Vret = delta-Vsep + delta-Ver – delta-Vdrag

            And then there is the added delta-V associated with getting the Falcon 9 S-1 on a trajectory to intercept the desired landing spot, delta-Vadd,

            delta-Vret = delta-Vsep + delta-Ver – delta-Vdrag + delta-Vadd

            MR2 = e (delta-Vret / (g * Isp))

            The final mass-ratio is that amount needed to actually safely land the F9 S-1, including any maneuvering fuel and fuel safety mass-margins.

            I’m confident I’ve missed a few points that might have a material affect on any solution.

            When I’ve asked practicing engineers who dealt with Shuttle about how much payload mass would be penalized in order for a return Falcon 9 first-stage, they just laughed and said a whole lot.

            The question to answer is whether the lost revenue from lost mass not put into orbit is outweighed by the savings of the recovered and refurbished first-stage. Looking at what I’ve written, that is going to be a very interesting loop to close.

            I’ll put this together in a spreadsheet and link to it so that others can tweak it. As all of this in analytical, the solution will be extremely rough as to almost unusable. But at least the exercise is worth the fun.

            • Thanks Jim, I really appreciate you going to all the trouble.

              If we can get a rough cut at how much payload capacity would be lost, it would be a good start at making cost trades (which would also be rough – but enlightening).

            • I think you’re over-thinking this Jim.

              First of all, the reusable Falcon 9 may not be an option for payloads that require the full payload capability advertised for the Falcon 9. Notice there are no pricing options advertised yet for using the reusable Falcon 9 (or Falcon Heavy), so we don’t know what payload range the reusable Falcon 9 will be available for.

              As to your item #4, who says the Falcon 9 was not originally built for reusability? The first launches were testing out parachutes so they could test out reusability, and the rocket itself has been described by SpaceX as “over-built” according to NASA’s structural requirements for crew. I would also imagine that a water landing is much harder on the structure than a tail-first landing, even with the mid-air maneuvering that it will have to do.

              As to your comment about the Falcon Heavy (i.e. “It’s a paper rocket at this point”), customers don’t see it that way, as they are already starting to buy them. And since a Falcon Heavy in it’s basic version is made from Falcon 9 components, it’s is far from a “paper” rocket like the SLS would be.

              As to your comment about “lost revenue”, you and Joe seem to forget that SpaceX has no problem selling payload launches on completely disposable rockets, so they have no downside here – if a customer has a payload that won’t fit on a reusable rocket, they will sell them a standard disposable one. What that also means is that SpaceX can iterate their reusability, which no doubt will take a few years. In the meantime, they will continue to dominate the disposable rocket market.

              It’s a pretty nice business model.

              • I appreciate that you think I am over-thinking this–I’m not, and not by a long shot. Someone with decades of actual flight experience like Dan Adamo would look at that and laugh himself to death at the simpleminded approach. I’m just doing the very most basic back-of-the-envelope of an approximate sol’n. The answer would be off by 1-2 orders of magnitude at least.

                In 2006, when work on the Falcon 9 began in earnest, SpaceX had just managed to finally get the Falcon 1 to not blow-up or otherwise fail during launch. And that miracle came about because of NASA-MSFC, not SpaceX, engineers. So if SpaceX engineers didn’t know to incorporate something so basic as baffles in the Falcon 1 tankage, I think it unfounded optimism to conclude that its engineers knew how to design Falcon 9’s first stage to reenter the Earth.

                As mentioned earlier, McDonnell-Douglass, Masten, MSFC, and JSC have all dealt with, and had difficulties in solving, the problems of vertical landing. Anyone with even a smattering of an exposure to guidance and controls or structures appreciates how difficult this is going to be. And this isn’t about smarts. This boils down to creativity in thinking of the unk-unk’s out there that will cause the returning F-9 S-1 to, if lucky, break-up in flight and burn or, worse case, make for a real interesting day to be somewhere between Cocoa Beach, Orlando, or Daytona Beach.

                But I am confident, given that SpaceX engineers are taking an incremental approach to learn about vertical flight that once those engineers better understand the controls challenges, that will give SpaceX’s structural engineers insight into just what loads the F-9 S-1 will face during reentry. But right now this whole talk is just so much SciFi.

                As for SpaceX’s business model, SpaceX has how many successful commercial launches under its belt? Don’t get me wrong; Elon Musk is the consummate salesman. But as reported in AvWeek by Amy Svitak, “Hitting Home”, some of those who have signed launch contracts with SpaceX are now working on their plan-B options. In particular, SES is concerned that it may have to launch on the Ariane 5 because the Falcon 9 1.1 won’t be ready by mid-2013, as promised.

                According to AvWeek, this isn’t the first slip-up by SpaceX–in 2009 London-based Avanti Corporation had to switch from Falcon 9 to an Ariane 5 to launch its Hylas 1 broadband satellite. And Iridium is rearranging plans for its 72 Iridium Next constellation to give SpaceX more time.

                SpaceX’s business model will get tested. But it is erroneous to think it has been already tested and found successful. Remember in the 1990’s when the commercial launch sector was going to make commercial space launches more affordable? I do. I remember the boastful talk well. Today those companies either do not exist or they went through bankruptcy costing investors billions, with a “B”. So lets return to the discussion of how successful SpaceX’s business model is when SpaceX has launched its first several Falcon 9 1.1 Iridium and SES missions. I look forward to catching-up in mid-2013 with you on this topic.

                • Well Jim, you are certainly right when you said:

                  But right now this whole talk is just so much SciFi.

                  Which is why trying to apply formula’s to unknown problems is definitely “SciFi”. Especially when your assumptions are dependent on business models that aren’t yet apparent. Until you know what SpaceX will be selling, you won’t know what capacities to use for your calculations.

                  As to the Falcon 1, well, that was a long time ago. SpaceX was a pretty small outfit then, but they have grown far larger, including lots of new engineers. Assuming they’ll make the same mistakes – just on a larger scale – has not proven out with the three successful Falcon 9 launches, as well as the two successful Dragon missions. And assuming they didn’t consider the reusability factors everyone is discussing prior to committing is a little speculative. So far SpaceX has show they can learn, and even think ahead.

                  As to the Aviation Week article, you misread the article – it is a common practice for satellite operators to purchase backup launches, and SES isn’t just now working on a plan B, but planned for it when they bought the SpaceX launch – satellite operators are a very conservative bunch, and they do the same for other launch providers too, so SpaceX is not being singled out.

                  As to the loads being applied to the reusable Falcon 9 1st stage after it stages and starts it’s return to Earth, I wonder how they compare to the loads applied during it’s launch? I would think the most stressful part of the post-separation phase would be right after the 2nd stage leaves and the 1st stage has to slow down without going out of control. In any case, more speculation until we see what SpaceX actually does.

                  • I don’t think I misread the AvWeek article. I know all satellite launch customers have a plan-B. But Avanti had to launch its payload on an Ariane 5 because SpaceX was late in delivering the Falcon 9. Iridium is reworking its Next constellation schedule because of concerns that F-9 v1.1 won’t be ready on time. SES is worried too for the same reason. Certainly, those launches could still go as planned. I have no knowledge as to the status of F-9 v1.1

                    The process I outlined would be what the SpaceX engineers did maybe two years ago on a coffee counter or restaurant table in trying to get a broad strokes idea of whether return to launch was even feasible. Time will tell after testing grasshopper.

  3. I am not an engineer but a few points present themselves to my mind.
    1/ The 1st stage is nearly empty and has dropped the 2nd stage when it starts to turn back. This will reduce the requirements substantially.
    2/ Musk thinks it will work. This is no guarantee. He thought just parachutes might work. It didn’t. But at least he’s trying this approach which is more than anybody else is doing.
    3/ Reusability will require a much higher flight rate. Sure. But as prices fall flight rates will rise. This is fundamental economics. Some in the launch industry have claimed there is no price elasticity, but SpaceX claims to have already demonstrated this with their already lower launch prices.

    • A couple of points:

      – “The 1st stage is nearly empty and has dropped the 2nd stage when it starts to turn back. This will reduce the requirements substantially.” To perform a retrograde maneuver the first stage must dissipate its down range momentum, and then reverse it to return to (basically) the launch site. It must then still have the fuel to perform the landing (and landing struts to perform the landing). I am not trying to characterize what the payload penalty for that would be (at this point) that is why I am asking someone with more expertise in this area to evaluate.

      – “Reusability will require a much higher flight rate. Sure. But as prices fall flight rates will rise.” The problem with that theory is that it assumes an initial market high enough (a minimum of 50 flights/year) to make prices fall in the first place and that is (sadly) not currently the case.

      But again, lets (at first) limit the discussion to what a reusable Falcon 9 can do. When that is established, we can begin to discuss a wider range of issues.

  4. Joe, there are a few things working in their favor:
    1. The F9 already has quite a few reserves in its performance envelope. The F9 1.1 with the MUCH improved Merlin 1D will have even more.
    2. They will change the flight envelope slightly. From what I understood, the first stage will reach a greater hight and will not cover that much distance horizontally in return.
    3. Slowing the almost empty stage down and returning it back it its landing site will cost a lot less fuel than it took to accelerate the full stage with the second stage and the payload on top. Plus it wont have to fight gravity. If you assume it to need about 20% of the fuel (and that is probably waaaay to much), then you can guess that the much improved performance will probably be able to cover this. If a failure occurs and they need the reserves, they will just write it off as an expendable like the do now.
    This leads me to
    4. The cost. If a throw away F9 costs X amount of money, than a reusable one can only be chaper. The whole flght rate nonsense only applies to multi billion USD government programs with huge standing armies in the ground crews to refurbish and support the thing. The development of the reusable F9 first stage will have cost a lot less than that and the operation will be cheaper than building a new rocket and engines every time.

  5. Aside from the economic of things.I just don’t feel that we have master the technology to land these things on it’s tail. There is awfully lot of crashes or close calls. McDonnell Douglas crashed the DC-X, Masten crashes its Xaero, NASA crashed it’s Morpheus… Landing a matchstick at it’s end is just asking for trouble.

    Had anyone looked at the potential down turn of the telecommunication satellite bossiness’ affect will have on the launcher market?

    The market was on a 15 year replacement cycle, and is now on the verge of a down turn. In fact, I heard that Space System Loral(which was doing very well in the last few years in the commercial market) had just had its first layoff since they climb out of their last down cycle. This coming cyclic down turn that is predicted in FAA reports, and in Arianespace’s own annual reports. With the winding down of the wars in Mid. East., Government need for some of those transponders will vanish as well.

    Elon Musk must of know this as well, that is why he is so focus on getting the NASA business in the last few years. and that is probably why the Dragon gets all the R&D money, and to be launch ahead of any GEO satellites.

    With the ISS reaching the end of it’s life soon, and the worldwide expectation of government budget cuts, I feel that the government side of the satellite market is going to shrink as well.

    In a year or two, I think we may see a business climate that is very similar to the late 90’s an early 2000’s when the end of that GEO satellite boom ended. That which killed Beal Aerospace, killed X-33, killed X-34.

    • “Aside from the economic of things.I just don’t feel that we have master the technology to land these things on it’s tail. There is awfully lot of crashes or close calls. McDonnell Douglas crashed the DC-X, Masten crashes its Xaero, NASA crashed it’s Morpheus… “

      Actually I will have to defend the DC-X. The crash was not caused by the complicated parts of the operation. To save money on the test vehicle the hydraulic system that extended and held in place the landing struts had to be mated/de-mated in the turnaround process (which was not to be the case with the actual vehicle). A new manager tried to “streamline” the turnaround process by (among other things) deleting the quality inspection step for that process. As a consequence one of the struts support systems did not get reattached. The vehicle actually did a nominal landing, and then fell over on its side.

      “Landing a matchstick at its end is just asking for trouble.”

      You, make a good point about the height vs. width of the Falcon 9 first stage, Jim’s comments above about the reusable version needing heat shielding will almost certainly raise the Center of Mass making control on landing more difficult.

      The more you look at what would be required to make the Falcon 9 reusable the more complicated it gets. That is the reason I suggested seeing the outcome of the payload analysis before moving on to the other issues.

  6. “Landing a matchstick at its end is just asking for trouble.”

    The engines should almost all of the mass, and the tanks would be near empty. That would put the C of M very low.

    For delta-v, wikipedia says that with the reusable version they plan to stage at 2 kps, or about mach 6. They need to cancel this velocity, but their first stage has a mass ratio of more than 25:1 (!). So that says it won’t take a lot of fuel to do so.

    Once that velocity is cancelled, they *do not* need to expend 2 kps of delta v to get back. They’re only 200 km or so downrange! A ballistic return trajectory will only require about 1.5 kps or so. That’s mach 4.5, which should not require much shielding.

    Alternately, if there is a landing site a few hundred km downrange (old oil platform?), it could skip the return. It then only has to cancel its original 2 kps, and can re-enter at mach 1 or less. If a skydiver in a suit can do it, why can’t an aluminum balloon?

    And air drag should be sufficient for braking all the way to the last 10 seconds of descent.

  7. The USA , in the shape of SpaceX and Robert Bigelow et al, are once again innovating new Space technologies, and, if reusability does deliver major cost/kilo reductions, there will be new users and businesses generated.

    As with computers, better tech and lower costs called into being numerous unforeseen apps.
    A kind of Moore’s Law applied to space transportation seems to be in the works- say a halving of launch costs per kilo every 5 years, to be conservative. If so, then
    Europe is going entirely in the wrong direction, and by 2018 could well be out of the launch business. Seen any dodos lately?

    Their only hope is Skylon- which is spooking out several of the ESA officials- who now say it is feasible, but for some reason will not commit to it. In 2014, they will have a last chance to compete, by backing Skylon, which is planning to test a full-up SABRE engine in 2014-5)

    If Europe/ESA does not back Skylon fully in 2014, then the UK could, should, and maybe even will, go it alone. It would cost less than a third of the development costs of the planned High Speed Train- and, unlike that project, would actually REDUCE costs to its customers.

    Space transportation is the ONLY transport area offering to reduce users’ costs.

    NO wonder politicians and bureaucrats are scared!

2 Pings & Trackbacks

  1. Pingback:

  2. Pingback:

MSL Project Scientist Details Curiosity’s Discovery of Martian Streambed

The 2012 Space Race Is On!