I don't think we'll see bigger diameter Starships until at the earliest 2030. There's still a ton of work for the 9m starship to do, and it's already comedically oversized for the vast majority of payloads. Great analysis as always!
That's what I don't really understand why Blue Origin or ULA are going for bigger fairings. The largest satellite ever sent into orbit was with a regular Falcon 9 fairing. The issue isn't size, it's cost. And they're going backwards from there. Starship I completely understand the fairing size. They're their own customer. Starlink V2 (and possibly V3 satellites. Something Musk briefly mentioned during the Everyday Astronaut tour) are massive and heavy. So they need the extra size for them. But for the rest, I don't really think a bigger fairing is all that important tbh
@@snakevenom4954 I will re-review the video again, but I thought the discourse there indicated that Vulcan Centaur had a decent diameter, given the fuels and materials used, at 5.4 meters?
@@citizenblue I agree. But if the largest satellite fit into a Falcon 9 fairing, how many satellites will need a larger fairing? I'm sure some will. But enough to be a market? Not likely
Entering my 3rd year as an Aerospace Engineering student and I’ve got to say, I love your videos. These videos are incredibly informative and well thought out. They cleared up some confusions I’ve had and gave me a new perspective of spacecraft elements. Thanks ES
Dunno, the increase in diameter and height dictates a completely new structure rather than just putting a plug in an existing structure to make it longer, metal structures being a bit less limited to the moulding needed to make a composite structure like falcon.
Falcon 9 just grew in length. Growing in diameter is a completely different matter. And Starship / Super Heavy won't change the diameter either, as far too many very expensive things have to change.
Nice video! Quick addition: the mean height of a rocket is also limited by its thrust density. One of the reasons to taper a rocket like the Saturn V is to fit more engines underneath while still getting sensible aspect ratios in the tanks above it. Starship gets to have as high an aspect ratio as it does in significant part because of how unusually powerful its engines are and how packed its flamey end is with them.
Yes, this is an often overlooked factor. Short rockets don't need to pack the engines in very tightly. Electron has lots of space between engine bells, and the bells only take up a small percentage of the area. F9 is quite closely packed, and Super heavy is just insane with engines jammed in so tightly that most can't gimbal. So an 18m Starship full stack isn't going to ever be twice as tall as the 9m. It just physically can't be.
Another reason for the steps in the Saturn V was the design process. The Third stage was a reused Saturn 1 / 1B Second stage & thus was designed first. The First stage was designed second; because without it the Saturn V (originally C4) was not going anywhere. This left the Second stage to be designed last. This gave them all sorts of problems, not only did they have to fit between the other two; but weight overruns were left to them to sort out. The Second stage thus had the most advanced engineering solutions of all stages. Such as the common bulkhead between the hydrogen Lox tanks.
You forgot to talk about the heat shield! The heat shield area increases in a linear fashion to an increase in diameter. Also, a starship of greater diameter may actually slow down more quickly in the atmosphere because there’s a greater area for the atmosphere to work on. so, increasing the diameter of the starship may not only increase the amount of propellant that can be carried, but also have some real increases in the effectiveness and efficiency of the heat shield. Since the heat is getting heavier at this point because of the addition of the ablative layer, Looking for ways that the heat shield will become a smaller fraction of a total weight of the upper stage could be a real advantage.
@@EagerSpace Nonetheless, a nice video. I wonder what you think of the booster catch attempt for flight five? My question is what does the last second abort look like? If SpaceX loses control of the booster (which is never landed completely successfully, even on flight four) Where will the booster crash if it’s already close to the tower? I’m puzzled why they’re rushing this because I don’t see her upside for it, but I do see a lot of downsides.
If you add that question to my "ask me a question" video, I'll probably be able to answer it there in detail. I personally don't see much downside as long as they think they can avoid dropping it on the propellant tanks. For abort they could probably drop it off in another direction, but we've seen very few issues with running raptors.
@@EagerSpace That, and the fact SpaceX in general has their software in tiptop condition. To date I don't know of any software error ever causing major problems. In sharp contrast to that other aerospace company who's name starts with a B and charges 5 times more.
@@wbwarren57Also the heatshield has to work less because a bigger vehicle create a shockwave that is further away from the vehicle itself, it's the reason why capsules are shaped the way they are on the bottom.
Something else to consider with Starship is the reentry cross section. A shorter but wider Starship will have a different belly cross section than a tall and thin Starship. Although a larger cross section means means more surface area that needs tiles, it also means a gentler reentry. That too has some optimization equation, but it's probably more complicated due to heating limits and such.
*No need for tiles at all. just drill lots of micro holes. then pump out dry ice out of those holes to form a cold co2 insulating boundary layer. you dont even really need a pump. the heat of re-entry will cause melting of the dry ice and high pressure dry ice co2 to come out of the micro holes to form the insulating boundary layer.*
There is an advantage a bigger rocket gains on reentry too: A bigger radius object pushes the boundary layer further away from itself on reentry, therefore the heating is reduced. So you can get away with a thinner, weaker, and thus lighter heatshield than a smaller diameter vehicle.
@@EagerSpace Scott Manley has a great video on it: ua-cam.com/video/hLHo9ZM3Bis/v-deo.html NASA also published this presentation on reentry thermodynamics, that talks a little about the stagnation point, and how the radius is important, but it discusses a bunch of other stuff too: tfaws.nasa.gov/TFAWS12/Proceedings/Aerothermodynamics%20Course.pdf
Thanks for the paper. Scott does great videos but I don't watch them if its a topic I might want to cover - which is most of his videos - because I don't want to be derivative.
Fantastic video. I was thinking you would not talk about pressure, but you did. The math is simple and enough for people to start thinking instead of just talking.
There are some difficulties with rockets over 10 meters in diameter, having to do with airflow over the frontal surface area. Lengthening the rocket is much more efficient than increasing the diameter.
I would have mentioned the issue with thrust density with regard to height. Unless SpaceX comes up with continually better engines, Starship can get wider, but it can’t get taller. You no longer have room for enough engines at the bottom. 150 meters is pretty much the max you can get to with current engine technology (Raptor V3). (Assuming same diameter for the whole rocket.)
BTW, this means that if you don’t have really good engines, you are limited to a lower height. It is an important design constraint. As long as New Glenn uses the ORSC BE-4 engine, you’re probably not going to see a version reach much beyond 100 meters, as an example. This could push designers to increase from a 7 meter diameter to 9 or 12 meters, or beyond. If the need arises for a larger more capable rocket.
Agreed. I think a larger diameter will come as soon as they start generating revenue from operations (Starlink, etc...?). Starship is mostly a fuel tanker and the demand for payload is limitless. It's a KC135 refueling an A380.
@@SpaceAdvocate Actually they wouldn't be able to make an 150m design with raptor v3 and will have to go for a higher thrust in future iterations. They also seem to have those issues with fuel leaks and if they would be able to build an engine that can run at 300T of power they would be able to take Starship height up to 150m. (Also curious about raptor LEET design, which might be able to get up to 330T and allow them to increase the height of Starship a bit more than what they're currently planning)
@@RandomPerson-V They are planning to make a 150 meter tall version with Raptor V3. It will have 10,000 tons of thrust, using 35 Raptor V3 engines (~285 tons of thrust per engine). It would be able to reach 150 meters quite comfortably. The liftoff mass would for the 150 meter version be around 7000 tons, so the T/W would be around 1.43. That's quite good. Even using Raptor 2 engines, the 150 meter version would be able to lift off, at a T/W of around 1.15. Or about the same as the Saturn V.
Make the booster a cone. You can widen it at the bottom to increase thrust. It would minimise the changes to the ship, while allowing for a more powerful booster capable of throwing the ship further.
The factory would be easy enough I think, most of the stuff in there i suspect is just jigs and such which would be easy to retool. But the bays would be a pain since they are designed for the current diameter.
Why not both? 9 meters for payload is absolutely huge relative to anything we’d want to put up in the near future (say… 20 years out?). I could see a scenario in which the 9m starships handle the actual payloads, while the absolutely gigantic 12m starships are just used for refueling. Eventually, the 12m rockets could be phased in for payloads after that.
I never heard of your UA-cam channel but it popped up in my suggestions. I must say, you do an excellent job. Thank you. One tthing you didn't mention was he increase in the number of rocket engines goes up as diameter increases. Maybe the mass and thrust balances out, but it would be nice to hear you say so.
As you increase the diameter, the area and mass both (approximately) increase with the square of the radius, so it balances out. You can’t increase the height, though. That increases mass to the area for engines.
Elon recently twitted that wider version of starship will be developed at some point The moment i saw that i immediately remembered that i watched this came back to leave a comment and feed the algorithm
Ok, you concluded the most important factor is choosing rocket diameter based the amount of propellant you want. There is a strong reason to go bigger NOW not in 10-20 years. They want to bring as much propellants as possible in the tanker ships, to reduce the amount of tanker flights they need to refill for Mars and Moon trajectories. I think they should build 12m tankers, to refill 9m cargo and crew Starships. Some are claiming the imagined 9m version of a tanker would take 12-17 flights. If is that bad, they might want to go to 12m tankers asap. (for efficiency rather than to shut up the skeptics).
I don't think number of tanker flights is something they care about directly. They do care about how much refueling will cost and they care about the hassle of supporting tanker flights if the numbers per refueling get big enough to be operationally problematic, but you are saying that they should throw - let's pick a number - a few billion dollars towards a 12 meter variant just to reduce the number of tanker flights. It's not clear that at 12 meter new version is a win from a cost perspective.
One other constrain with diameter is surface of bottom where you put your engines. At some point going bigger is just getting wider if you want to keep the TWR
Excellent work. It should be noted that the maximum height of a rocket is set by the thrust per unit area of the base of the rocket. And thrust per unit area of the base will be constant with a given engine type and engine packing density.
@TheEvilmooseofdoom Yeah I wasn't very clear. How about this... Making a cylindrical rocket wider wont give it any advantage in being able to stack propellant higher.
@@TheEvilmooseofdoom You can’t add height without adding mass. If you’re only increasing the height, you can’t add more engines, and the thrust to weight will drop.
The assumption that the tank is the same thickness is simply wrong. If the tank is twice the diameter, and designed to hold to the same pressure, then the walls need to be twice as thick. Also, rockets have height limits if all the engines ar on the bottom. There's a limit to how many engines you can bit under each stage. So it's impractical to build cylindrical rockets much bigger than starship. This is why the N1 was conical despite being even smaller than starship.
@chrissouthgate4554 Fair point. A better example would be the flared bottom of the Saturn V, to account for the massive F1 engines. No matter how you slice it, you will reach a point where there isn't enough space on the bottom of the rocket to leverage enough thrust to get the rocket off the ground. A single raptor engine can only lift the weight of a column of water with its own footprint 200m tall. Since you can't fill 100% of the area under the rocket with engines and you need a thrust to weight ratio of at least 1.5:1 to efficiently reach orbit, the 150m tall starship V3 will be pushing the limits of how tall a vehicle can be and still fly.
@EagerSpace The rocket derives its rigidity from the internal pressure. The taller the rocket, the higher the necessary pressure to maintain rigidity. Basically, you want the walls to always be under tensile load and never have a significant compressive load.
This was fascinating, thanks for helping me understand that. I noticed that SpaceX’s original Mars vehicle had a 12m diameter and a substantially higher payload capacity. But then they shrank the vehicle to 9m and now it seems like they are stretching it instead of making it wider. I wonder if they will return to 12m for the Mars colonization rockets. I’m curious what impact the number of refueling rockets needed for those deeper space missions.
Thickness is a function of diameter but not height for internal pressure. Vertical thrust loads are a buckling problem that is dependent on both diameter and height. So both types of load impact the thickness. You can add stiffeners for the thrust loads, which SpaceX already does but in either case of larger diameter or larger height will like require more stiffeners.
Thank you for elevating my entire approach to rocket design. Your approach to thoughtful analysis revolutionized my understanding of the market and state of technology. You changed my world
As of a day or two ago, the vertical tank farm has been completely cut down and harvested. Do you have access to newer photos? RGV might be willing to share.
This IS a very interesting topic. Another topic I've been curious about is the issue of how many engines is most beneficial. For example, Superheavy can loose an engine or two and still boost Starship enough for it to make orbit. There is a new design with 35 raptors. What is the benefit of more engines? Obviously more thrust, at the cost a little more weight per engine, but that gets the rocket up higher faster, so I assume that cuts into gravity losses? And Starship has an upcoming design revision where it will have nine engines rather than six. And it will hold more fuel. What is the cost of the additional engines and what is the benefit? I imagine there are some interesting graphs to be drawn re. those figures.
You are correct - the faster you burn the fuel, the quicker you can stage. That not only reduces gravity losses but for the first stage, means you are closer to the launch site so it takes a little bit less fuel to get back.
I did comment that I believed it would take SpaceX a long time to go for 12m never mind 18m when this video came out, but assuming SpaceX is serious about going to Mars which I have my doubts, and really even going to the moon on a semi regular basis a 18m Starship makes a bunch of sense and not even bothering with a 12m Starship because as I mentioned the launch infrastructure would have to be built new. However what an 18m Starship does for you is enables you to lift 600T to 800T in one launch reducing the number of launches required to do refuelling in LEO using to 3 - 4 launches from like 12 to 20 when you account for boil off, which getting it done in less launches also reduces how much you lose to boil off, this assumes you are using a 9m V3 for getting to the Moon or Mars. So you would have both being produced simultaneously although you would only ever need 3x 18m Starships for refuelling, maybe a couple of extra so that your not just using them for refuelling because it would be a bit of a waste to build all the infrastructure just for a fuel truck just encase anyone want to launch fricking 600T - 800T in 1 go. Also with a 18m Starship you can eventually just leave one in orbit as a fuel depo with the power generation and equipment to do the cooling required to stop boil off being an issue so that it's viable to do an 18m Starship to Mars, again it would require 12 launches assuming 800T of cargo, but only 12 launches because you don't have to worry about boil off between launches. Well that or either you would launch a depo into orbit, because you would want to outfit a depo with insolation to reduce boil off, protection from micrometeoroids as it's going to be getting left in space for a long time, which retrofitting that in space would be problematic, and building a starship on Earth capable of that would almost certainly be more trouble than building modules and launching those modules, because those models don't also have to be built to be a rocket so it reduces the complexity and therefore the cost and you do get some economy of scale because you would be doing 14 - 16 modules to the cost of the latter half would cost less than the first half. No 36m Starship pleas, that is getting truly ridiculous, 18m is already like launching a fairly skinny high-rise building into space.
Another complicating factor is reentry heat load. Smaller vehicles are better suited for reentry because they exhibit faster deceleration and can slow down more at higher altitude.
That’s inaccurate. You want a low mass relative to the aerodynamic cross section. That’s possible to achieve on both smaller and larger vehicles. For crewed vehicles, I would think it’s easier to accomplish on larger vehicles, as the volume needed for the crew is easier to work around.
Starship factory is at the launch site. This is an advantage when growing diameter. There is too much advantage to growing diameter. It will eventually happen. There could be a reducer for using prior generation 2nd stage Starships on a booster. I look forward to changes.
There's a couple of other factors to consider. Firstly fuel boil off during longer duration flights. Intuition says that a larger volume of fuel vs surface area would experience less fuel boil off than splitting the payload into two rockets, thus proportionately less payload would have to be dedicated to managing the temperature of the fuel, with proportionally smaller radiators, etc. The thicker sidewalls of a larger rocket also help shield against the radiation of deep space, at least a little. And I'm sure the size of the rocket plays a part in reentry flight profiles and heating, although I'm nowhere near smart enough to work out whether a larger vehicle is beneficial or a bigger challenge. Perhaps here the tables are turned and the larger volume to surface area increases the heating as there's more energy to dissipate per unit area... If Musk is still eyeing a bigger rocket I presume the tradeoffs work out positively.
Is there a possibility that before they go with a larger diameter rocket that SpaceX will go for a Starship Heavy as an interim stage and then flare up the upper stage to 12 meters? Seeing a rocket with 105 engines take off will be spectacular. Also going to 18 meters will Raptor be enough or do they need something the size of Rocketdyne F-1?
@@mostevil1082 They haven't yet proven that they can eliminate the 10t+ engine shielding which stops one Raptor RUD cascading into, well, more. Raptor 3 is supposed to solve this problem, I guess by being more reliable and robust? Anyway, if they don't solve the problem, then providing shielding for 33 or 35 engines will continue to be a substantial payload penalty. At some point 5 big engines look easier to shield than 35 small ones. When it comes to human payloads eliminating the shielding gets even harder, NASA is a bit of a stickler that way....
14:52 "The acceleration due to gravity goes up during the flight of the stage." Surely the acceleration due to gravity remains fairly constant at low altitudes and can only decrease as you get farther from the Earth? Would the correct wording be, "The acceleration of the rocket goes up during flight because the rocket constantly loses mass. (F = ma ⇒ a = F/m) This will increase the pressure on the tank." Edit: Looking at the equation right now, and I'm not sure how my explanation would work...
There is not that much advantage from increasing the width as height of the rocket is limited by the thrust by surface of the engines. Next rocket is more likely to be 27 meter wide or something, unless some breakthrough with rocket engines appears or we get some super materials like carbon nanotubes, metallic glass or hot temperature superconductors. Also, the 27 meter wide rocket is likely to be the biggest conventional rocket, as by that time we should have large moon and asteroid resource mining and manufacturing, so biggest Earth exports are likely to be complex things like computer chips and humans.
Asteroid mining would be better done from a base on Mars, lower gravity and much closer to the asteroid belt. Then refine the raw materials on mars and only transport highly refined materials back to earth.
@@schrodingerscat1863 I know it's hard to believe, but asteroid mining is not a thing economically. You can't mine resources from moon, mars or asteroid belt and make it cheaper than what you get on Earth, the math just does not check out. Asteroid mining will basically be exclusive to only space related activities, like making space habitats, spaceships and other space related stuff. Unless price of earth materials drastically raises, like by orders of magnitude, and we figure out good mass drivers, it's not going to happen. This is why I specifically said that Earth made rockets will have to compete with Space made rockets, as for the same reason why it's expensive to send stuff back to earth, it's expensive to send stuff out of earth, so we probably will not be delivering THAT much cargo to orbit with gigantic rockets, we are more likely to just deliver humans and chips, and then spaceships and fuel for spaceships will be made in space.
@@Ormusn2o May not currently be a thing but that will change as technology makes such a venture economically viable. May be 100 years away but it will happen, the resources available in the asteroid belt are vast.
Unless they begin constructing interplanetary ships in space There is a limit to the weight that existing rocket engines can lift into orbit without constantly destroying the launch pad!
1:37 as a layperson, this seems to be a pretty big assumption to make. It would have been nice if you'd added two or three sentences to justify its validity. For example, what are the typical pressures in those tanks? What are the structural loads?
since the raptor engines are not operating at full thrust and they don't need to re-light all the engines for landing either stage, it makes sense that you'd use up more of that untapped energy margin
Yes, just not width wise for a LONG ass time, it will go in length over time, just as F9 did, reason being is that all the launch infrastructure would need changing to accommodate a wider starship a longer starship is fine though, and we already know a longer startship is coming, but that's V2 I will assuming there will be a V3 and maybe even or 4 and 5 as the engines improve their thrust.
Take 2 boosters 9 meter diameter each side by side and one starship on top with 9*1.41 meter diameter or 12.6 meters. Then eliminate the need for the displacement ring by firing 2 starship raptors that are not covered by the boosters during separation. Or take 3 boosters bundled together with one super sized starship on top. If the boosters are already available, then why not bundle them for bigger payloads?
@nonegone7170The cross section of two super heavy boosters is two circles shaped like an eight. The cross section foot print of the 2nd stage on top in the center is a circle resting on both the super heavy boosters. The top of the boosters would require a structural adaptation to transfer half the weight of the 2nd stage along 180 degrees of its circular footprint to each super heavy booster. Apart from this transition interface between the first and second stages, all structural loads are vertical (same as is now). And since some of the second stage Raptors are outside the figure eight cross section, their rocket exhaust does not hit the boosters. This eliminates the structural impact of the raptor exhaust hitting the booster.
It would be much simpler to manufacture the Super Heavy booster from an Al-Li alloy and thereby increase the system's payload capacity for the following reasons: 1. SpaceX has decades of experience with this alloy, 2. most of the booster does not require any heat resistance, 3. the booster does not need to be mass-produced due to the extremely short turnaround time.
I would agree, but I think the reasons are that #1: Steel is much stronger than any aluminum alloy, meaning less material is used. #2: It is more easily repaired. #3 The spacecraft can be entirely made out of Steel whereas an aluminum craft has to have carbon fiber components which are much more expensive.
Starship is presently a one size fits all design, but Starship has a number of different roles which have different design requirements. Firstly, as a tanker and an orbital refuelling station, the bigger the better resulting in fewer launches and lower boil off of propellant while waiting. As a lunar transfer vehicle, it is much bigger than the Orion capsule so flexibility in terms of cargo and crew number. If it were just to move crew to lunar orbit then bring them back to Earth orbit, then the present starship with further development of the heat shield would probably the way to go. Cargo to the lunar surface initially would be one way with maximum mass to fuel ratio - just enough to land. The vehicle could be reused as habitation and/or construction material. A lunar lander needs to carry less cargo and more fuel to return to orbit. It could be a stripped down standard Starship. A lot depends on in situ fuel production. For Mars again bigger tankers in orbit, but landers should be more specialise to utilise the thinner atmosphere- perhaps a more elliptical cross section to maximise aerobraking. An interesting decade is guaranteed.
The starship can technically carry enough fuel to not need fuel production on the moon. It can get there from low Earth orbit and back on a full tank. Without in-situ fuel production on the moon, you will just be more limited in how much material you can bring back.
You did not touch on wind resistance of a much larger rocket. I have a design for a space station that requires some 60 foot diameter tanks in space and I don't want to build them in space so that means I need an 18 meter diameter rocket to launch them. One launch would put as many as 3 such tanks in orbit with "some assembly" on the third one. But the wind resistance is going to be intense so I think maybe there may be a practical limitation penalty for going larger in diameter in our thick atmosphere. If it could be launched from a two mile high mountain that would save a lot of fuel and probably negate the wind resistance issue as it would be miles higher before picking up speed.
To get into orbit, you need to generate sufficient velocity, deal with gravity losses, add potential energy to get to a higher location, and deal with aerodynamic drag. I think those are in order of magnitude from largest to smallest. Last time I looked, aerodynamic drag is only a couple hundred meters per second of delta v. .
So, out of curiosity, how would an estimate on internal volume of a 12 meter crewed starship compare to the ISS? That aside, while I don't disagree that the 9m version is likely to be the standard for some time, I would at the same time be not overly surprised if a 12 meter were to come along down the line as the next step, as it were. It likely would have to wait until after establishment of market demands driven by 9m Starship, but still, it's within the industrial capacity to implement, unlike for gains for 15m+ rockets of sufficient length to be necessary. As a different aside, how practical / feasible / likely do you think it would be for us to see in the future a space station that is just 4 modified crew starships docked to a single central hub module? And how would it compare to current / historical / likely other private space station programs going on right now?
My guess is that they will go bigger than 12m - 15m or 18m - and with that I think they are going to need a super raptor. They would (roughly) need 4x the thrust to go with 30 ish engines on an 18m vehicle.
How does aerodynamic drag play into tank diameters? Wouldn't you always want to stay a little on the skinny side to keep aero drag lower? Or is it a small component of the overall performance?
Nice work, busy busy I see. One thing you didn’t mention is thrust density. It is easier to stuff more engines under a bigger diameter rocket, and they are already struggling to meet their thrust goals for the new Starship versions - which you nicely summarized in your video covering the importance of Raptor. At some point they won’t be able to squeeze more performance out of Raptor (safely or reliably), at which point an increase in diameter is required for any further increase in thrust (=payload). >90% of Starship launches are fuel tankers, so the real analogy is the KC-135. For Elons Mars program, >99% of launches from earth are fuel trips to fill cargo Starships bound for Mars. Starship is almost entirely a fuel tanker, and there is no limit to the need for more payload. If Musk is really serious about Mars, and I believe he is, then an increase in diameter is inevitable, at least as soon as they start generating some revenue via Starlink (presumably?). It will be interesting to see where the revenue comes from to support the Mars program……???
Currently SpaceX is targeting superheavys with 35 engines instead of 33 while maintaining the same diameter, there are already unofficial designs showing that it is possible, so they still have room to improve the current design
@@APMI-OFICIAL Yeah, but if they are increasing they chamber thrust they should be able to increase the diameter at sea level, so they're probably leaving some sea level performance on the table. It's pretty early in the program to have run out of ability to stretch it any further.
@@marksinclair701 They will max out the raptor and that will likely end development of starship. Another larger version might be built later but likely not for a decade or more.
Before Musk started talking about orbital fuel depots I had always believed SpaceX would ultimately build a number of Booster/Starships that were twice as wide and twice as tall to facilitate refueling Moon and Mars bound Starships with one launch. Doubling both diameter and height would allow roughly 8 times the payload delivery to LEO compared to the normal size.
Say existing booster has 3400 mT prop and ss has 1200 mT prop, that's 86 m of combined tank wall. Doubling tank diameter and height will require thicker steel, approx additional 27 mT in beefier bulkheads and 255.5 mT in bigger & thicker tank walls (and I'm not adjusting the non-tank masses.) Your dry mass is now at least 115% heavier than before. The net dV requirements aren't changing, but drag losses will be much higher, so prop fraction will be worse. At best you're getting a 3.75x increase in delivered payload for 8x increase in prop. The rocket equation is a stinker. Loading more raptors on the bigger cross-section won't improve the prop fraction and will greatly increase the dry mass.
Regarding raptor mass, 18m diameter shell can fit 3-6 sea level raptors x 1.5mT and ~36 vac raptors x 1.75mT, so the additional 33 vac raptors will add ~58mT to dry mass. I think I jumped from considering the whole system to just booster mass when I said 115% additional. If dry booster = 250 and dry ship = 150, the old dry mass is 400. Add 27 from bulkheads, 255.5 from tank walls, 58 from extra vac raptors gives 85% extra dry mass., which should give 4.3x payload for 8x prop.
9:45 your charts for Starship look very much as if you calculate the full LOX and LCH4 tank surface areas separately and independently, then add the two together. In doing so, you are ignoring the fact that the Common Dome is shared between the two tanks in Starship (on both stages). So you're double-counting the common dome area, and therefore all your subsequent conclusions are strongly - and wrongly - biased toward higher fineness ratios (i.e. you incorrectly favor the thin-and-tall rocket body over a short-and-fat one).
There is an error in the reasoning. The thickness can't be assumed to be the same. It has to become thicker for the larger diameter tank (all other things being equal, such as the maximum pressure difference it can withstand).
I talk about that in the video. The question is whether the skin size is driven by the needs of it as a pressure vessel or the loads that it carries as the main structural member of the rocket holding either the you're stage or the payload.
Aerodynamic drag is not a big factor in rockets because they move slowly in the lower atmosphere and spend most of their time out of the atmosphere. In numerical terms, it takes about 9400 meters per second of Delta v to get into orbit, and drag is perhaps 200 of that.
I've heard that the crew area of Starship is bigger than the ISS. How much bigger would the crew area of the largest Starship variant be? Would it make much of a difference when sending colonists to Mars?
Great presentation. One thing to note is that while pressure vessels would keep a constant mass ratio, things like heat shielding or insulation benefit massively from the square cube law. This would indicate that there is a break-even point for hydrolox where the performance outweighs the added insulation and handling issues (see extreme 1960's fully reusable ssto proposals). speaking of which, why aren't there huge expander cycle engines using multiple chambers (like the russian engines do) to overcome the power/ surface area limitation?
The pumps in the expander cycle discharge at 1.5x to 2x the chamber pressure. The tubes in the thrust chamber and nozzle walls have to hold at least this pressure so that makes them heavier, thus increasing the regenerative surface area increases the weight more than in other types of engines. The extra weight claws back some of the higher performance of the expander cycle. Fabrication is also quite involved, specially shaped tubes and silver alloy filler are manually fitted and pieced together, then it's fused together in an oven. Not every defect can be repaired, which can scrap the whole chamber. Needing multiple chambers for each engine would make them even more expensive
SpaceX will take whatever gains they can get at this point. It's not like they can keep leave efficiency on the floor. This entire endeavor rides on the edges of what is physically possible.
Back when I looked at RP-1 I tried to pick a decent spot for viscosity. You can get it a little colder but you get close to gel, and since RP-1 is a standard rather than a specific mix my guess is that you would want a little margin.
Elon announced the 18m version is coming 2034. IMO, they'll probably start upgrading facilities in 5 years. I don't think the plane comparison makes sense without competition. Starlink and direct to cellular service is expanding rapidly. Landing a high fineness Starship on uneven soft regolith and Martian winds seems like a massive pita.
Martian winds are insignificant, because the force exerted by wind is a linear function of air density - and Mars atmospheric density is about 1% that of Earth. Meaning, same-speed winds on Mars impart only 1/100th the push they do on Earth.
@@Spherical_Cow but it's made of CO2 instead of nitrogen, so it carries much more energy than the density would suggest. See Scott Manley's video where he calculated wind power is about ¼th as effective on Mars compared to earth.
@@ch4.hayabusa density is by definition a quantity of mass per unit volume. When we say Mars' air density is 1% of Earth's, we literally mean a liter of Martian atmosphere has 1% the mass compared to a liter of Earth atmosphere at the same altitude (roughly speaking; Mars has a larger scale height but that doesn't affect the discussion of Martian wind strength per se). The specifics of atmospheric composition don't matter in this context: the force exerted by wind is purely a function of speed and density.
5:35 “Falcon 9 diameter is limited by the roads” So fly it. Not in an aircraft, it can fly itself. With just an aerodynamic fairing, effectively no cargo mass, it ought to manage a few thousand kms. Point to point in 20-25 minutes.
It wasn't clear whether that would work when they were designing the rocket and it would take a *lot* of work to allow F9 first stage flight over populated areas.
Honestlly there must be a practical limit when the advantages of reusability face the lack of misions and the problems of catching a even bigger starship and booster. So i wouldn't expect starship to get much bigger than maybe 14m. Maybe at some poin a fully disposable bigger vesion might be made for things like sending up a gravity drum of a bigger weeb.
Just take a look at a can of soup. It too is a cylinder designed to minimize surface area for a given volume. If minimizing surface area was really the main issue then you would choose a spherical shape. Spheres don’t have great aerodynamics, so getting them out of Earth’s atmosphere is a drag. This is where long pointy rockets do better. How to optimize between these factors and manufacturing limitations is kinda complicated.
Eager Space been cooking this week
Cooking with liquid gas!
@@PetesGuideuhhh
Maybe meth...
@@rexmann1984 ...ane
Yessir
I don't think we'll see bigger diameter Starships until at the earliest 2030. There's still a ton of work for the 9m starship to do, and it's already comedically oversized for the vast majority of payloads.
Great analysis as always!
That's what I don't really understand why Blue Origin or ULA are going for bigger fairings. The largest satellite ever sent into orbit was with a regular Falcon 9 fairing. The issue isn't size, it's cost. And they're going backwards from there.
Starship I completely understand the fairing size. They're their own customer. Starlink V2 (and possibly V3 satellites. Something Musk briefly mentioned during the Everyday Astronaut tour) are massive and heavy. So they need the extra size for them. But for the rest, I don't really think a bigger fairing is all that important tbh
@snakevenom4954 makes engineering a payload a lot simpler, and often cheaper, when you reduce the mass/volume constraints.
@@snakevenom4954 I will re-review the video again, but I thought the discourse there indicated that Vulcan Centaur had a decent diameter, given the fuels and materials used, at 5.4 meters?
@@michaeldunne338 Hydrogen preferred a smaller diameter but they kept the same diameter for a little less performance
@@citizenblue I agree. But if the largest satellite fit into a Falcon 9 fairing, how many satellites will need a larger fairing? I'm sure some will. But enough to be a market? Not likely
Entering my 3rd year as an Aerospace Engineering student and I’ve got to say, I love your videos. These videos are incredibly informative and well thought out. They cleared up some confusions I’ve had and gave me a new perspective of spacecraft elements. Thanks ES
This plan is very similar to how they developed Falcon 9, it took them several years and iterations before they got to the Block V booster.
Dunno, the increase in diameter and height dictates a completely new structure rather than just putting a plug in an existing structure to make it longer, metal structures being a bit less limited to the moulding needed to make a composite structure like falcon.
Falcon 9 just grew in length.
Growing in diameter is a completely different matter. And Starship / Super Heavy won't change the diameter either, as far too many very expensive things have to change.
Nice video! Quick addition: the mean height of a rocket is also limited by its thrust density. One of the reasons to taper a rocket like the Saturn V is to fit more engines underneath while still getting sensible aspect ratios in the tanks above it. Starship gets to have as high an aspect ratio as it does in significant part because of how unusually powerful its engines are and how packed its flamey end is with them.
Yes, this is an often overlooked factor.
Short rockets don't need to pack the engines in very tightly. Electron has lots of space between engine bells, and the bells only take up a small percentage of the area. F9 is quite closely packed, and Super heavy is just insane with engines jammed in so tightly that most can't gimbal.
So an 18m Starship full stack isn't going to ever be twice as tall as the 9m. It just physically can't be.
Another reason for the steps in the Saturn V was the design process. The Third stage was a reused Saturn 1 / 1B Second stage & thus was designed first. The First stage was designed second; because without it the Saturn V (originally C4) was not going anywhere. This left the Second stage to be designed last. This gave them all sorts of problems, not only did they have to fit between the other two; but weight overruns were left to them to sort out. The Second stage thus had the most advanced engineering solutions of all stages. Such as the common bulkhead between the hydrogen Lox tanks.
Flamey end lol
Eager Space is the Excel version of Perun with powerpoint.
Absolutely outstanding stuff
Why is this channel so underrated? Some of the best space content out there
You forgot to talk about the heat shield! The heat shield area increases in a linear fashion to an increase in diameter. Also, a starship of greater diameter may actually slow down more quickly in the atmosphere because there’s a greater area for the atmosphere to work on. so, increasing the diameter of the starship may not only increase the amount of propellant that can be carried, but also have some real increases in the effectiveness and efficiency of the heat shield. Since the heat is getting heavier at this point because of the addition of the ablative layer, Looking for ways that the heat shield will become a smaller fraction of a total weight of the upper stage could be a real advantage.
Definitely true.
I was going to do that originally but I thought there was already too many graphs and too much mass.
@@EagerSpace
Nonetheless, a nice video. I wonder what you think of the booster catch attempt for flight five? My question is what does the last second abort look like? If SpaceX loses control of the booster (which is never landed completely successfully, even on flight four) Where will the booster crash if it’s already close to the tower? I’m puzzled why they’re rushing this because I don’t see her upside for it, but I do see a lot of downsides.
If you add that question to my "ask me a question" video, I'll probably be able to answer it there in detail.
I personally don't see much downside as long as they think they can avoid dropping it on the propellant tanks. For abort they could probably drop it off in another direction, but we've seen very few issues with running raptors.
@@EagerSpace That, and the fact SpaceX in general has their software in tiptop condition. To date I don't know of any software error ever causing major problems. In sharp contrast to that other aerospace company who's name starts with a B and charges 5 times more.
@@wbwarren57Also the heatshield has to work less because a bigger vehicle create a shockwave that is further away from the vehicle itself, it's the reason why capsules are shaped the way they are on the bottom.
Something else to consider with Starship is the reentry cross section. A shorter but wider Starship will have a different belly cross section than a tall and thin Starship. Although a larger cross section means means more surface area that needs tiles, it also means a gentler reentry. That too has some optimization equation, but it's probably more complicated due to heating limits and such.
*No need for tiles at all. just drill lots of micro holes. then pump out dry ice out of those holes to form a cold co2 insulating boundary layer. you dont even really need a pump. the heat of re-entry will cause melting of the dry ice and high pressure dry ice co2 to come out of the micro holes to form the insulating boundary layer.*
@@esecallumwhat lmao
@@furriesinouterspaceUnited learn to read and listen
@@esecallum Uh you spatted some absolute nonsense, none of that will work.
@@furriesinouterspaceUnited i proved it it lab with bunsen burner-e and an air compressor. you are regurgitaing it cant be done
your recent videos have blown up in views and you deserve it. Keep up the amazing videos!
Thanks. I have made some changes but I think it started once I got to 5000 subs.
Advantage of a 18 meter diameter starship is it would be possible to send up a 18 meter wide telescope mirror.
There is an advantage a bigger rocket gains on reentry too: A bigger radius object pushes the boundary layer further away from itself on reentry, therefore the heating is reduced. So you can get away with a thinner, weaker, and thus lighter heatshield than a smaller diameter vehicle.
Interesting. Do you know if any references I can read?
@@EagerSpace Scott Manley has a great video on it: ua-cam.com/video/hLHo9ZM3Bis/v-deo.html
NASA also published this presentation on reentry thermodynamics, that talks a little about the stagnation point, and how the radius is important, but it discusses a bunch of other stuff too: tfaws.nasa.gov/TFAWS12/Proceedings/Aerothermodynamics%20Course.pdf
Thanks for the paper. Scott does great videos but I don't watch them if its a topic I might want to cover - which is most of his videos - because I don't want to be derivative.
re-watching your more recent videos over and over again, keep up the great work please!!!
Glad to see those chapters in the description + increased video posting cadance + Hot sauce Topics 😉
This is... Surprisingly simple to be honest. I expected to understand nothing but I was easily able to follow every step
Fantastic video. I was thinking you would not talk about pressure, but you did. The math is simple and enough for people to start thinking instead of just talking.
There are some difficulties with rockets over 10 meters in diameter, having to do with airflow over the frontal surface area. Lengthening the rocket is much more efficient than increasing the diameter.
I LOVE YOUR SHOW. NO BS, very interesting stuff.
9:34 Please get carried away like this more often.
you never miss
Wait for my next video, I think you might change your mind.
Thank you for making the airliner class analogy.
I would have mentioned the issue with thrust density with regard to height. Unless SpaceX comes up with continually better engines, Starship can get wider, but it can’t get taller. You no longer have room for enough engines at the bottom.
150 meters is pretty much the max you can get to with current engine technology (Raptor V3). (Assuming same diameter for the whole rocket.)
BTW, this means that if you don’t have really good engines, you are limited to a lower height. It is an important design constraint. As long as New Glenn uses the ORSC BE-4 engine, you’re probably not going to see a version reach much beyond 100 meters, as an example. This could push designers to increase from a 7 meter diameter to 9 or 12 meters, or beyond. If the need arises for a larger more capable rocket.
Agreed. I think a larger diameter will come as soon as they start generating revenue from operations (Starlink, etc...?). Starship is mostly a fuel tanker and the demand for payload is limitless. It's a KC135 refueling an A380.
@@SpaceAdvocate Actually they wouldn't be able to make an 150m design with raptor v3 and will have to go for a higher thrust in future iterations. They also seem to have those issues with fuel leaks and if they would be able to build an engine that can run at 300T of power they would be able to take Starship height up to 150m.
(Also curious about raptor LEET design, which might be able to get up to 330T and allow them to increase the height of Starship a bit more than what they're currently planning)
@@RandomPerson-V They are planning to make a 150 meter tall version with Raptor V3. It will have 10,000 tons of thrust, using 35 Raptor V3 engines (~285 tons of thrust per engine).
It would be able to reach 150 meters quite comfortably. The liftoff mass would for the 150 meter version be around 7000 tons, so the T/W would be around 1.43. That's quite good.
Even using Raptor 2 engines, the 150 meter version would be able to lift off, at a T/W of around 1.15. Or about the same as the Saturn V.
Make the booster a cone. You can widen it at the bottom to increase thrust. It would minimise the changes to the ship, while allowing for a more powerful booster capable of throwing the ship further.
Dude this was seriously awesome
reduced surface area also reduces heat gain, ie. boil-off.
MOAR Eager Space! Yesss!! 😊
I do wonder what challenges SpaceX would face if they had to retool factories for a wider starship
The factory would be easy enough I think, most of the stuff in there i suspect is just jigs and such which would be easy to retool. But the bays would be a pain since they are designed for the current diameter.
Bays? 🤔
Glad i found this channel
It's great. No fluff. Just dives right in to the good stuff.
Terran Space Academy is also great for the same reasons.
@@citizenblue love channels that do that
I LOVE this channel!
I wonder if the ship necessarily needs to be the same size as the booster since they use hot staging now.
great discussion good job
Why not both? 9 meters for payload is absolutely huge relative to anything we’d want to put up in the near future (say… 20 years out?). I could see a scenario in which the 9m starships handle the actual payloads, while the absolutely gigantic 12m starships are just used for refueling. Eventually, the 12m rockets could be phased in for payloads after that.
WOW. Mind blown with the what is starship comparison to CRJ, 737/320, and 380. Absolutely nailed that question
I never heard of your UA-cam channel but it popped up in my suggestions.
I must say, you do an excellent job. Thank you.
One tthing you didn't mention was he increase in the number of rocket engines goes up as diameter increases. Maybe the mass and thrust balances out, but it would be nice to hear you say so.
Thanks.
You do get more engines because the area for engines increases with the square of the radius.
As you increase the diameter, the area and mass both (approximately) increase with the square of the radius, so it balances out.
You can’t increase the height, though. That increases mass to the area for engines.
Elon recently twitted that wider version of starship will be developed at some point
The moment i saw that i immediately remembered that i watched this came back to leave a comment and feed the algorithm
Ok, you concluded the most important factor is choosing rocket diameter based the amount of propellant you want. There is a strong reason to go bigger NOW not in 10-20 years. They want to bring as much propellants as possible in the tanker ships, to reduce the amount of tanker flights they need to refill for Mars and Moon trajectories. I think they should build 12m tankers, to refill 9m cargo and crew Starships. Some are claiming the imagined 9m version of a tanker would take 12-17 flights. If is that bad, they might want to go to 12m tankers asap. (for efficiency rather than to shut up the skeptics).
I don't think number of tanker flights is something they care about directly.
They do care about how much refueling will cost and they care about the hassle of supporting tanker flights if the numbers per refueling get big enough to be operationally problematic, but you are saying that they should throw - let's pick a number - a few billion dollars towards a 12 meter variant just to reduce the number of tanker flights.
It's not clear that at 12 meter new version is a win from a cost perspective.
One other constrain with diameter is surface of bottom where you put your engines. At some point going bigger is just getting wider if you want to keep the TWR
Excellent work. It should be noted that the maximum height of a rocket is set by the thrust per unit area of the base of the rocket. And thrust per unit area of the base will be constant with a given engine type and engine packing density.
How does that determine height? Mass I can see...
@TheEvilmooseofdoom Yeah I wasn't very clear. How about this...
Making a cylindrical rocket wider wont give it any advantage in being able to stack propellant higher.
@@TheEvilmooseofdoom You can’t add height without adding mass. If you’re only increasing the height, you can’t add more engines, and the thrust to weight will drop.
Great video!
Something you neglected to mention that might be a significant consideration: wouldn't air resistance increase very significantly with gains in width?
Air resistance is not a big energy loss when launching to orbit. Rockets are slow in the lower atmosphere and do most of their work in vacuum.
The assumption that the tank is the same thickness is simply wrong. If the tank is twice the diameter, and designed to hold to the same pressure, then the walls need to be twice as thick. Also, rockets have height limits if all the engines ar on the bottom. There's a limit to how many engines you can bit under each stage. So it's impractical to build cylindrical rockets much bigger than starship. This is why the N1 was conical despite being even smaller than starship.
He talks about this at 13:50
The N1 was conical mostly because they were using spherical propellant tanks. That's the ratio between Lox & Kerosene.
Is the pressure driving the thickness or is it the structural load?
@chrissouthgate4554 Fair point. A better example would be the flared bottom of the Saturn V, to account for the massive F1 engines. No matter how you slice it, you will reach a point where there isn't enough space on the bottom of the rocket to leverage enough thrust to get the rocket off the ground.
A single raptor engine can only lift the weight of a column of water with its own footprint 200m tall. Since you can't fill 100% of the area under the rocket with engines and you need a thrust to weight ratio of at least 1.5:1 to efficiently reach orbit, the 150m tall starship V3 will be pushing the limits of how tall a vehicle can be and still fly.
@EagerSpace The rocket derives its rigidity from the internal pressure. The taller the rocket, the higher the necessary pressure to maintain rigidity. Basically, you want the walls to always be under tensile load and never have a significant compressive load.
This was fascinating, thanks for helping me understand that.
I noticed that SpaceX’s original Mars vehicle had a 12m diameter and a substantially higher payload capacity. But then they shrank the vehicle to 9m and now it seems like they are stretching it instead of making it wider.
I wonder if they will return to 12m for the Mars colonization rockets. I’m curious what impact the number of refueling rockets needed for those deeper space missions.
Thickness is a function of diameter but not height for internal pressure. Vertical thrust loads are a buckling problem that is dependent on both diameter and height. So both types of load impact the thickness. You can add stiffeners for the thrust loads, which SpaceX already does but in either case of larger diameter or larger height will like require more stiffeners.
This is definitely a case where the simplifications I made detract from getting a real answer.
Hell yeah! Can't wait to watch this
Thank you for elevating my entire approach to rocket design. Your approach to thoughtful analysis revolutionized my understanding of the market and state of technology.
You changed my world
Banger video
As of a day or two ago, the vertical tank farm has been completely cut down and harvested. Do you have access to newer photos? RGV might be willing to share.
10:43 nice Princess Bride reference!
It originally said something like "large rockets", and I am so pleased I came up with "rockets of unusual size"
@@EagerSpace You even managed to keep the same acronym! R.O.U.S.
Ideally, we want to get to the point of building starships in space. Size will not matter then!!
Rockets of Unusual Size? I don't think they exist.
Thanks for this video
Imagine there will be alotta variations for different purposes.
Would personnel love my own mini version for personal trips to distant moons.
There is a function in the grapth plot to make the grapth smooth, so the ridges from the calculation is not vissible
Are you saying that I could have smoothed them?
I prefer not to do that.
@EagerSpace well i guess. Its a matter of tast.
This IS a very interesting topic. Another topic I've been curious about is the issue of how many engines is most beneficial. For example, Superheavy can loose an engine or two and still boost Starship enough for it to make orbit. There is a new design with 35 raptors. What is the benefit of more engines? Obviously more thrust, at the cost a little more weight per engine, but that gets the rocket up higher faster, so I assume that cuts into gravity losses? And Starship has an upcoming design revision where it will have nine engines rather than six. And it will hold more fuel. What is the cost of the additional engines and what is the benefit? I imagine there are some interesting graphs to be drawn re. those figures.
You are correct - the faster you burn the fuel, the quicker you can stage. That not only reduces gravity losses but for the first stage, means you are closer to the launch site so it takes a little bit less fuel to get back.
Is it possible to mitigate the noise for a wider diameter rocket like this?
I did comment that I believed it would take SpaceX a long time to go for 12m never mind 18m when this video came out, but assuming SpaceX is serious about going to Mars which I have my doubts, and really even going to the moon on a semi regular basis a 18m Starship makes a bunch of sense and not even bothering with a 12m Starship because as I mentioned the launch infrastructure would have to be built new.
However what an 18m Starship does for you is enables you to lift 600T to 800T in one launch reducing the number of launches required to do refuelling in LEO using to 3 - 4 launches from like 12 to 20 when you account for boil off, which getting it done in less launches also reduces how much you lose to boil off, this assumes you are using a 9m V3 for getting to the Moon or Mars.
So you would have both being produced simultaneously although you would only ever need 3x 18m Starships for refuelling, maybe a couple of extra so that your not just using them for refuelling because it would be a bit of a waste to build all the infrastructure just for a fuel truck just encase anyone want to launch fricking 600T - 800T in 1 go.
Also with a 18m Starship you can eventually just leave one in orbit as a fuel depo with the power generation and equipment to do the cooling required to stop boil off being an issue so that it's viable to do an 18m Starship to Mars, again it would require 12 launches assuming 800T of cargo, but only 12 launches because you don't have to worry about boil off between launches.
Well that or either you would launch a depo into orbit, because you would want to outfit a depo with insolation to reduce boil off, protection from micrometeoroids as it's going to be getting left in space for a long time, which retrofitting that in space would be problematic, and building a starship on Earth capable of that would almost certainly be more trouble than building modules and launching those modules, because those models don't also have to be built to be a rocket so it reduces the complexity and therefore the cost and you do get some economy of scale because you would be doing 14 - 16 modules to the cost of the latter half would cost less than the first half.
No 36m Starship pleas, that is getting truly ridiculous, 18m is already like launching a fairly skinny high-rise building into space.
I imagine a century from now, starship will be remembered as the Ford Modle T of early space flight.
More like the first deep water sailing ships, like the caravel or the galleon.
I guess more like the Ford Edsel.
Another complicating factor is reentry heat load. Smaller vehicles are better suited for reentry because they exhibit faster deceleration and can slow down more at higher altitude.
That’s inaccurate. You want a low mass relative to the aerodynamic cross section. That’s possible to achieve on both smaller and larger vehicles. For crewed vehicles, I would think it’s easier to accomplish on larger vehicles, as the volume needed for the crew is easier to work around.
@SpaceAdvocate Incorrect. The larger the vehicle, the thicker the walls need to be, the greater the mass to cross-sectional area.
Is there ever a time when payload width overpowers the need for the most efficient rocket?
Yes. When whoever (NASA) wants that payload is willing to throw in a few extra billion for the cost of development&operation.
Starship factory is at the launch site. This is an advantage when growing diameter. There is too much advantage to growing diameter. It will eventually happen. There could be a reducer for using prior generation 2nd stage Starships on a booster. I look forward to changes.
It might but at the huge expense of scrapping much of what they just built.
A larger diameter pushes the shockwave further away during re-entry. So there's that consideration too.
Another banger video, sir
It makes sense.
There's a couple of other factors to consider. Firstly fuel boil off during longer duration flights. Intuition says that a larger volume of fuel vs surface area would experience less fuel boil off than splitting the payload into two rockets, thus proportionately less payload would have to be dedicated to managing the temperature of the fuel, with proportionally smaller radiators, etc. The thicker sidewalls of a larger rocket also help shield against the radiation of deep space, at least a little.
And I'm sure the size of the rocket plays a part in reentry flight profiles and heating, although I'm nowhere near smart enough to work out whether a larger vehicle is beneficial or a bigger challenge. Perhaps here the tables are turned and the larger volume to surface area increases the heating as there's more energy to dissipate per unit area...
If Musk is still eyeing a bigger rocket I presume the tradeoffs work out positively.
Is there a possibility that before they go with a larger diameter rocket that SpaceX will go for a Starship Heavy as an interim stage and then flare up the upper stage to 12 meters?
Seeing a rocket with 105 engines take off will be spectacular.
Also going to 18 meters will Raptor be enough or do they need something the size of Rocketdyne F-1?
More raptors seems more likely than bigger ones. Scaling up rarely works linearly. Bigger engines add length too.
Clustering engines has already proven itself and has many advantages. There is no need for larger engines no matter how you scale the rocket.
Just go Mega-Soyuz with four carrot boosters and a potato masher core stage
@@mostevil1082 They haven't yet proven that they can eliminate the 10t+ engine shielding which stops one Raptor RUD cascading into, well, more. Raptor 3 is supposed to solve this problem, I guess by being more reliable and robust? Anyway, if they don't solve the problem, then providing shielding for 33 or 35 engines will continue to be a substantial payload penalty. At some point 5 big engines look easier to shield than 35 small ones. When it comes to human payloads eliminating the shielding gets even harder, NASA is a bit of a stickler that way....
14:52 "The acceleration due to gravity goes up during the flight of the stage." Surely the acceleration due to gravity remains fairly constant at low altitudes and can only decrease as you get farther from the Earth? Would the correct wording be, "The acceleration of the rocket goes up during flight because the rocket constantly loses mass. (F = ma ⇒ a = F/m) This will increase the pressure on the tank."
Edit: Looking at the equation right now, and I'm not sure how my explanation would work...
Yes, your explanation is correct; it's about the effective gravity not the actual gravity.
@@EagerSpace Ahh, "effective acceleration", there we go
There is not that much advantage from increasing the width as height of the rocket is limited by the thrust by surface of the engines. Next rocket is more likely to be 27 meter wide or something, unless some breakthrough with rocket engines appears or we get some super materials like carbon nanotubes, metallic glass or hot temperature superconductors. Also, the 27 meter wide rocket is likely to be the biggest conventional rocket, as by that time we should have large moon and asteroid resource mining and manufacturing, so biggest Earth exports are likely to be complex things like computer chips and humans.
Asteroid mining would be better done from a base on Mars, lower gravity and much closer to the asteroid belt. Then refine the raw materials on mars and only transport highly refined materials back to earth.
@@schrodingerscat1863 I know it's hard to believe, but asteroid mining is not a thing economically. You can't mine resources from moon, mars or asteroid belt and make it cheaper than what you get on Earth, the math just does not check out. Asteroid mining will basically be exclusive to only space related activities, like making space habitats, spaceships and other space related stuff. Unless price of earth materials drastically raises, like by orders of magnitude, and we figure out good mass drivers, it's not going to happen. This is why I specifically said that Earth made rockets will have to compete with Space made rockets, as for the same reason why it's expensive to send stuff back to earth, it's expensive to send stuff out of earth, so we probably will not be delivering THAT much cargo to orbit with gigantic rockets, we are more likely to just deliver humans and chips, and then spaceships and fuel for spaceships will be made in space.
@@Ormusn2o May not currently be a thing but that will change as technology makes such a venture economically viable. May be 100 years away but it will happen, the resources available in the asteroid belt are vast.
Unless they begin constructing interplanetary ships in space There is a limit to the weight that existing rocket engines can lift into orbit without constantly destroying the launch pad!
1:37 as a layperson, this seems to be a pretty big assumption to make. It would have been nice if you'd added two or three sentences to justify its validity. For example, what are the typical pressures in those tanks? What are the structural loads?
nevermind, should've watched the whole video first :'D
I did handwave a lot in the section on pressure and structural analysis but I think it's a very complex topic.
@@EagerSpace that section is pretty good and more than I expected. I simply wrote my comment before watching that far.
No worries.
since the raptor engines are not operating at full thrust and they don't need to re-light all the engines for landing either stage, it makes sense that you'd use up more of that untapped energy margin
Yes, just not width wise for a LONG ass time, it will go in length over time, just as F9 did, reason being is that all the launch infrastructure would need changing to accommodate a wider starship a longer starship is fine though, and we already know a longer startship is coming, but that's V2 I will assuming there will be a V3 and maybe even or 4 and 5 as the engines improve their thrust.
Take 2 boosters 9 meter diameter each side by side and one starship on top with 9*1.41 meter diameter or 12.6 meters. Then eliminate the need for the displacement ring by firing 2 starship raptors that are not covered by the boosters during separation. Or take 3 boosters bundled together with one super sized starship on top. If the boosters are already available, then why not bundle them for bigger payloads?
@nonegone7170The cross section of two super heavy boosters is two circles shaped like an eight. The cross section foot print of the 2nd stage on top in the center is a circle resting on both the super heavy boosters. The top of the boosters would require a structural adaptation to transfer half the weight of the 2nd stage along 180 degrees of its circular footprint to each super heavy booster. Apart from this transition interface between the first and second stages, all structural loads are vertical (same as is now). And since some of the second stage Raptors are outside the figure eight cross section, their rocket exhaust does not hit the boosters. This eliminates the structural impact of the raptor exhaust hitting the booster.
Could we get there using the Brisbane tank designs?
It would be much simpler to manufacture the Super Heavy booster from an Al-Li alloy and thereby increase the system's payload capacity for the following reasons: 1. SpaceX has decades of experience with this alloy, 2. most of the booster does not require any heat resistance, 3. the booster does not need to be mass-produced due to the extremely short turnaround time.
I would agree, but I think the reasons are that #1: Steel is much stronger than any aluminum alloy, meaning less material is used. #2: It is more easily repaired. #3 The spacecraft can be entirely made out of Steel whereas an aluminum craft has to have carbon fiber components which are much more expensive.
Starship is presently a one size fits all design, but Starship has a number of different roles which have different design requirements. Firstly, as a tanker and an orbital refuelling station, the bigger the better resulting in fewer launches and lower boil off of propellant while waiting. As a lunar transfer vehicle, it is much bigger than the Orion capsule so flexibility in terms of cargo and crew number. If it were just to move crew to lunar orbit then bring them back to Earth orbit, then the present starship with further development of the heat shield would probably the way to go. Cargo to the lunar surface initially would be one way with maximum mass to fuel ratio - just enough to land. The vehicle could be reused as habitation and/or construction material. A lunar lander needs to carry less cargo and more fuel to return to orbit. It could be a stripped down standard Starship. A lot depends on in situ fuel production.
For Mars again bigger tankers in orbit, but landers should be more specialise to utilise the thinner atmosphere- perhaps a more elliptical cross section to maximise aerobraking. An interesting decade is guaranteed.
The starship can technically carry enough fuel to not need fuel production on the moon. It can get there from low Earth orbit and back on a full tank.
Without in-situ fuel production on the moon, you will just be more limited in how much material you can bring back.
You did not touch on wind resistance of a much larger rocket. I have a design for a space station that requires some 60 foot diameter tanks in space and I don't want to build them in space so that means I need an 18 meter diameter rocket to launch them. One launch would put as many as 3 such tanks in orbit with "some assembly" on the third one. But the wind resistance is going to be intense so I think maybe there may be a practical limitation penalty for going larger in diameter in our thick atmosphere. If it could be launched from a two mile high mountain that would save a lot of fuel and probably negate the wind resistance issue as it would be miles higher before picking up speed.
To get into orbit, you need to generate sufficient velocity, deal with gravity losses, add potential energy to get to a higher location, and deal with aerodynamic drag.
I think those are in order of magnitude from largest to smallest. Last time I looked, aerodynamic drag is only a couple hundred meters per second of delta v. .
So, out of curiosity, how would an estimate on internal volume of a 12 meter crewed starship compare to the ISS?
That aside, while I don't disagree that the 9m version is likely to be the standard for some time, I would at the same time be not overly surprised if a 12 meter were to come along down the line as the next step, as it were. It likely would have to wait until after establishment of market demands driven by 9m Starship, but still, it's within the industrial capacity to implement, unlike for gains for 15m+ rockets of sufficient length to be necessary.
As a different aside, how practical / feasible / likely do you think it would be for us to see in the future a space station that is just 4 modified crew starships docked to a single central hub module? And how would it compare to current / historical / likely other private space station programs going on right now?
Can you add those questions to the "ask me a question" video I dropped recently?
And that is why they call it Rocket science.
Great info, thanks.
Eager Space you any reference material for this topic ?
Anyone have speculation of how many raptors / configurations would be on the 12 m super heavy and starship?
My guess is that they will go bigger than 12m - 15m or 18m - and with that I think they are going to need a super raptor. They would (roughly) need 4x the thrust to go with 30 ish engines on an 18m vehicle.
How does aerodynamic drag play into tank diameters? Wouldn't you always want to stay a little on the skinny side to keep aero drag lower? Or is it a small component of the overall performance?
Drag is a small part of the energy to get to orbit. Only about 1% iirc
@@EagerSpace Yeah thats very low! Gravity is a cruel mistress.
Make it big enough to fit 2-4 f-22's in it. The rapid strike capabilities of that would be profound.
Nice work, busy busy I see.
One thing you didn’t mention is thrust density. It is easier to stuff more engines under a bigger diameter rocket, and they are already struggling to meet their thrust goals for the new Starship versions - which you nicely summarized in your video covering the importance of Raptor. At some point they won’t be able to squeeze more performance out of Raptor (safely or reliably), at which point an increase in diameter is required for any further increase in thrust (=payload).
>90% of Starship launches are fuel tankers, so the real analogy is the KC-135. For Elons Mars program, >99% of launches from earth are fuel trips to fill cargo Starships bound for Mars. Starship is almost entirely a fuel tanker, and there is no limit to the need for more payload. If Musk is really serious about Mars, and I believe he is, then an increase in diameter is inevitable, at least as soon as they start generating some revenue via Starlink (presumably?). It will be interesting to see where the revenue comes from to support the Mars program……???
Currently SpaceX is targeting superheavys with 35 engines instead of 33 while maintaining the same diameter, there are already unofficial designs showing that it is possible, so they still have room to improve the current design
And starlink currently has profits of the order of 6 billion dollars
@@APMI-OFICIAL Yeah, but if they are increasing they chamber thrust they should be able to increase the diameter at sea level, so they're probably leaving some sea level performance on the table. It's pretty early in the program to have run out of ability to stretch it any further.
@@marksinclair701 They will max out the raptor and that will likely end development of starship. Another larger version might be built later but likely not for a decade or more.
Before Musk started talking about orbital fuel depots I had always believed SpaceX would ultimately build a number of Booster/Starships that were twice as wide and twice as tall to facilitate refueling Moon and Mars bound Starships with one launch. Doubling both diameter and height would allow roughly 8 times the payload delivery to LEO compared to the normal size.
Say existing booster has 3400 mT prop and ss has 1200 mT prop, that's 86 m of combined tank wall. Doubling tank diameter and height will require thicker steel, approx additional 27 mT in beefier bulkheads and 255.5 mT in bigger & thicker tank walls (and I'm not adjusting the non-tank masses.) Your dry mass is now at least 115% heavier than before. The net dV requirements aren't changing, but drag losses will be much higher, so prop fraction will be worse. At best you're getting a 3.75x increase in delivered payload for 8x increase in prop. The rocket equation is a stinker. Loading more raptors on the bigger cross-section won't improve the prop fraction and will greatly increase the dry mass.
Regarding raptor mass, 18m diameter shell can fit 3-6 sea level raptors x 1.5mT and ~36 vac raptors x 1.75mT, so the additional 33 vac raptors will add ~58mT to dry mass. I think I jumped from considering the whole system to just booster mass when I said 115% additional. If dry booster = 250 and dry ship = 150, the old dry mass is 400. Add 27 from bulkheads, 255.5 from tank walls, 58 from extra vac raptors gives 85% extra dry mass., which should give 4.3x payload for 8x prop.
9:45 your charts for Starship look very much as if you calculate the full LOX and LCH4 tank surface areas separately and independently, then add the two together. In doing so, you are ignoring the fact that the Common Dome is shared between the two tanks in Starship (on both stages). So you're double-counting the common dome area, and therefore all your subsequent conclusions are strongly - and wrongly - biased toward higher fineness ratios (i.e. you incorrectly favor the thin-and-tall rocket body over a short-and-fat one).
There is an error in the reasoning.
The thickness can't be assumed to be the same. It has to become thicker for the larger diameter tank (all other things being equal, such as the maximum pressure difference it can withstand).
I talk about that in the video. The question is whether the skin size is driven by the needs of it as a pressure vessel or the loads that it carries as the main structural member of the rocket holding either the you're stage or the payload.
Noob here so i don't really know what I am talking about.. But wouldn't a wider rocket also have significantly higher drag needing a lot more fuel?
Aerodynamic drag is not a big factor in rockets because they move slowly in the lower atmosphere and spend most of their time out of the atmosphere.
In numerical terms, it takes about 9400 meters per second of Delta v to get into orbit, and drag is perhaps 200 of that.
If it gets bigger , maybe its safer to launch the crew on a smaller rocket, snd use Spacex for habitats, etc.
I've heard that the crew area of Starship is bigger than the ISS. How much bigger would the crew area of the largest Starship variant be? Would it make much of a difference when sending colonists to Mars?
My guess is that the longer designs just have bigger tanks.
Great presentation. One thing to note is that while pressure vessels would keep a constant mass ratio, things like heat shielding or insulation benefit massively from the square cube law. This would indicate that there is a break-even point for hydrolox where the performance outweighs the added insulation and handling issues (see extreme 1960's fully reusable ssto proposals). speaking of which, why aren't there huge expander cycle engines using multiple chambers (like the russian engines do) to overcome the power/ surface area limitation?
The pumps in the expander cycle discharge at 1.5x to 2x the chamber pressure. The tubes in the thrust chamber and nozzle walls have to hold at least this pressure so that makes them heavier, thus increasing the regenerative surface area increases the weight more than in other types of engines. The extra weight claws back some of the higher performance of the expander cycle.
Fabrication is also quite involved, specially shaped tubes and silver alloy filler are manually fitted and pieced together, then it's fused together in an oven. Not every defect can be repaired, which can scrap the whole chamber. Needing multiple chambers for each engine would make them even more expensive
Radius=1/2 x Diameter.
Volume= (PI x Radius x Radius) x Height
Height = 100 meters
Diameter=9 meters. Radius = 4.5 meters
Volume = (3.14*4.5*4.5)*100 = 6,361 cu meters
Height = 150 meters
Volume = (3.14*4.5*4.5)x150 = 9,542 cu meters
Height = 100 meters
Diameter=18 meters. Radius = 9 meters
Volume = (3.14*9*9)*100 = 25,442 cu meters
Height = 150 meters
Volume = (3.14*9*9)x150 = 38,170 cu meters
2x the diameter yields 4x the volume.
The optimum height of a rocket is 10-20x the diameter.
Which starship is easier to reenter the earth 9 meters or 18?
SpaceX will take whatever gains they can get at this point. It's not like they can keep leave efficiency on the floor. This entire endeavor rides on the edges of what is physically possible.
NISTIR 6646 records viscosities for RP-1 down to -30°C, where it could reach a density of 836 kg/m^3
Back when I looked at RP-1 I tried to pick a decent spot for viscosity. You can get it a little colder but you get close to gel, and since RP-1 is a standard rather than a specific mix my guess is that you would want a little margin.
Make the booster big enough to hold 100 engines and the starship big enough to hold 30 sea level and 30 vacuum.
Oh, that picture of the fast food chicken fried burger was totally unfair at just barely 2 minutes into your video
What about drag?
Drag is generally not a big factor because rockets don't move very quickly in the lower atmosphere.
Elon announced the 18m version is coming 2034. IMO, they'll probably start upgrading facilities in 5 years.
I don't think the plane comparison makes sense without competition. Starlink and direct to cellular service is expanding rapidly. Landing a high fineness Starship on uneven soft regolith and Martian winds seems like a massive pita.
Martian winds are insignificant, because the force exerted by wind is a linear function of air density - and Mars atmospheric density is about 1% that of Earth. Meaning, same-speed winds on Mars impart only 1/100th the push they do on Earth.
@@Spherical_Cow but it's made of CO2 instead of nitrogen, so it carries much more energy than the density would suggest. See Scott Manley's video where he calculated wind power is about ¼th as effective on Mars compared to earth.
@@ch4.hayabusa density is by definition a quantity of mass per unit volume. When we say Mars' air density is 1% of Earth's, we literally mean a liter of Martian atmosphere has 1% the mass compared to a liter of Earth atmosphere at the same altitude (roughly speaking; Mars has a larger scale height but that doesn't affect the discussion of Martian wind strength per se). The specifics of atmospheric composition don't matter in this context: the force exerted by wind is purely a function of speed and density.
5:35 “Falcon 9 diameter is limited by the roads”
So fly it. Not in an aircraft, it can fly itself. With just an aerodynamic fairing, effectively no cargo mass, it ought to manage a few thousand kms. Point to point in 20-25 minutes.
It wasn't clear whether that would work when they were designing the rocket and it would take a *lot* of work to allow F9 first stage flight over populated areas.
Honestlly there must be a practical limit when the advantages of reusability face the lack of misions and the problems of catching a even bigger starship and booster. So i wouldn't expect starship to get much bigger than maybe 14m. Maybe at some poin a fully disposable bigger vesion might be made for things like sending up a gravity drum of a bigger weeb.
Just take a look at a can of soup. It too is a cylinder designed to minimize surface area for a given volume. If minimizing surface area was really the main issue then you would choose a spherical shape. Spheres don’t have great aerodynamics, so getting them out of Earth’s atmosphere is a drag. This is where long pointy rockets do better. How to optimize between these factors and manufacturing limitations is kinda complicated.