Why was the ITS downscaled?

[BringBackSuperHeavies!]

The Starship is not a true interplanetary vessel, it's more tailored for Mars. You could still send large payloads out to Europa or wherever, but the Starship would not be reused or you would instead use a kick stage. On Mars, fuel production is possible. On places such as Callisto, it might be possible, but unlike with Mars just making a bigger spaceship could be more feasible. The original ITS ship carried 1950 metric tonnes of propellant. For comparison, the Saturn V first stage was slightly more massive, at 2200 metric tonnes if I remember right. A second stage almost as heavy as the Saturn V moon rocket definitely has a lot more potential than something half that. Now, if carrying a 100 ton payload the current Starship has a delta-v of 6.9 km per sec (or was it 7.9). But, after its mission the Starship has to return to Earth (unless you want to expend it and spend a hundred million more bucks), and that consumes delta-v, so this means that actual payload capacity to destinations such as Ceres would be less. The ITS ship's size means transit times for Mars could be reduced, so less time floating in space and having your body slowly degrade. Unlike Starship, the ITS ship could go to outer solar system destinations and return, reducing costs. It might actually have payload to GEO without orbit refueling and with reuse. It might make Mars colonisation more practical than with the current Starship. It could house even larger payloads in the bay than the current Starship. Maybe a Mercury landing could be a lot more practical with this. Given all of these benefits, why downscaled it? At the 2018 IAC, Musk said that they couldnt come up with ideas to pay for ITS and the BFR could be used for more than just interplanetary travel and maybe make some money. But what is the reason ITS ship couldn't deploy satellites and do ISS operations and things like that. It's even bigger. Maybe we could have two part missions (first drop off this very big satellite in 1000km orbit, do a burn, then resupply and service the ISS, then go back home) with this kind of thing. They already paid for Starship while spending less than 5% of resources on it. Why couldn't they do the ITS project? Even given higher development cost such as more powerful Raptors. (Except: first, drop the carbon fibre. Second, the vacuum engines should gimbal, should they wish to redo the ITS)

[redneck]

Even the Starship is probably oversized for this particular step in the journey. Right now, Earth to orbit is the proper target and keeping the steps small enough to surmount fairly painlessly should be the method.   Doing ITS in the extreme scale without intermediate steps is riding for a fall.  Learn your lessons and make your mistakes on the cheap if possible and then take that knowledge to the next level.   If SpaceX had done Falcon9 without Falcon1 as a precursor, the lessons learned would have been more expensive and time consuming.   This is not a popular viewpoint.*

Also, you seem to be skipping the concept that mass ratio doesn’t change much after a certain size is reached.   The ITS that you suggest would need refueling just as Starship will to do similar missions.   

*My opinion being that a smaller precursor to Starship could have been in revenue service last year while retiring various questions in handling methane vehicles. 

[eriblo]

ITS did not have that much more performance compared to Starship - it was still powered by the same chemical propulsion. It had roughly twice the maximum payload and slightly better dv due to lighter carbon fibre construction. A dedicated expendable kicker Starship has even higher dv capability and Starship with refueling can do everything ITS could unless the payload is a dense and indivisible 300+t unit. AFAIK no such use case or payload has ever been proposed.

The ITS was downscaled so that SpaceX could afford it. Nobody else is going to pay you the equivalent of several years of global launch market revenue to develop a system sized to launch payloads 10x larger than any current ones at 10-100x the global launch rate.

SpaceX made Starship as small as it could be while still being fairly certain that it can perform its main goal - Mars settlement (without knowing exactly what that would entail). The switch to stainless steel is even more important in reducing cost of development and testing. Even so they had to become the most profitable launch company ever and invent a whole new much larger launch market (Starlink) in order to to get enough investment.

[spacenut]

I would like to have seen a 5.5m or 6m diameter smaller version with say 9-12 engines with a reusable upper stage.  This would have been a single stick fully reusable launch vehicle with a larger but similar payload capability of Falcon Heavy.  Test everything for full reusability and replace Falcon Heavy.  Then go with the 9m diameter vehicle they have now.  It would probably already be in service launching Starlinks, refueling for heavier payloads to the moon and beyond.  We might be later with Starship, but now we get Starship earlier. 

The only thing that SpaceX might have done first would have been building it in Florida to launch at the Cape with very few holdups other than weather.  They could have gotten another pad to build and launch from.  Then build at Boca Chica.  We would probably already have launched the second vehicle by now. 

[Twark_Main]

The Starship is not a true interplanetary vessel, it's more tailored for Mars. You could still send large payloads out to Europa or wherever, but the Starship would not be reused or you would instead use a kick stage.

Mars is a planet. Therefore Starship is a "true" interplanetary vessel.

There's no requirement that a "true" interplanetary vessel be able to visit every planet non-expendable configuration. That's not a thing in anyone's book. You're just making up your own definitions now. 

[InterestedEngineer]

. The original ITS ship carried 1950 metric tonnes of propellant.

Starship carries 1200t of propellant.   1950-1200 = 750t more.  At 113t/ring that's 7 more rings.

800t more of payload means 1200t more takeoff thrust, or 36t additional thrust per Raptor  That's already projected by Elon and peak-tested on the test stand.

That's easily within projected capabilities.

So not, ITS wasn't wasn't downscaled. It was merely iterated to, which is always a better way to go.

[meekGee]

Even the Starship is probably oversized for this particular step in the journey. Right now, Earth to orbit is the proper target and keeping the steps small enough to surmount fairly painlessly should be the method.   Doing ITS in the extreme scale without intermediate steps is riding for a fall.  Learn your lessons and make your mistakes on the cheap if possible and then take that knowledge to the next level.   If SpaceX had done Falcon9 without Falcon1 as a precursor, the lessons learned would have been more expensive and time consuming.   This is not a popular viewpoint.*

Also, you seem to be skipping the concept that mass ratio doesn’t change much after a certain size is reached.   The ITS that you suggest would need refueling just as Starship will to do similar missions.   

*My opinion being that a smaller precursor to Starship could have been in revenue service last year while retiring various questions in handling methane vehicles.

Yes, see how much faster other smaller rockets have gone.

SpaceX's goal is Mars.  LEO is just a staging area, and they chose (variants of) the same vehicle to get to the staging area and to go to Mars from it. Which simplified things.

You have a long argument to make if you think they should have made a smaller LEO-as-a-goal ship first.

[InterestedEngineer]

. The original ITS ship carried 1950 metric tonnes of propellant.

Starship carries 1200t of propellant.   1950-1200 = 750t more.  At 113t/ring that's 7 more rings.

800t more of payload means 1200t more takeoff thrust, or 36t additional thrust per Raptor  That's already projected by Elon and peak-tested on the test stand.

That's easily within projected capabilities.

So not, ITS wasn't wasn't downscaled. It was merely iterated to, which is always a better way to go.

Furthermore, the difference in deltaV pof the mass ratio of (125 + 200 + 1200) / (125 + 200) = 4.69 which is what we get with Raptor3 and no additional fuel rings, and (125+200+1950) / ( 125+200) = 7, is an addition of a mere 1400m/sec, so all that additional fuel buys almost nothing, and there are few destinations that small amount of deltaV makes additionally possible.

It really behooves you to solve the rocket equation before making an assertion about rocket performance.  In engineering, you show your work.  Your instincts are likely wrong especially when there's a logarithm in the equation.

Furthermore, refueling at the destination is required to reuse the final stage for any ITS, and there aren't many places besides Mars where that's possible.   Maybe Titan, but power is hard to collect that far out there, which will be needed to create oxygen.  The round trip times to Mars are already problematic for reuse, and that's the fastest round trip time to any planet (except maybe Venus?)¹

Add on the fact that refueling in a +2.5km/sec elliptical orbit gets you 2.5km/sec of deltaV.   Staging² always beats larger tanks, it's in the nature of the rocket equation.

That +2.5km/sec plus the additional Oberth boost of an periapses burn gets you solar system escape velocity.   So the existing system is already interstellar, let alone interplanetary.

1. The Moon is going to require those extra rings, there's an entire thread about this
2. refueling is effectively the same thing as staging

[deltaV]

My guess is Starship's size is a compromise between several goals. The first goal is SpaceX's Mars ambitions, which favors a launcher as big as possible to keep price per mass low. The second goal is launching mega constellations such as Starlink, which also favors a big launcher to keep price per mass low. The third goal is having Starship replace Falcon for all their commercial and government missions so they have only one launcher to pay fixed costs for, and this goal needs price per flight to be low and would be easiest with a launcher that's no bigger than needed, i.e. around 30-50 tonnes to LEO like Vulcan, Terran R, and New Glenn are. A final less important goal is replacing SLS; practically any sized vehicle could do that but having a vehicle that's more capable than SLS will help reduce the excuses SLS supporters make.

[redneck]

Even the Starship is probably oversized for this particular step in the journey. Right now, Earth to orbit is the proper target and keeping the steps small enough to surmount fairly painlessly should be the method.   Doing ITS in the extreme scale without intermediate steps is riding for a fall.  Learn your lessons and make your mistakes on the cheap if possible and then take that knowledge to the next level.   If SpaceX had done Falcon9 without Falcon1 as a precursor, the lessons learned would have been more expensive and time consuming.   This is not a popular viewpoint.*

Also, you seem to be skipping the concept that mass ratio doesn’t change much after a certain size is reached.   The ITS that you suggest would need refueling just as Starship will to do similar missions.   

*My opinion being that a smaller precursor to Starship could have been in revenue service last year while retiring various questions in handling methane vehicles.

Yes, see how much faster other smaller rockets have gone.

SpaceX's goal is Mars.  LEO is just a staging area, and they chose (variants of) the same vehicle to get to the staging area and to go to Mars from it. Which simplified things.

You have a long argument to make if you think they should have made a smaller LEO-as-a-goal ship first.

I found several threads discussing this when I went looking. It's not a long argument. Trying to do too many things at once on the largest scale ever is time consuming and risky. Other companies doing smaller ships don't have a track record of moving fast so that argument doesn't hold.

 It's almost 3 years from the start of upper stage test flights on this one. Do you think a smaller version emulating successful Falcon technique would have taken more than another year or so to bring into service? Falcon9 profile using Raptors and stainless steel with RTLS baselined. If Raptors and methane are going to be fast turnaround and inexpensive to operate on Starship, then they would be the same on a smaller unit. Daily turnarounds of ships flying Starlink2 would be immanent. Starlink2 is revenue waiting to happen. The 6 meter diameter with 9 engines suggested above should be capable of 20-25 full Starlink2 per launch.

The above would take some pressure off of the full up Starship and be working from an experience base. Environmental and other permitting problems would be far less involved. But this has all been hashed out before. As I said, not a popular opinion that SS/SH at this scale could be a mistake for the first iteration of new technologies.

[DJPledger]

Dev. cost. Plain and simple. The original ITS plan was simply unaffordable so was scaled back to what we now know as Starship to make it affordable.

[meekGee]

Even the Starship is probably oversized for this particular step in the journey. Right now, Earth to orbit is the proper target and keeping the steps small enough to surmount fairly painlessly should be the method.   Doing ITS in the extreme scale without intermediate steps is riding for a fall.  Learn your lessons and make your mistakes on the cheap if possible and then take that knowledge to the next level.   If SpaceX had done Falcon9 without Falcon1 as a precursor, the lessons learned would have been more expensive and time consuming.   This is not a popular viewpoint.*

Also, you seem to be skipping the concept that mass ratio doesn’t change much after a certain size is reached.   The ITS that you suggest would need refueling just as Starship will to do similar missions.   

*My opinion being that a smaller precursor to Starship could have been in revenue service last year while retiring various questions in handling methane vehicles.

Yes, see how much faster other smaller rockets have gone.

SpaceX's goal is Mars.  LEO is just a staging area, and they chose (variants of) the same vehicle to get to the staging area and to go to Mars from it. Which simplified things.

You have a long argument to make if you think they should have made a smaller LEO-as-a-goal ship first.

I found several threads discussing this when I went looking. It's not a long argument. Trying to do too many things at once on the largest scale ever is time consuming and risky. Other companies doing smaller ships don't have a track record of moving fast so that argument doesn't hold.

 It's almost 3 years from the start of upper stage test flights on this one. Do you think a smaller version emulating successful Falcon technique would have taken more than another year or so to bring into service? Falcon9 profile using Raptors and stainless steel with RTLS baselined. If Raptors and methane are going to be fast turnaround and inexpensive to operate on Starship, then they would be the same on a smaller unit. Daily turnarounds of ships flying Starlink2 would be immanent. Starlink2 is revenue waiting to happen. The 6 meter diameter with 9 engines suggested above should be capable of 20-25 full Starlink2 per launch.

The above would take some pressure off of the full up Starship and be working from an experience base. Environmental and other permitting problems would be far less involved. But this has all been hashed out before. As I said, not a popular opinion that SS/SH at this scale could be a mistake for the first iteration of new technologies.

Even the Starship is probably oversized for this particular step in the journey. Right now, Earth to orbit is the proper target and keeping the steps small enough to surmount fairly painlessly should be the method.   Doing ITS in the extreme scale without intermediate steps is riding for a fall.  Learn your lessons and make your mistakes on the cheap if possible and then take that knowledge to the next level.   If SpaceX had done Falcon9 without Falcon1 as a precursor, the lessons learned would have been more expensive and time consuming.   This is not a popular viewpoint.*

Also, you seem to be skipping the concept that mass ratio doesn’t change much after a certain size is reached.   The ITS that you suggest would need refueling just as Starship will to do similar missions.   

*My opinion being that a smaller precursor to Starship could have been in revenue service last year while retiring various questions in handling methane vehicles.

Yes, see how much faster other smaller rockets have gone.

SpaceX's goal is Mars.  LEO is just a staging area, and they chose (variants of) the same vehicle to get to the staging area and to go to Mars from it. Which simplified things.

You have a long argument to make if you think they should have made a smaller LEO-as-a-goal ship first.

I found several threads discussing this when I went looking. It's not a long argument. Trying to do too many things at once on the largest scale ever is time consuming and risky. Other companies doing smaller ships don't have a track record of moving fast so that argument doesn't hold.

 It's almost 3 years from the start of upper stage test flights on this one. Do you think a smaller version emulating successful Falcon technique would have taken more than another year or so to bring into service? Falcon9 profile using Raptors and stainless steel with RTLS baselined. If Raptors and methane are going to be fast turnaround and inexpensive to operate on Starship, then they would be the same on a smaller unit. Daily turnarounds of ships flying Starlink2 would be immanent. Starlink2 is revenue waiting to happen. The 6 meter diameter with 9 engines suggested above should be capable of 20-25 full Starlink2 per launch.

The above would take some pressure off of the full up Starship and be working from an experience base. Environmental and other permitting problems would be far less involved. But this has all been hashed out before. As I said, not a popular opinion that SS/SH at this scale could be a mistake for the first iteration of new technologies.

Starlink sats need to be launched at the same rate as current ones, if not faster...  so you need the large launcher.

Also, Mars and Starlink are not sequential. Starlink will fund the big colony drive, but that another 6-8 years out.

A smaller stepping stone would have just prolonged the timeline...  It may have taken a little bit less time, but they would have then had to repeat the exercise..  not a short cut.

SpaceX did "chicken out" on a 12 m design, but only because they believe 9 m will suffice for a Mars drive, at least for now.

Maybe 10 years from now we'll see a post-raptor ship that's 12-15 m..  but right now the 9 m ship is being developed as rapidly as any other small rocket if not faster.

[Brigantine]

The "colony drive" is a lot further in the future than many people here acknowledge... Starship is still an investment that's expected to make a return in a reasonable time frame - Humans on Mars is a different beast.

Falcon was too small for fully reusable upper stages to be cost-effective, since the added weight reduces the payload capability by a huge fraction.

Theory: Starship's size is just beyond what you need for a 2-stage fully reusable LEO vehicle to have a respectable ratio of payload mass : launch mass. (and thus unlock the "fully reusable" paradigm just when competitors manage booster recovery)

Test: What is the trade-space around diameter and payload-mass:launch-mass ratio? Where are these other threads about mini-starships? (particularly ones with re-usable upper stages)

[TomH]

To the OP: I recommend going into the archives. There are several hundred pages at least (in multiple threads) regarding the optimal size for BFR, Colonial Transporter, Starship. There are also many citations of tweets made by Elon himself re.the optimal size. It boiled down to this, something big enough to get to Mars, yet not so big that it could not take over commercial launches from Falcon 9. First and foremost, the thing had to be able to make money as a very frequently launched LV that would place commercial payloads into LEO (and some large DoD payloads into high energy orbits). It needed to be of optimal size to have a high launch cadence, and had to be able to out-compete competitors on price. Getting to Mars is the cherry on top, but there had to be a cash cow providing the milk for the ice cream Sunday holding that cherry up. There is also a well written historical documentation of the entire development of the LV by HyperionV.

SS will become an LV with roughly 3 times the thrust and 3 times the payload of a Saturn V. That is sufficient for the Mars objective with orbital refueling. It is capable of being a LEO workhorse and income earner as well. Elon felt that anything bigger might be overkill, not have the commercial launch flexibility, not be able to sustain as high a cadence.

This issue was debated in extreme detail years ago. There is also the excellent history of the entire project authored by HyperionV. The issue does not need to be debated again now for the benefit of someone not familiar with the highly documented history.

The OP is referred to the archives and I recommend to the moderators that this thread be locked. There is no need at all to rehash this.

P.S. There is also the very technical issue of how tall a rocket can be. The SS stack will be stretched, however the ability to do that is limited. When you make a rocket taller, you increase the thrust requirements to get the thing to adequate T/W (thrust to weight ratio.) You might say, Well, make it wider also, but that doesn't help at all due to the fact that mass is based on three dimensional volume, while thrust surface area is limited to two dimensional area on the bottom of the thrust puck surface. Think of it this way, you reach the maximum iSP and combustion pressure of the Raptors, cram them slightly closer together,  and build the 9m rocket to maximum height. You cannot make the thing taller, simply because you can't get more power out of the Raptors. Can you make it wider and thus also taller? Imagine clustering 7 full stack Starships, 1 center, surrounded by 6 more, strapped together. Independently, none of them can get taller, simply because you have reached all the thrust you can apply to the thrust surface of 𝝅r². Even if you strap them all together, none of them can get any taller due to the ratio of total mass (of which 3 dimensional volume is a factor) to the 2 dimensional surface area of the thrust surface. Now, merge all 7 of them together into a single immense LV with a diameter of 23-24 meters. That ratio of mass (predicated on 3 dimensions) to 2 dimensional thrust surface area, does not change. You now have a very fat LV, but you simply cannot make it taller (assuming an inability to increase Raptor thrust).

So let's say you build such a monster. The fineness (height to diameter ratio) of the upper SS stage changes dramatically. You wind up with a squat cylinder that is the same height, but is ∼2.5 x as wide. This changes its reentry aerodynamics immensely. One huge factor for the diameter of the full stack is the fineness ratio of the SS US in relation to its reentry aerodynamics, i.e. its ability to fly back to the tower and be caught by the chopsticks. The surface area to mass ratio is decreased and you have greater stresses on the fuselage during entry as well as less ability to neutralize the kinetic energy during entry and descent.

Elon and SX do not need a bunch of amateurs second guessing the design of this very complex machine all these years after the fact. Some here are qualified rocket scientists, but most of us, while pretty well educated, really are not practicing experts in the field of rocket science. Assuming that we might know more than these aeronautical engineers shows some hubris on our part. There's even a name for the condition:

https://en.wikipedia.org/wiki/Dunning–Kruger_effect  (software not including full string as hyperlink)

[BringBackSuperHeavies!]

Surprised they already went over it so long ago. But where are the archives anyways? Also does a 'bunch of amateurs' refer to the many people who have replied on this thread?

[TomH]

Surprised they already went over it so long ago. But where are the archives anyways? Also does a 'bunch of amateurs' refer to the many people who have replied on this thread?

1. No disrespect, but your surprise at that shows a lack of knowledge and understanding (read Dunning-Kruger). Engineers do a phenomenal degree of theorizing and modeling. SpaceX is famous for testing at the same time that it models and reiterating as it progresses. There is much pro/con analysis and comparisons that are done, both early-on and during development. Examination of trade-offs and swapping one feature for another for some reason is referred to as trades. They don't just throw these things together with no forethought. Discussion here is not at the same level, but most of it is far more advanced than you would hear between persons with no science/engineering background.
2. The archives are the older threads. They are chronological. Use the search function. Also, find HyperionV's history.
3. Most of us, absolutely yes. Many of us are well educated in science, physics, engineering, but are not rocket scientists. There are a number of SpaceX engineers who post on the site, but most of them are in the L2 section. There is one person who worked on Von Braun's Saturn V team, one who was an engineer on the older Atlas rockets, a NASA guy who handles payloads, including controlling the Mars rovers from loading to landing and a couple of rocket engineering professors. But most of us are scientists and engineers in other fields and not rocket PhDs. Our backgrounds give us some understanding of what is being discussed. Do you have a PhD in aeronautical engineering applied to rocket design? I don't, but I read accurate sources a lot. Again, see Dunning-Kruger. Then go read....extensively. One can become a very well versed amateur, but that still doesn't put one on par with Space X's designers. Thinking we might have better ideas about this rocket than they do is folly.

[redneck]

Surprised they already went over it so long ago. But where are the archives anyways? Also does a 'bunch of amateurs' refer to the many people who have replied on this thread?

1. No disrespect, but your surprise at that shows a lack of knowledge and understanding (read Dunning-Kruger). Engineers do a phenomenal degree of theorizing and modeling. SpaceX is famous for testing at the same time that it models and reiterating as it progresses. There is much pro/con analysis and comparisons that are done, both early-on and during development. Examination of trade-offs and swapping one feature for another for some reason is referred to as trades. They don't just throw these things together with no forethought. Discussion here is not at the same level, but most of it is far more advanced than you would hear between persons with no science/engineering background.
2. The archives are the older threads. They are chronological. Use the search function. Also, find HyperionV's history.
3. Most of us, absolutely yes. Many of us are well educated in science, physics, engineering, but are not rocket scientists. There are a number of SpaceX engineers who post on the site, but most of them are in the L2 section. There is one person who worked on Von Braun's Saturn V team, one who was an engineer on the older Atlas rockets, a NASA guy who handles payloads, including controlling the Mars rovers from loading to landing and a couple of rocket engineering professors. But most of us are scientists and engineers in other fields and not rocket PhDs. Our backgrounds give us some understanding of what is being discussed. Do you have a PhD in aeronautical engineering applied to rocket design? I don't, but I read accurate sources a lot. Again, see Dunning-Kruger. Then go read....extensively. One can become a very well versed amateur, but that still doesn't put one on par with Space X's designers. Thinking we might have better ideas about this rocket than they do is folly.

The flip side to your last sentence is the attitude that the experts have all the answers. Pretty clearly not the case, see SLS and company. Otherwise I am in agreement with what you said along with most of yesterdays' comments.

[laszlo]

Where are these other threads about mini-starships? (particularly ones with re-usable upper stages)

https://spacenews.com/stoke-raises-seed-round-to-work-on-fully-reusable-rockets/

[spacenut]

Some things about Starship/Superheavy.  They use the same diameter and tooling.  They use the same engine/engine components.  The only difference is the fins and TPS on the Starship.  Therefore Starship can be used without fins and TPS as an expendable upper stage or used as the moon lander by only adding good legs.  Also, without fins and TPS it can be used as a fuel depot.  It is very cost effective.

No different upper stage fuel.  No different upper stage engines except nozzle extensions on the vacuum versions. 

I, however when they got away from the composite idea, wish they would have went ahead with a 12m version, using the same number of engines, but have a shorter booster and starship.  Then after perfecting the landings, added even more engines or more powerful engines and stretched the vehicles so more fuel could go with more engines.  The wider version would have allowed for better legs to be contained within the bottom skirt of the starship for safer landings.  Stretching the vehicles only after mastering landings.  The starship version would have had a lower center of gravity also making for safer landings.

Then, I heard recently they are working on another engine that would give around 700,000 lbs thrust, which would allow for either the 9m version to be stretched or they could make a wider version.  It also uses metholox. 

[laszlo]

... Also, find HyperionV's history. ...

So much easier said than done. I just tried and this website's totally miserable search function for once agrees with a google site search. The only reference to HyperionV that shows up is in this thread.

Without a decent search function, it's not so much archives as a very large pile of papers in a very large room.

Until the site admins finally add a usable search function, posters should provide actual links to references they make that they want others to look at. At the very least, narrow it down to the appropriate thread. Otherwise no one will look at the "archived" info and we will be having this kind of thread ad infinitum.