Countdown to new smallsat launchers

[TrevorMonty]

Why would you say that?
Some things don’t scale well, like computers and transmitters, and so are proportionally greater burdens to small vehicles. This is critical, probably fatal, to small SSTO, which need high mass fractions per the rocket equation.
And if they achieve the high mass fraction, final acceleration becomes untenably high without extremely low throttling.
We have seen lots of attempts at reducing the costs, recurring and initial. SSTO is not common for good reasons.

That's for vertical take off SSTO, where T/W ratio must exceed 1:1 just to lift off. HTOL opens more options if you have a suitable engine.

HTOL means additional lifting surfaces which is more dry mass, SSTO worst enemy.  Can do airlaunch but if Virgin is anything to go by its not cheap option.

I won't be surprised when they decide to switch to TSTO LV.

[edzieba]

SSTO (gloss over the 30kg bit) to grab venture-capital bucks for the shiny SSTO buzzword, then stack a shorter version of your SSTO stage on top with a bell extension and have a practical TSTO with reasonable payload. Or just make your single stage so danged cheap (Somewhat Big Dumb Booster) that it's still viable even when only launching a handful of cubesats.
The 'lets build a tiny FFSC Methalox engine' bit does seem a step too far into untrodden ground to get it done on the cheap, though.

[john smith 19]

And frankly, in an age when orbital rockets land "just like in Buck Rogers" I don't see why we should be dismissive of the idea that they could launch that way too.

Becase for a VTOL design you get 2-3x the payload on a TSTO that you get on a SSTO?

That has historically been the real killer of VTO SSTO concepts. 

For the same money (the theory goes) you can get 2-3 the payload mass to orbit.

The inverse argument is the driving idea behind pure Bi/Triemese concepts. Building a multi-stage vehicle with a single-stage budget.
[EDIT it is also something better than a rocket engine is needed to HTO to allow you to afford the mass of wings and landing gear while matching the payload fraction of a VTO TSTO ]

[trimeta]

Generically these smallsat launch concepts have been called "bricklifters" given their payload is around the mass of a housebrick.  I suppose they should be called cubesat launchers but I don't think there's a mass limit on the cubesat spec's, just what you can pack into a litre of volume.

Looking at the currently-active version of the CubeSat spec, there is a firm upper limit on mass for a 1U unit: 1.33 kg. It actually seems like this is going to change in the next revision, going up to 2 kg. But in either case, there is a well-defined upper bound. To what extent launchers care about that boundary is a separate question, but it is part of the spec nonetheless.

[Davidthefat]

Generically these smallsat launch concepts have been called "bricklifters" given their payload is around the mass of a housebrick.  I suppose they should be called cubesat launchers but I don't think there's a mass limit on the cubesat spec's, just what you can pack into a litre of volume.

Looking at the currently-active version of the CubeSat spec, there is a firm upper limit on mass for a 1U unit: 1.33 kg. It actually seems like this is going to change in the next revision, going up to 2 kg. But in either case, there is a well-defined upper bound. To what extent launchers care about that boundary is a separate question, but it is part of the spec nonetheless.

In my experience, the Cube Sat standards are more of a guideline than a spec. I've seen payloads well above the 5 kg limit for 3Us.

[Asteroza]

SSTO (gloss over the 30kg bit) to grab venture-capital bucks for the shiny SSTO buzzword, then stack a shorter version of your SSTO stage on top with a bell extension and have a practical TSTO with reasonable payload. Or just make your single stage so danged cheap (Somewhat Big Dumb Booster) that it's still viable even when only launching a handful of cubesats.
The 'lets build a tiny FFSC Methalox engine' bit does seem a step too far into untrodden ground to get it done on the cheap, though.

I dunno, with the recent discussion on plastic engines, the bar may have been lowered for small engine development? Either way, the original Bricklifter/Mockingbird concept is the concept of record that nobody seems to be actively emulating, which always struck me as odd.

[niwax]

SSTO (gloss over the 30kg bit) to grab venture-capital bucks for the shiny SSTO buzzword, then stack a shorter version of your SSTO stage on top with a bell extension and have a practical TSTO with reasonable payload. Or just make your single stage so danged cheap (Somewhat Big Dumb Booster) that it's still viable even when only launching a handful of cubesats.
The 'lets build a tiny FFSC Methalox engine' bit does seem a step too far into untrodden ground to get it done on the cheap, though.

I dunno, with the recent discussion on plastic engines, the bar may have been lowered for small engine development?

Turbo pumps scale badly enough that going electric is not only a valid option, but possibly better at that scale. Having two even smaller ones makes no sense at all. Not to mention neither coking nor a tiny isp increase is worth noting for this application.

[john smith 19]

SSTO (gloss over the 30kg bit) to grab venture-capital bucks for the shiny SSTO buzzword, then stack a shorter version of your SSTO stage on top with a bell extension and have a practical TSTO with reasonable payload. Or just make your single stage so danged cheap (Somewhat Big Dumb Booster) that it's still viable even when only launching a handful of cubesats.
The 'lets build a tiny FFSC Methalox engine' bit does seem a step too far into untrodden ground to get it done on the cheap, though.

I dunno, with the recent discussion on plastic engines, the bar may have been lowered for small engine development?

Turbo pumps scale badly enough that going electric is not only a valid option, but possibly better at that scale. Having two even smaller ones makes no sense at all. Not to mention neither coking nor a tiny isp increase is worth noting for this application.

It's why John Whiteheads team at Sandia developed reciprocating positive displacement pumps at small scale.

VTO SSTO is a radical strategy. IMHO the only serious attempts in the US have been Rotary Rocket and DC-X. Everything else in the US has been  Vugraphs and Powerpoints.

[john smith 19]

I dunno, with the recent discussion on plastic engines, the bar may have been lowered for small engine development? Either way, the original Bricklifter/Mockingbird concept is the concept of record that nobody seems to be actively emulating, which always struck me as odd.

Perhaps you should look at the original paper on the subject. here
That's a long way from a "plastic" engine, but it suggests directions for research.

Mockingbird was a reciprocating engine concept specifically because turbo pumps are difficult at this scale and they wanted to show better than pressure fed was possible at this scale, which they did.

[john smith 19]

Looking at the currently-active version of the CubeSat spec, there is a firm upper limit on mass for a 1U unit: 1.33 kg. It actually seems like this is going to change in the next revision, going up to 2 kg. But in either case, there is a well-defined upper bound. To what extent launchers care about that boundary is a separate question, but it is part of the spec nonetheless.

I stand corrected.
When I'm considering outer mass limits I go worst case with Tungsten. That's 20Kg/l. Obviously I don't anyone's going to make a cubesat that's a solid lump of tungsten but if your launcher could  lift it then anything lighter will be pretty easy.

OTOH this obsession with fineness ratio means you end up with very narrow launchers on which to mount your payload. If people are looking at SSTO seriously they have to consider the mass per unit length  of payload fairing (which got the DC-Y plan to put the payload bay between the tanks.

[john smith 19]

Now, if they were proposing some crazy new tech to make their SSTO work (aerospike, air-breathing, beamed power, etc.), then I would call this is nonsense. But they talk about using existing technologies. And when you consider the greatly increased performance of modern engines, the decreased size of computers (at least when compared to the rest of spaceflight history), and new materials technologies like composite cryogenic tanks... I think it's a concept viable enough to warrant an attempt.

SSTO ELV has been theoretically possible since the early 1960's with the Titan stages (whose design brief was basically "Deliver the Atlas ICBM payload but don't use pressure stabilization to do it").

Ultimate engine performance is set by the thermodynamics of the F/O combination. And note SX built possibly the largest composite LOX tank ever and rejected it in favor of a steel grade around since at least the early 1950's.

SSTO is a good idea only if it allows you to offer a cheaper launch price and need fewer launches to break
even.

SSTO is about as much a "slot in" concept to a design as building a heavy lift LV out of 3 regular boosters. In that case the core becomes a very different design because of the different stress paths unless you retain a totally common design and every booster gets heavier to cope with the off chance it will have to serve as a core in a 3 stick launch one day.

There's a whole playbook of design choices and hacks that engine and stage designers have developed to improve performance. Some can be retrofitted but others have to be in the design from day 1, starting with fuel and aspect ratio choices. Some have never been tried (differential throttling has been talked about for decades) How serious a design team is about SSTO can be gauged by how much the team has considered or incorporated these options in their design. 

But the killer is payload fraction. AFAIK no VTO SSTO has promised payload mass fraction of a TSTO (c2-3.5% of GTOM) because the structurally mass fraction is tight. 

So as a VC (just to be clear I'm not a VC IRL) why should I put my $ in your startup when the next guy I'm seeing will give me 2-3x the payload for the same investment (just like every other VTO TSTO startup that's looking for my funding in fact)?

Because anyone saying they can offer TSTO payload fraction in a SSTO will need to consider every aspect of their design from the ground up. It would be quite a potentially interesting idea though. 1 set of GNC, engines structure etc. Might be cheaper.

BTW historically spherical tanks (maximum volume, minimum surface area --> minimum mass) have been rejected on the grounds of drag and mfg complexity but (depending on the size) there are at least 2 ways to make pretty big spherical tanks either by spinning halves or by hydroforming from cylinders. Both are 1 step processes. Water jet cutting allows part to be cut in two while preserving the properties of the base alloy (no heat affected zone) and Holko at Ames in the early 70's demonstrated (and patented) ways to do large size diffusion bonding with low imposed pressures provided  the surfaces were very flat (16microinches, which is viable with sanding) and edge sealed. You need a big furnace (or high temperature heat blankets and lots of HT insulation on top of them) but you don't need the big press as well and you don't need a vacuum to make it work.

The drag issue should be put in perspective. Saturn V had 40m/s drag losses but more like 1200 m/s gravity losses. IE 30x higher. Drag is really a thing for cruise vehicles and LV that's cruising is in serious trouble. 

We'll see where they go with their design.

[TrevorMonty]

Now, if they were proposing some crazy new tech to make their SSTO work (aerospike, air-breathing, beamed power, etc.), then I would call this is nonsense. But they talk about using existing technologies. And when you consider the greatly increased performance of modern engines, the decreased size of computers (at least when compared to the rest of spaceflight history), and new materials technologies like composite cryogenic tanks... I think it's a concept viable enough to warrant an attempt.

SSTO ELV has been theoretically possible since the early 1960's with the Titan stages (whose design brief was basically "Deliver the Atlas ICBM payload but don't use pressure stabilization to do it").

Ultimate engine performance is set by the thermodynamics of the F/O combination. And note SX built possibly the largest composite LOX tank ever and rejected it in favor of a steel grade around since at least the early 1950's.

SSTO is a good idea only if it allows you to offer a cheaper launch price and need fewer launches to break
even.

SSTO is about as much a "slot in" concept to a design as building a heavy lift LV out of 3 regular boosters. In that case the core becomes a very different design because of the different stress paths unless you retain a totally common design and every booster gets heavier to cope with the off chance it will have to serve as a core in a 3 stick launch one day.

There's a whole playbook of design choices and hacks that engine and stage designers have developed to improve performance. Some can be retrofitted but others have to be in the design from day 1, starting with fuel and aspect ratio choices. Some have never been tried (differential throttling has been talked about for decades) How serious a design team is about SSTO can be gauged by how much the team has considered or incorporated these options in their design. 

But the killer is payload fraction. AFAIK no VTO SSTO has promised payload mass fraction of a TSTO (c2-3.5% of GTOM) because the structurally mass fraction is tight. 

So as a VC (just to be clear I'm not a VC IRL) why should I put my $ in your startup when the next guy I'm seeing will give me 2-3x the payload for the same investment (just like every other VTO TSTO startup that's looking for my funding in fact)?

Because anyone saying they can offer TSTO payload fraction in a SSTO will need to consider every aspect of their design from the ground up. It would be quite a potentially interesting idea though. 1 set of GNC, engines structure etc. Might be cheaper.

BTW historically spherical tanks (maximum volume, minimum surface area --> minimum mass) have been rejected on the grounds of drag and mfg complexity but (depending on the size) there are at least 2 ways to make pretty big spherical tanks either by spinning halves or by hydroforming from cylinders. Both are 1 step processes. Water jet cutting allows part to be cut in two while preserving the properties of the base alloy (no heat affected zone) and Holko at Ames in the early 70's demonstrated (and patented) ways to do large size diffusion bonding with low imposed pressures provided  the surfaces were very flat (16microinches, which is viable with sanding) and edge sealed. You need a big furnace (or high temperature heat blankets and lots of HT insulation on top of them) but you don't need the big press as well and you don't need a vacuum to make it work.

The drag issue should be put in perspective. Saturn V had 40m/s drag losses but more like 1200 m/s gravity losses. IE 30x higher. Drag is really a thing for cruise vehicles and LV that's cruising is in serious trouble. 

We'll see where they go with their design.

You are right about drag, LVs are through worst of atmosphere with first minute and clear of it after 2minutes which leaves another 7-8minutes of flight in vacuum to reach orbit.

[edzieba]

On the other hand, most of the cost in a launch startup is R&D, not manufacture. Almost every one I can think of has failed before reaching orbit, not after.
The cost to design an 'enormous' (for cubesats, small for 'regular' launch vehicles) SSTO stage is not significantly different from designing a small TSTO stage, and you only need to design one of them. If your response to "but the payload fraction!" is "So what?", why not build a 'mass inefficient' rocket that gets a payload to orbit with the cost to develop one stage, rather than the cost to develop two stages (and additional R&D time for staging)? By the time you're even concerned about increased BoM cost and operations costs, you've by definition already succeeded in getting your rocket to orbit and generating revenue.

Or in terms of cubesats: say you want an investment of $Xm to develop an SSTO stage that launches a single 3U cubesat, or $2Xm to develop two stages that can launch 3x 3U cubesats. But is there a market for 3x co-located cubesats that will pay 2x the price per launch to justify your 2x investment? In the face of "it's cheaper, but you have to share" SpaceX rideshares, Rocketlab continue to win dedicated launches after all.

[john smith 19]

The cost to design an 'enormous' (for cubesats, small for 'regular' launch vehicles) SSTO stage is not significantly different from designing a small TSTO stage, and you only need to design one of them. If your response to "but the payload fraction!" is "So what?", why not build a 'mass inefficient' rocket that gets a payload to orbit with the cost to develop one stage, rather than the cost to develop two stages (and additional R&D time for staging)? By the time you're even concerned about increased BoM cost and operations costs, you've by definition already succeeded in getting your rocket to orbit and generating revenue.

The key issue has always been that vehicle mass control must be ruthless 
Of course with million row spreadsheets tracking all mass properties of every single major part of a new design (and how they affect the overall GTOM, MoT, CoG etc) should be fairly easy
[EDIT
Key players in this area are the Society of Allied Weight Engineers
]
With the design uncertainties of a new vehicle almost no one has felt confident enough of their margins to believe they can get away with a SSTO.  They need the comfort of the possibility of absorbing any booster weight growth in a performance increase in the US (or vice versa).

Overcoming decades of the fear of delivering a stage that can't reach orbit is a very difficult thing to overcome. 

[john smith 19]

You are right about drag, LVs are through worst of atmosphere with first minute and clear of it after 2minutes which leaves another 7-8minutes of flight in vacuum to reach orbit.

Drag loss is one of those things that needs a whole trajectory integration to get a realistic estimate of.

Historically that was weeks of work. Today OTS software can handle that if you're a specialist and a spreadsheet model can give a crude understanding of what's going on in an afternoon, as long as you remember when to use the resolved components of a force and when to use the force along the axis of the vehicle.

But the folklore from the 40's,50's and 60's dies hard.   

The "Lower the drag loss by leaving the atmosphere ASAP" give high gravity losses (long period in pure vertical flight) and  high steering losses (as the nozzle is substantially off axis WRT to the instantaneous direction of the vehicle when you decide to start tipping the vehicle).

But by starting to go horizontal (even if it's just a fraction of a degree per second of flight) you extend your time in the atmospheres.  Oh noes.    Drag losses increase.

For best overall performance (IE lowest total loss number) you have to consider all three, but remember gravity losses are always 10x of times bigger than the others.

If you thinking "Who cares?" consider the difference between a high loss LV (Saturn V) and a low one (Delta 7934 7925) is about 1 Mach number. Last time I ran an SSTO calculation that's a difference of about 7 seconds of Isp needed to get SSTO performance.

And this can all be done before a  single piece of hardware gets ordered.

[Lars-J]

The "Lower the drag loss by leaving the atmosphere ASAP" give high gravity losses (long period in pure vertical flight) and  high steering losses (as the nozzle is substantially off axis WRT to the instantaneous direction of the vehicle when you decide to start tipping the vehicle).

This is so wrong, I don't even know where to start. High steering losses? LOL. A properly designed trajectory has almost none of that.

Google "Gravity turn": https://en.wikipedia.org/wiki/Gravity_turn

All a rocket needs to do is a slight initial pitch maneuver after lift-off - gravity does the rest.

[Asteroza]

I dunno, with the recent discussion on plastic engines, the bar may have been lowered for small engine development? Either way, the original Bricklifter/Mockingbird concept is the concept of record that nobody seems to be actively emulating, which always struck me as odd.

Perhaps you should look at the original paper on the subject. here
That's a long way from a "plastic" engine, but it suggests directions for research.

Mockingbird was a reciprocating engine concept specifically because turbo pumps are difficult at this scale and they wanted to show better than pressure fed was possible at this scale, which they did.

I meant plastic engines for an easier dev cycle allowing high physical iteration, not the final engine.

Also, due to nomenclature, I feel I should point out for those reading that the rough 5000lbs rule of thumb for turbopumps vs something else (reciprocating, specifically Whitehead envisioning a third gas driving a piston pump) is specifically referring to the difficulty of manufacturing/designing the turbine driving the compressor pump (axial or centrifugal). Which is why when Rocketlabs substituted an electric motor for the turbine, they get most of the turbopump benefits.

Looking at the currently-active version of the CubeSat spec, there is a firm upper limit on mass for a 1U unit: 1.33 kg. It actually seems like this is going to change in the next revision, going up to 2 kg. But in either case, there is a well-defined upper bound. To what extent launchers care about that boundary is a separate question, but it is part of the spec nonetheless.

I stand corrected.
When I'm considering outer mass limits I go worst case with Tungsten. That's 20Kg/l. Obviously I don't anyone's going to make a cubesat that's a solid lump of tungsten but if your launcher could  lift it then anything lighter will be pretty easy.

OTOH this obsession with fineness ratio means you end up with very narrow launchers on which to mount your payload. If people are looking at SSTO seriously they have to consider the mass per unit length  of payload fairing (which got the DC-Y plan to put the payload bay between the tanks.

Well, there is the rather specific tungsten payload of flechettes, known as "Rods from God", but that isn't very civil...

You are right about drag, LVs are through worst of atmosphere with first minute and clear of it after 2minutes which leaves another 7-8minutes of flight in vacuum to reach orbit.

There's also the ugly structural issue that sending up a high cross-sectional area onion shaped SSTO means your Max Q dynamic pressures will affect the structural design, and thus weight, of your vehicle. The gravity losses may stand alone, but aero drag losses don't stand alone.

[niwax]

SSTO (gloss over the 30kg bit) to grab venture-capital bucks for the shiny SSTO buzzword, then stack a shorter version of your SSTO stage on top with a bell extension and have a practical TSTO with reasonable payload. Or just make your single stage so danged cheap (Somewhat Big Dumb Booster) that it's still viable even when only launching a handful of cubesats.
The 'lets build a tiny FFSC Methalox engine' bit does seem a step too far into untrodden ground to get it done on the cheap, though.

I dunno, with the recent discussion on plastic engines, the bar may have been lowered for small engine development?

Turbo pumps scale badly enough that going electric is not only a valid option, but possibly better at that scale. Having two even smaller ones makes no sense at all. Not to mention neither coking nor a tiny isp increase is worth noting for this application.

It's why John Whiteheads team at Sandia developed reciprocating positive displacement pumps at small scale.

VTO SSTO is a radical strategy. IMHO the only serious attempts in the US have been Rotary Rocket and DC-X. Everything else in the US has been  Vugraphs and Powerpoints.

I don't necessarily agree that there has been no serious attempt, it's more like there has been no serious demand. A large number of first stages could have been readily incorporated into an SSTO had someone asked for it. Ever since the original Atlas you could have bolted a cubesat to an existing rocket to make . A Falcon 1C has 8900m/s in the first stage and engine and avionics commonality with all other SpaceX rockets at the time. No one asked to put the fairing straight on it and keep flying when Kestrel was retired. An Electron could just about get a cubesat or two to orbit if they put on the battery jettison from the second stage. Instead they're adding a third stage on most missions.

If an idea gets less interest the closer it is the being real, that's not a great look. Yes, you could say that VTO SSTO is pretty radical because people only look at HTO. But you could also say that people only looked at SSTO when it became a necessity because they wanted to do any sort of HTO vehicle it's the simplest way to get a viable product out if they to SSTO first, while everyone who can afford to do stages does so immediately.

[john smith 19]

This is so wrong, I don't even know where to start.

Then perhaps you shouldn't? 

Engineering is done with numbers, not opinions. So here are some actual numbers.

According to the wiki article both Shuttle and Saturn V used the gravity turn maneuver. Saturn has the highest gravity loss (and the highest total loss) of any of the systems listed, but also the lowest drag loss. Bravo!

OTOH its steering losses are the 2nd worst of the list. Shuttle did worse on steering and drag losses but did much better on gravity losses, which are much bigger to begin with.

Keep in mind that both were hydrolox systems with the highest Isp fuels available in the second stage (and with shuttle in the first as well).

Now look at that Delta 7925 (7000 series, 9 solids, 2=hypergolic 2nd stage, 5m fairing). It has nearly the worst drag (despite its much narrower dia than the Saturn V) 3.5x higher. Oh noooes. 
OTOH its steering losses are less than 1/7 that of Saturn V and its gravity losses less than 3/4. Note Delta is a kerolox 1st stage, hypergolic second stage, with 6 solids firing at launch and 3 (vacuum optimized nozzles) firing late in first stage just prior to 2nd stage separation.

And note the absolute  values of those losses and the Delta orbital velocity the rocket has to reach is higher than either Shuttle or Saturn V.

The total delta V needed by the Saturn V is 453m/s higher than the Delta 7925, despite getting more assist from the earths rotation and having the lowest drag losses of any vehicle and the highest performance fuels to use.

TBF when Saturn was designed trajectory simulation was a sloooow task. So folklore and rules-of-thumb saved time, which was very important.  Today if you're running a for-profit rocket mfg company there is no excuse not to do these simulations.

IMHO what these results show is a) Gravity losses are always the biggest part of the loss budget. b) Steering and drag losses need to be considered together c) The goal is to minimize the total losses. Delta sacrifices 96m/s of drag to gain 210 m/s in steering loss improvement compared to Saturn.

I don't know if the 7925 uses a gravity turn or not. But gravity turns don't seem to have helped either Saturn or Shuttle with their losses much. 

[Steven Pietrobon]

All a rocket needs to do is a slight initial pitch maneuver after lift-off - gravity does the rest.

As I discovered when I tried to simulate the Saturn V, that only works for the first stage. You need to do a controlled pitch up after that in order to get into your desired orbit. Otherwise, you end up in the drink or into a highly elliptical orbit. I believe this is the main contribution of steering losses, which can be reduced by using a higher thrust second stage.

I suspect the reason why Delta 7925 has a lower gravity loss is the higher initial acceleration from all those solids. Same with the Space Shuttle. Drag losses are mainly related to the mass to area ratio of the vehicle. A higher mass vehicle for the same area will have less drag losses, which I like to call the Titanic effect. The bigger your vehicle is, the less drag loss you will have, which is evident in the Saturn V.

The way I think a vehicle should be designed is around the initial acceleration. If you have a certain thrust available is it better to carry less propellant for a lower gravity losses (but less payload without any losses) or do you try to maximise the amount of propellant? My inclination is to carry as much propellant as possible for the thrust available. This effectively equates to what is the best initial acceleration, which is a value which can be easily optimised in your design. I always tend to go to 1.2g as being close to "optimum".

The argument of "Let's increase thrust to decrease gravity losses" in my mind goes to "Let's also increase propellant mass to get the maximum performance." My argument goes as follows. If I start at say 1.5g I will be able to put a payload of a certain mass into say a 200 km circular orbit. Now lets load up with more propellant and start off at 1.2g. During the time from 1.2g to 1.5g, I will have gained a certain ideal speed (about ve*ln(1.5/1.2) ~= 670 m/s ideal for an exhaust speed of ve = 3 km/s). This ideal increase in delta-V will be reduced due to gravity losses and a higher first stage dry mass, but its still an increase which effectively allows for an increase in payload mass.