Catapult start?

[DDE]

15 minutes ago, DerekL1963 said:

That would make you a minority of one...   Nobody else treats an ejection charge as the first stage.

Hell, the designers of Gnom treated a solid rocket that propelled the whole thing to M=0.75 as a 'zeroth' stage.

[PB666]

Lets examine what happened to this thread, we went from rubber-band assisted launch mechanisms to arguing over whether the ejection charge is a first stage or not.

The problem is how to gain dV so as to reduce 1st stage size and increase 2nd stage. Its not an argument as to whether you would or would not blow up a nuclear power submarine if you did not use some sort of ejection charge as per the point no-one is going to launch a Mars targeting space craft from a 100 feet underwater.

This is quite easy, buy some land in Ecuador, build a launch site there, take the rocket to the site and launch. If you don't like equador find an active volcanoe in the pacific

Hawai'i (i.e.  19°34′N 155°30′W ) 9' closer to the equator relative to Cape canaveral. Note that I used to use Ecuador as a launch site, the problem though is equitorial is only a really good ideo for GTO targeting, its not such a great idea for targeting objects elsewhere in the solar system because theta required to plane match another celestial there is a time of day of launch resulting in a AtP burn to escape that matches the required theta .

Elevation 4000 meters.
Reduction of specific thermodynamic energy = 40,000 J/kg 
total thermodynamic  energy required to make orbit ~ 1400000 j/Kg
total kinetic energy required to sustain orbit = 122448558

Minimum dV gained by launching 4000 m higher = 1.2 m/s (pure) dV.

dV gain as a result of ISP gain - engine based calculation

dV gained as a result of lower drag below Mach 1 - form based calculation

dV gained as a result of using a more efficient engine design - design based calculation.

dV gained as a result of reaching Max Q (Super Mach) higher in the atmosphere - again form and structural based calculation

dV gained as a result of having a higher TWR.

dV gained as a result of having a larger second stage, at least 4 times the composite of all the other dV gains.

 

 

 

[radonek]

Vertical catapult is also impractical on safety grounds - what if you throw up that rocket and then first stage fails to ignite?

[DDE]

44 minutes ago, radonek said:

Vertical catapult is also impractical on safety grounds - what if you throw up that rocket and then first stage fails to ignite?

Nothing good!

That's why Soviet naval VLSes were, allegedly, slightly canted to send those things overboard.

[kerbiloid]

10 hours ago, DerekL1963 said:

That would make you a minority of one...   Nobody else treats an ejection charge as the first stage.

That's exactly why I have added "very". i know where the "1st" and "2nd" stages are.

Edited November 20, 2017 by kerbiloid

[Nuke]

i think the general consensus is that its cheaper to build a slightly larger fuel tank than it is to use assistive technologies. 

like i had an idea to send a tunnel boring machine through a mountain at like a 30 degree angle. at the base you build a massive pressure vessel connected to the tunnel, a rocket is placed on a a rail system with a piston to seal the tube. in front of the rocket the tunnel is evacuated of most of its air. the end is sealed with a cap to keep the tunnel from repressurizing. when the valves open, the rocket will be driven up the tube, this compress what little air is left in the tunnel until its enough to blow the cap (which would be on hinges so it would clear the way for the rocket). you only need to exceed the pressure at the top of the tube which is significantly less than surface pressure. as soon as the cap is blown the engines are throttled up. this is at higher altitude so there are a lot fewer losses to drag. of course were back to the problem of being way too elaborate where a little bit more sheet metal and some more fuel was what you really needed.  

Edited November 20, 2017 by Nuke

[Bill Phil]

41 minutes ago, Nuke said:

i think the general consensus is that its cheaper to build a slightly larger fuel tank than it is to use assistive technologies. 

like i had an idea to send a tunnel boring machine through a mountain at like a 30 degree angle. at the base you build a massive pressure vessel connected to the tunnel, a rocket is placed on a a rail system with a piston to seal the tube. in front of the rocket the tunnel is evacuated of most of its air. the end is sealed with a cap to keep the tunnel from repressurizing. when the valves open, the rocket will be driven up the tube, this compress what little air is left in the tunnel until its enough to blow the cap (which would be on hinges so it would clear the way for the rocket). you only need to exceed the pressure at the top of the tube which is significantly less than surface pressure. as soon as the cap is blown the engines are throttled up. this is at higher altitude so there are a lot fewer losses to drag. of course were back to the problem of being way too elaborate where a little bit more sheet metal and some more fuel was what you really needed.  

True. However, the tyranny of the rocket equation begins to hit hard. 

Beyond reducing rocket size, it also improves payload mass fraction, whereas adding more fuel reduces it. If we can get the initial velocity high enough, then SSTOs become practical with relatively moderate mass ratios and decent payload mass fractions. Of course, dynamic pressure issues will be a signficant issue as well as an increase in atmospheric drag, but going up the side of a mountain may help with that. 

Rocket sleds have gone up to Mach 8.5. Nearly 3 km/s. Maglev would have almost no friction and electricity is pretty cheap. Maybe something could come of it in the future.

[DerekL1963]

1 hour ago, Bill Phil said:

Maglev would have almost no friction and electricity is pretty cheap.

It's not the operating costs that kill you - it's the cost of the not cheap part, the construction of the track.

 

2 hours ago, Nuke said:

i think the general consensus is that its cheaper to build a slightly larger fuel tank than it is to use assistive technologies. 

 

That's not a consensus, it's a stone cold fact.   The various forms of assisted launch only make economic sense when you have a high enough flight rate to amortize the capital costs of construction.
 

[radonek]

2 hours ago, DerekL1963 said:

That's not a consensus, it's a stone cold fact.   The various forms of assisted launch only make economic sense when you have a high enough flight rate to amortize the capital costs of construction.

Not long ago, stone cold fact was that it's uneconomical to land and reuse spent stages.

[DDE]

1 hour ago, radonek said:

Not long ago, stone cold fact was that it's uneconomical to land and reuse spent stages.

Red card! That’s a Galileo Gambit right there!

[Bill Phil]

4 hours ago, DerekL1963 said:

It's not the operating costs that kill you - it's the cost of the not cheap part, the construction of the track.

I'd make the argument that what kills it is the cost of developing a new launch vehicle that is designed to take advantage of the system. 

Maglev cost per kilometer is already in the tens of millions (or hundreds of millions) of dollars. But to get a few hundred m/s, you wouldn't need much length. The real costs would be in the buildings for integrating payloads and the other buildings you'd need.

 

[Xorth Tanovar]

Why not the following, it's a lot simpler...

 

[kerbiloid]

That rocket would get worn out against this rounded railway.

[DerekL1963]

3 hours ago, Bill Phil said:

Maglev cost per kilometer is already in the tens of millions (or hundreds of millions) of dollars.

That's for level track designed for (relative to a launch assist system) low acceleration, low speed, and low loads.  Apples and oranges.

And that's before figuring in that a launch assist system will be at least two parallel tracks for stability.

 

3 hours ago, Bill Phil said:

The real costs would be in the buildings for integrating payloads and the other buildings you'd need.

0.o  Buildings (even these types) are simple and cheap.  The labor alone (even including the internal systems) would be a fraction of that a kilometer of track requires.

Edited November 20, 2017 by DerekL1963

[magnemoe]

4 hours ago, radonek said:

Not long ago, stone cold fact was that it's uneconomical to land and reuse spent stages.

In one way you are right, consensus was that developing recovery would be very expensive and would increase the size and complexity of first stage so much it would have limited return value.
However the real killer was the development cost. 
Then came spaceX, first they simplified design a lot by landing with the engine rather than wings. Yes this works in part because they use lots of smaller engines. 
Second they cut a lot in development cost in testing on stages after separation, the grasshopper was their only pure test rocket. 
Crashing rocket until it worked was very kerbal but it worked. 

Now an catapult launch would give an far less saving and be very expensive to build. 
It has an added downside in the countdown procedure. you want to start the engines before you launch to check if all is running well. 
SpaceX take it so far that they ramp up to full trust before releasing. 
Waiting with starting the engines until you are up in the air and you increase the chance of failure, this could easy cost you the special launch pad you had build. 

Now for spaceplanes this make some sense as in trolley launch. 
First you only need the landing wheels to support the almost empty plane, not one filled with fuel. Secondary the trolley can give extra trust, jet, rocket or even electrical to help getting up to takeoff speed. Separation will be the dangerous part, you must probably have some sort of arms who let the plane pull it nose up while still connected to trolley and then separate at that would be wheels off. 
 

[Helmetman]

Not that this proves much cheaper. Or that it is a good idea for a return profit at all. The idea of catapulting just isn't.

But what about building a several kilometer deep elevator. A elevator that can push several hundred Metric ton and accelerate it to, euhm, let's just say to sub/transonic velocities. 
In the last few hundred meters reaching the surface acceleration will stop, clamps will be released and the engine ignited. The rocket pushes on it's own power a 100 or so meters below the surface when the rocket clamps releases the elevator from the rocket. The elevator platform is then timely decelerated to stop before the end of the tunnel so it doesn't collide with the installation on the surface.
The rocket then shoots out from below at transonic velocities only to maintain velocity up to and after Max Q. 
So you need less thrust only to maintain it. So a TWR of 1.05 should suffice, so that's less engines (less weight), less thrust (fuel consumption) and less time spent for the firing rocket engines in the denser part of the atmosphere.

The elevator probably needs a lot of equipment, air circulation to maintain sane air pressures within the tunnel it's launched from. This is completely hypothetical, I wouldn't know how to get a orbital rocket several clicks under, but let's assume all that equipment were in place.
But as for any randomly thrown out idea there's probably a gazillion of reasons why this is impossible, or not? It's probably not worth the investment. It would pay back over a very long time period as the project would be quite costly. But for the big bucks invested it will pay back over time if you could properly manage this.
 

[DDE]

@Helmetman Shorter tunnel, bigger lasers!

[Helmetman]

@DDE Even better  

[FleshJeb]

Some excellent background info and costing here:

https://en.wikipedia.org/wiki/StarTram

I recall seeing some discussion somewhere of Kilimanjaro (stratovolcano) as a site for mass-driver launch tubes. An extinct shield volcano would probably be better, due to the consistent and gentle slopes.

[GoSlash27]

15 hours ago, DerekL1963 said:

The various forms of assisted launch only make economic sense when you have a high enough flight rate to amortize the capital costs of construction.

^Succinct.
 Pure science folks often forget that there are financial and practical constraints at work as well as technical. Ultimately, orbiting payloads isn't just a technical exercise, it's a business model. If you can't deliver your payloads cheaper and more reliably than the other guy, then you may as well forget about it. All this talk about tracks up mountains, boring holes hundreds of miles long, building the world's largest circus cannon, etc... It's a whole lot cheaper for the customer and attractive to the investor to just go ahead and build a bigger rocket.

Best,
-Slashy

[PB666]

6 hours ago, FleshJeb said:

Some excellent background info and costing here:

https://en.wikipedia.org/wiki/StarTram

I recall seeing some discussion somewhere of Kilimanjaro (stratovolcano) as a site for mass-driver launch tubes. An extinct shield volcano would probably be better, due to the consistent and gentle slopes.

Imagine how many space fantasies begin with If funding starts now will be ready in 10 years. We have an example, funding on the fusion reactor began in the 1960s, there is still no fusion reactor.

The problem is this, don't build a track, find the highest suitable mountain, don't levitate use wheels and/or teflon. Find a pre-existing structure.

Lest say you rail launched from 10000 feet under ground (at sea level) mount everest at 29,000 feet. Lets say you accelerated at 10 meters per second at a 30 degree angle. So 29,000 + 10000 = 38000 feet  = 11582 meters. Now lets say that prior to launch the tube was evacuated, so no air resistance the only constraint is the structural 2g limit. That means acceleration is limited to 20 meters per second.

So lets do the math, tube at 30' means that the hypotenuse is twice the height or 23164 meters. The net acceleration is 14.7 a. 23164 = 0.5 * 14.7 * T^2, T = 56  seconds.
56 * 14.7 = 823.4 m/s (Mach 2.5). The Concorde aircraft typically broke Mach speed at 25,000 feet and cruise at mach 2 between FL 580 and 620. The SR71 black bird flew at Mach 3 between FL750 and FL800. Thus the air craft is going relatively fast for an altitude of 30,000 feet. In addition a typical space craft at this altitude would have a angle of attack very close to its direction of motion. This craft at 30' would have to have a thrust of at least 2g to maintain vertical velocity even as its angle of attack falls to the horizon (horizontal velocity would increase where as vertical velocity would remain constant or rise slowly causing the surface apparent angle to fall). Such a craft would have to begin steering vertically to prevent loosing most of its thrust as drag force created by a low pitch and low gain of altitude with thrust.

It needs to be mentioned that aircraft do not steer well over Mach 1 so steering the aircraft up after leaving the tube is problematic, and winging it up is limited by the shape of the tube.

Most of the gain you obtained from the tube could be gained by efficiencies by simply launching from the top of mount Everest and making a somewhat more rapid gravity turn (and the more space suitable engines one can use at the low atmospheric pressure)>

[Brofessional]

Imagine this, but with a Saturn V.

 

[wumpus]

I'll admit up front I hate gravity losses and the tyranny of the rocket equation.  I threw together a python program to check how bad Saturn V fuel efficiency was for the first minute:

0 s : 1.54625223835 m/s : 1.54625223835 m (height) : 1.1577647422 =TWR
5 s : 10.0469603887 m/s : 34.2612803397 m (height) : 1.18409158743 =TWR
10 s : 19.8735115925 m/s : 113.433200768 m (height) : 1.21164360424 =TWR
15 s : 31.0884883467 m/s : 245.877244367 m (height) : 1.24050835371 =TWR
20 s : 43.7590114972 m/s : 438.735000052 m (height) : 1.27078194432 =TWR
25 s : 57.9571899464 m/s : 699.498434434 m (height) : 1.30257010107 =TWR
30 s : 73.76062748 m/s : 1036.03632806 m (height) : 1.33598939901 =TWR
35 s : 91.2529956488 m/s : 1456.62344005 m (height) : 1.37116869187 =TWR
40 s : 110.52468331 m/s : 1969.97276257 m (height) : 1.40825077228 =TWR
45 s : 131.67353548 m/s : 2585.2712853 m (height) : 1.4473943091 =TWR
50 s : 154.805696652 m/s : 3312.21976104 m (height) : 1.48877611686 =TWR
55 s : 180.036576818 m/s : 4161.07704799 m (height) : 1.53259382616 =TWR
60 s : 207.491962264 m/s : 5142.70970671 m (height) : 1.57906903985 =TWR

Note that this ignores aero losses and assumes the spacecraft is traveling vertical and has full gravity losses (since 5km/s is ~mach .5 this seems reasonable).

By 60 seconds (and ~200 m/s delta-v) it has burned 784 metric tons of fuel (it burned 65 tons just to clear the tower). The whole thing weighs just under 3000 tons.

Let's forget about catapult launch and consider something already constructed: stratolaunch (no, it can't carry a falcon 9 much less a Saturn V.  It is just an example of a non-ground launch).
Two big features: the biggest (unless you *need* an inclination change) is that at 10km, atmospheric pressure should be 25% of sea level.  This should move your Isp to much closer to Isp (vacuum) and allow for bell design much closer to vacuum (increasing it more).  The other is about 450m/s in delta-v, as that is what it "costs" to get to 10km (you have to add any additional velocity (probably close to 200m/s [for mach .5]) via the pythagorean theorem as the "delta-v vectors" will be at right angles).

[reality check for those lusting after that  "450m/s".  The Saturn V example should include an extra "313 m/s" due to altitude, which can be simply added to the 207m/s at the end (both are straight up).  Also I'm pretty sure the Falcon 9 FT (don't know about the next one) should be pretty close to launching at ~1.5.  It should be possible to add an extra 500m/s via simply adding fuel until you take off at TWR=1.15  Adding fuel until TWR gets close to zero is reasonable if launching from a pad, but redesigning your aircraft because you increased the mass of your rocket isn't going to happen].

So that's pretty big.  But no plane will lift a Saturn V (and the reality check makes us less happy about planes anyway).  So lets do what Jeb wants and real rocket scientists do when they have to: MOAR BOOSTERS!!!

Same thing now with 4 SR-118 boosters (which wouldn't be possible since the Saturn V was dead in the 1970s and the SR-118 showed up in the 1980s.  But the obvious solid booster (in the Thor) had far too long a burn time.

0 s : 3.67342251632 m/s : 3.67342251632 m (height) : 1.3748007873 =TWR
5 s : 23.110690987 m/s : 79.6301232657 m (height) : 1.41145410163 =TWR
10 s : 44.4025650675 m/s : 258.297313217 m (height) : 1.45011536423 =TWR
15 s : 67.6523604544 m/s : 549.25460682 m (height) : 1.49095422156 =TWR
20 s : 92.9722739953 m/s : 962.624449832 m (height) : 1.53415998455 =TWR
25 s : 120.484431702 m/s : 1509.11961067 m (height) : 1.57994456286 =TWR
30 s : 150.322096065 m/s : 2200.09637888 m (height) : 1.62854594076 =TWR
35 s : 182.63106263 m/s : 3047.61435481 m (height) : 1.68023231482 =TWR
40 s : 217.571282612 m/s : 4064.50388391 m (height) : 1.7353070455 =TWR
45 s : 255.318756904 m/s : 5264.44239817 m (height) : 1.79411461536 =TWR
50 s : 296.067757844 m/s : 6662.04118585 m (height) : 1.85704784147 =TWR
55 s : 340.033449274 m/s : 8272.94443327 m (height) : 1.92455666085 =TWR
60 s : 367.488834719 m/s : 10054.5614543 m (height) : 1.57906903985 =TWR

So we add 200 tons of boosters to 3000 tons of rocket (and 784 tons of spent fuel) and get 160m/s more delta-v (and maybe another 100m/s of altitude delta-v)?  Don't tell Tsiolkovsky (I'm sure the equation insists you get more).   Actually all we are doing is helping get rid of those horrific gravity loses.  Also doing such would certainly violate the warranty on your Saturn V (not just the sides having to bear the extra load from the pull of the boosters, but bearing the extra weight thanks to the extra thrust/acceleration).  My guess was the engineers simply worked out roughly what they needed and then kept adding fuel till TWR nearly hit one, then added another F-1 engine (repeat until you have enough delta-v).  Obviously they had a pretty good idea, but fine tuning simply included adding fuel.

I assume that rockets that use rings of SRBs typically are rare heavy models of rockets that typically have much lighter duty.  You certainly *can* get a rocket to work well this way, but more fuel is typically preferred.  367m/s may look small, but it certainly beats 200m/s for burning 784 tons of fuel.  That's the  tyranny of the rocket equation for you.  It isn't so much that flinging a rocket off Everest at 824m/s (plenty of rockets burn half their fuel getting to mach 2), it is just building such a ramp is so ridiculously expensive that it is far, far cheaper to build rockets that  are twice the size.

My code in case you want to point out obvious bugs I missed:

Spoiler

startweight=6478000.0
startthrust=7500000.0
burnrate=28806.1

w=startweight
t=startthrust
v=0
h=0
for i in xrange(61):
    twr=t/w
    a=(t/w-1.0)*9.801
    v=v+a
    h=h+v
    if not i%5:
        print i,"s :",v,"m/s :",h,"m (height) :",twr,"=TWR"
    w=w-burnrate
    
    
#saturn V with 4 SR-118 boosters (Peacekeeper missile/Minotaur IV first stage)

w=startweight
t=startthrust

w=w+4.0*49.0*2204.6
t=t+4.0*226.8*2204.6
v=0
h=0
srburnrate=45.3*2204.6/56.4

for i in xrange(56):
    twr=t/w
    a=(t/w-1.0)*9.801
    v=v+a
    h=h+v
    if not i%5:
        print i,"s :",v,"m/s :",h,"m (height) :",twr,"=TWR"
    w=w-burnrate
    w=w-(srburnrate*4.0)

w=startweight-56*burnrate
t=startthrust

for i in xrange(5):
    twr=t/w
    a=(t/w-1.0)*9.801
    v=v+a
    h=h+v
    if not (i+56)%5:
        print i+56,"s :",v,"m/s :",h,"m (height) :",twr,"=TWR"
    w=w-burnrate

    

 

source for SR-118 data: http://www.spacelaunchreport.com/mintaur4.html
much of the rest was pulled from wikis or less reliable media...
 

[PB666]

Gravity losses are the imaginary enemy, its a little kitty cat standing in front of the mirror that makes it look big. On the grand scale the couple minutes it takes to get to cleaner air and pass Mach speed thats the problem.

If the air did not exist the Saturn V could have launched at 2g thrust turned quickly to 45' pitch and had an acceleration of 0.414g in the vertical and 1.414g in the horizontal. The reason that space craft launch strait up from earth is plain and simple, climb to get out of the thick air and turn into the thin air and finish turning when there is no air.

That why any discussion of a catapult are essentially useless unless you have find a significantly high enough launch point.

IF we had to we could build rockets that accelerated at 10g and you could turn within a few seconds so that 98% of the thrust is applied toward circularization, but as long as air exist all those g-forces become dynamic stress, heat and drag.

Oddly almost everyone who has played the game long enough to land on the mun and takeoff with the same engines knows they can turn to a shallow angle moments after take off using a little bit more than theoretical dV required to make orbit. Imagine making munar orbit if  50 km mountains surrounded the landing site on the mun. That's how you should think of the atmosphere on Earth, a big obstructive mountain that delays the application of circularization dV.

Here is how it looks. At equator the earth is traveling 450 m/s and 7824 m/s is required to make orbit. If you had a 10g rocket you would need to apply 7374 dV over 75.16 seconds. This IIRC requires a lead of 37.5 seconds.

37.5 x 9.81 = 367.85. Thus you would If the rocket could accelerate at 10g then 9g would be allowed to accelerate 4.087 sec and gravity losses would only be 40 m/s. So that 1500 to 2500 dV lost getting stuff to orbit, thats pretty much all due to drag and averting excess drag.

And yet knowing these some players still think that venus is a good place to build a colony, how in the devil would you ever leave a venusian colony (only one way off, and its a 'hell' of a fall).

 

Edited November 22, 2017 by PB666

[kerbiloid]

On 20.11.2017 at 9:29 PM, FleshJeb said:

I recall seeing some discussion somewhere of Kilimanjaro (stratovolcano) as a site for mass-driver launch tubes. An extinct shield volcano would probably be better, due to the consistent and gentle slopes.

Kilimanjaro site also has a special meaning for KSP.
If take a look at Kerbin, one can see that Kerbin continents are deformed Earth ones.
Here KSC is placed somewhere on Zanzibar, the runway archipelago matches Madagascar. and the mountains to the NNW from KSC are where Kilimanjaro could be.

Spoiler

On 21.11.2017 at 2:13 AM, GoSlash27 said:

Pure science folks often forget that there are financial and practical constraints at work as well as technical.

While true science folks know that 90% of science is a battle for food and place under the Sun, without taking of prisoners.