Questions

[Spaced Out]

So why is it that rockets in real life burn for so much longer than in KSP? Is it because they need to get so much horizontal velocity before their apoapsis reaches the target height that they have to be at a really small angle so they can get as much as possible before their apoapsis gets there?

Edited December 10, 2017 by Spaced Out

[Steel]

Exactly. IRL you need about 7500 ms^-1 of horizontal velocity at LEO, in KSP you need about 2000 ms^-1. In both cases you only have to be about 150 km up, so you have to burn much longer and much more towards the horizon in real life.

Edited December 10, 2017 by Steel

[DDE]

1 hour ago, Steel said:

Exactly. IRL you need about 7500 ms^-1 of horizontal velocity at LEO, in KSP you need about 2000 ms^-1. In both cases you only have to be about 150 km up, so you have to burn much longer and much more towards the horizon in real life.

@Spaced Out, there's also the tendency of KSP players to use REALLY overpowered engines in upper stages.

[magnemoe]

4 minutes ago, DDE said:

@Spaced Out, there's also the tendency of KSP players to use REALLY overpowered engines in upper stages.

You need an overpowered upper stage as your burn until 80 km height is far shorter. both in height and in distance because the tiny planet. 
You can burn somewhat downward to move AP in front of you. Make an node just to get time to burn start a bit ahead and keep moving AP, this cost more fuel but let you use an low TWR upper stage with LV-N or ion engines. Then dock at an space station, we want full load of fuel for the nuclear stage, also top our lander and the probes. We would also like some science to study in transit, we have some spare parts for future missions to. 

 

[regex]

10 minutes ago, DDE said:

@Spaced Out, there's also the tendency of KSP players to use REALLY overpowered engines in upper stages.

While true, to be fair there's really no way around short launches, even with low TWR engines, on Kerbin because of the relatively low orbital velocity. Furthermore, because of the short times to orbit there is much less room for precision adjustment (radial burns), although it is certainly possible. Because of short times to orbit and lack of time to make precision adjustments (and the peculiarities of KSP's engines) most players prefer a circularization burn whereas most IRL rockets burn directly into orbit.

[DAL59]

1 hour ago, Steel said:

2000 ms

Only 2000?  I've been launching wrong.  

[Green Baron]

Orbital speed at 120km is around 2200m/s, at 80km it is ~2300m/s, dV from the ground to orbit is 3.300-3400, gravity loss, a small margin and friction included.

Edited December 10, 2017 by Green Baron

[PB666]

2295 at 70 km, and this does not belong in this group.

[Spaced Out]

23 minutes ago, PB666 said:

2295 at 70 km, and this does not belong in this group.

Well I supposed it did belong in this forum because this involves real life too, but mods feel free to move it.

[NSEP]

1 hour ago, DAL59 said:

Only 2000?  I've been launching wrong.  

2km/s is the orbital velocity. You need 3,4 to 4,1 km/s of total delta to get into orbit (because of the atmosphere and vertical takeoff)

[PB666]

35 minutes ago, NSEP said:

2km/s is the orbital velocity. You need 3,4 to 4,1 km/s of total delta to get into orbit (because of the atmosphere and vertical takeoff)

The minimum orbital velocity required is =SQRT(2*(70,0000*8.5 + 1/2(2295-175)^2)) = 2384m however because it takes two burns to orbit a body some of the first burn must include gravity losses. So lets say 2500.

The additional 1000+ dV required can be lowered by making parts with higher density and lower coefficient of drag. Modding alllows you to make these and I have a number of fuel tanks with much lower drag. For drag modding to work properly in KSP you need the drag mods to occupy the entire stack (IOW a stack of fuel tanks becomes one aerodynamic fuel tank.
The other thing is you need thrusters that can fight gravity at very low angles. If we were to make a hohman transfer from the surface of kerbin the launch pitch would nearly be zero, unfortunately we would have no lift.

To compensate if you launch at 10' pitch the the rise would be 0.173 so that to remain up you would need to produce 60 m/sec². To launch at 45' requires 12 m/Sec^2 but the rate of horizontal acceleration falls.

If we imagine a launch where you travel up to 10 K then turn at 45' and travel to 30K and turn to 30' pitch. At 30K plus are climb is 86.8% efficient, at 10 to 30k our climb is 70% efficient, at 0 to 10K our launch is 3% efficient compared.

So now the problem is time, time x inefficiency = loss. If it takes 2 minutes at 1.6 g to reach 10k, it means you spend 320 dV to get 1/7th the needed altitude and 1/12th the needed velocity, most of the other 1/12th is wasted because the direction vector is off by 90' and although it has value at  a displacement theta, you still loose most. In turning to 45' the next loss is from gravity and drag as speed increases the force squares as altitude increases the force drops as a function of scale.
The more aerodynamic the space craft the more rapidly it can turn, the more rapidly it converts dV into horizontal thrust and the higher thrust requirment the spacecraft can have, that means a smaller proportion of the thrust goes into keeping the craft up.

So for a craft of standard KSP design about 1.4 TWR is a good launch 65% is wasted.
With improved parts 2.0 TWR 50% is wasted until turn to 45 in which 40% is wasted.
With greatly improved parts 3.0 TwR 33% is wasted at launch and turn to 45 30% is waste< turn to 30' and 14% is wasted.

The problem is that drag is the square of speed, that affords 50% drop in drag = 41% higher speed or that speed about 3000 meters below.
So that if we cut drag by 3/4ths it really only means we can turn about 3000 meters lower and have 50% higher speed.

One earth it does not make that much of difference but on Kerbin it does. Gravity decreases faster on kerbin, getting into lower gravity faster saves up to 10%. The atmosphere is thinner which means you pay a cost in drag for a shorter period of time if you hammer your speed as long as you are moving up by 100s of meters per second.

 

[Spaced Out]

1 hour ago, PB666 said:

The problem is that drag is the square of speed, that affords 50% drop in drag = 41% higher speed or that speed about 3000 meters below.
So that if we cut drag by 3/4ths it really only means we can turn about 3000 meters lower and have 50% higher speed.

Can you explain this a bit more? I don't really get what you mean.

1 hour ago, PB666 said:

The more aerodynamic the space craft the more rapidly it can turn, the more rapidly it converts dV into horizontal thrust and the higher thrust requirment the spacecraft can have, that means a smaller proportion of the thrust goes into keeping the craft up.

Also can you explain this a bit more too?

[PB666]

First off if you don't have Mech Jeb, or Flight engineer get one of these.

Here are some targets. This is done with a sleek rocket and a powerful engine (TWR = 7)

If you launch strait up with a fairly large fuel mass and alllow the apoapsis at 70k, then a powerful burn 10s of second before hitting apo you can achieve orbit for 3122 m/s dV with a very streamline rocket waste is 750 dV)

If you launch strait up and gain a apoapsis of say 40k, then close to 40k turn to 20' pitch and fire until Apo is 70 k. and then circularize. Dv required is 2688 (waste is 304 dV)

If you launch to 60' and then turn to 45'  gain a apoapsis of say 72k, Apo is 70 k. and then circularize. Dv required is 2592 (waste is 210 dV)

Using standard KSP equipment.

Launching and turning slowly to 45' by 15,000 then to 30' by 25' then zero by 56' and circularizing at 70k. dV = 3406 (waste is 1022 m/s) [Aerodynamic rocket saved about 700 dV]

Launching strait up and circularizing at  to achieve 70k (i.e burn along 0'pitch starting 30 or so seconds before reaching 70k). dV = 4043 (waste is  1658 m/s) most lost through the slow burn at 1.4g.

In these scenarios we can see the loss, some of the losses are from having a not so streamlined rocket, some are from not being so powerful.

To achieve a good launch , launch strait up start with TWR = 1.4g until you have a good build strategy. At about 100m m/s tilt ' to 89'

From there tilt about 1' per 1000 meters, until you get close to 9,000 meters then start tilting faster.

Between 10 and 15 k meters if you have more power use it but once you get to 310 m/s hold speed past 18 k feet, at this point you should be around 45 to 60' pitch
Here is the formula I use from that point on. By the way trubngin your rocket fast will flip your rocket if it does not have AV-R8 winglets to keep it close to the true.

The way a control a rocket between 0 and 24k has alot to do with its shape and with other flight characteristics, if the rocket is sleek and powerful I turn it more quickly, If im hauling up a bulky payload I use less TWR and turn later.
The strategy for slow rockets is basically to peel off boosters until it safe to burn at a higher speed.

When the apo passes 26k  then pitch is 40',
when apo passes 36k pitch is 30',
when apo passes 46k pitch is 20', AV
when apo passes 56 k pitch is 10'.
Adjust apo and burn speed from that altitude to control the circularization process, highly dependent on the thrust of the space craft.

Although you can make orbit with nth-stage TWR below 1 on kerbin (not on earth) it is much easier if your vacuum engine has a TWR about 1.5 and you will waste. So that once you are over 45k meters and your speed is below 1600 m/s (orbital) TWR should be close to 2.

 

 

[Spaced Out]

54 minutes ago, PB666 said:

 

From there tilt about 1' per 1000 meters, until you get close to 9,000 meters then start tilting faster.

 

Thanks for answering, but I have another question. Don't you just need an initial pitchover, not manual control?

[Starman4308]

58 minutes ago, Spaced Out said:

Thanks for answering, but I have another question. Don't you just need an initial pitchover, not manual control?

If you have an aerodynamically stable rocket and are very careful and consistent making your initial launch, yes. In practice, most players are going to get that approximately right, but still keep their hands on the controls in case they turned too early or too late and want to pitch up/down a bit.

Edited December 11, 2017 by Starman4308

[wumpus]

5 hours ago, PB666 said:

One earth it does not make that much of difference but on Kerbin it does. Gravity decreases faster on kerbin, getting into lower gravity faster saves up to 10%. The atmosphere is thinner which means you pay a cost in drag for a shorter period of time if you hammer your speed as long as you are moving up by 100s of meters per second.

Gravity falls off faster on Kerbin (73% at 100km) than Earth (89% at 400km), but I suspect that Kerbal spacecrafts' lousy mass ratios (compared to real life) makes that a pretty moot point (assuming that nearly all the gravity losses occur in the first stage).

[PB666]

If you are really on top of things you can do this, but the problem is that if you pitch over too fast, what happens you end up losing all your dV to drag and may end up stressing parts or overheating solar panels. I pitched over too quickly awhile ago and lost my McJeb controller.

There is no uniform thrust situation, so its best to control things. If you have MechJeb you can look at drag force, this will tell you when the rockets form drag my be so high as to tilt the rocket or break something.

Theoretically on lift off the best acceleration gives you a drag that equals effective gravity, this however is not generally a good thing in rocket science. The reason is that once a rocket takes off its already got 1g of  of force, as it climbs the a increases the force on the upper parts increases while those on the engines only increase with ISP as you approach 273 m/s your drag force is no longer the square of speed but increases quite rapidly. Thus the force at the top of the rocket increases with acceleration and with drag creating an ephemeral quantity known as Q. At a certain Q the rocket is at risk of damage or flipping. Thats why its a good idea to go 'up' until Q goes down and then go 'sideways.' 

since pressure ~ 1/alt²  rougly to be precise below Mach speed. At exactly Mach speed is 3 times as great, and then it falls.

q =  P₀ * e^(-g*M*h/R*T) * V² /2

If you apply increasing acceleration (as a result of fuel loss) to a rocket going strait up, q will increase but because dp/dh has a dependency of v [ p is ~ e(-h)] as h get large quickly pressure drops quickly and q follows.

For this reason you want q to fall away because of the increase in h before attempting to break Mach Speed in many rocket designs (especially using KSP parts because they have a tendency to be bulky).

The problem with this that the rate at which the AoW (surface pitch bug) drops is both a function of pitch and speed, so that if your pitch is too high slow down and it will fall over more quickly, if your pitch is falling too low, speed up and it will fall less quickly. If your pitch is way too low (e.g 30' at 10,000) and your altitude is too low and you speed up, you burn all your dV fighting drag. So if your space craft is draggy, then travel up and turn later. If its very sleek turn sooner and power up to 2g.

The other reason you want to control the pitch is this: For your heaviest rockets you will be staging. Heed this warning, stage separation always works better when the wind bug and pitch are identical; however, this is particularly true for radial separation. The problem is for instance, you don't want to dump alot of used rocket parts on your space center, it costs money. So you lift turn down course a few degrees maybe  seperate then turn back to 90 and then turn again to match the bug and separate, then maybe let the rocket turn and separate more boosters. In my biggest rocket designs (32,000 tons) I want to hold back the turn of the rocket and 3 seconds before stage separation let it tilt over until exactly on the bug separation while still under full power. Following this I might separate structural parts offsetting the stage from the core (The hydrolic separator just does not provide the spatial clearance for huge boosters). IOW you don't want to be correcting pitch because of drag while you are trying to separate a 1kt boosters. You may see a ball of fire and some control part appears on its way back to the planet.

Edited December 11, 2017 by PB666

[regex]

1 hour ago, Spaced Out said:

Thanks for answering, but I have another question. Don't you just need an initial pitchover, not manual control?

As stated it depends on the rocket. In stock/vanilla with a TWR of around 1.25~1.4 on the pad, and depending on upper stage TWR, I start the turn at around 60~80m/s. I tilt about 10 degrees, wait until the prograde marker on the navball has reached that point, then slap the prograde button next to the navball, and wait. Once I stage I assess the altitude and apoapsis, and make adjustments during the burn. With a low TWR upper stage I turn later so it gets "lofted" to some extent, with a high TWR stage I can make a flatter trajectory. Circularize at apoapsis.

Obviously this all depends on the rocket and how it flies but that's generally how my launches go. Everyone has their favorite technique that works best for them; mine acommodates pretty much every weird payload I want to put into orbit on top of whatever launcher fits around it.

Interestingly, when making second stage burn adjustments, I tend to burn radially inward (towards the planet) in vanilla/stock because I want to bring down the apoapsis, whereas in RO/RSS I typically find myself burning radially outwards (away from the planet) to raise the apoapsis and lengthen the time to it due to the longer burn times.

All that is flying manually and trying to acommodate many different rockets. Computers that are made for the rocket in question can take the best path to orbit (I think there's a MechJeb module somewhere that can launch Ariane 5 perfectly, or you can always KOS it).

Edited December 11, 2017 by regex

[PB666]

17 minutes ago, regex said:

Interestingly, when making second stage burn adjustments, I tend to burn radially inward (towards the planet) in vanilla/stock because I want to bring down the apoapsis, whereas in RO/RSS I typically find myself burning radially outwards (away from the planet) to raise the apoapsis and lengthen the time to it due to the longer burn times.

The reason for that is in KSP the atmosphere height is 70 km and you can power up  your burn to horizontal about 24km which means the burn to horizontal meaning you have 46 km to burn out about 1700 dV and to add to that your gravity falls off quickly 0.5 g is 240 km above surface, and on top of that you are approaching w2r more quickly. Once you establish a suborbital trajectory you have plenty of thrust*time to circularize before you reach the apex and in general this is going to push the apo higher. If we go by the equation d =  0.5 a t² and a = 6.0 on average then you have 118 seconds to burn at 13+ a and that gives 1534 dV that with the spin of the planet and 400 or so horizontal V before the turn the 2295 you need to orbit.

On earth your orbit is 140 and you can start really powering up the burn to horizontal about 40 km, meaning you have 100 km to traverse but 6950 dV to burn. Since the average gravity is nearly 9.8 and w²r  is not realized as countering gravity until relatively much later the typical downward acclearatoin is around 8.0 a. 100,000 = 0.5 * 8 t² meaning that you have 158 seconds to burn the 6950 dV and thus using 1.3 TWR you would not achieve that speed (only 2054) with a lateral burn in the required time. Either you have to turn later getting a higher apogee or burn harder. Burning harder is very difficult to do and 140,000 m is rather low anyway, so a combination of both suffices.

If you have a 400 V_tan, Earth rotates at 450 m/s surface, and you have TWR of 2.0 during the burn. Total orbital velocity at the apex is 4010, which does not seem bad until your realize that at the end it only reduces effective gravity by about a quarter son instead of 9.8 effective gravity is a = 7.3. If the rocket has a TWR of 2.0 and you want to hold altitude you need a pitch of 21' which is not too bad because 93% of your thrust goes into horizontal acceleration and that increases to saturation quickly as the pitch falls. Are there ways to improve on this:
1. Increase the amount of thrust in your upper stages, this decreases the time and decreases the cosine losses.
2. Burn high quickly after launch and turn and burn as quickly as possible keeping the pitch about 10' above the prograde bug until omega-based waste reductions take effect.

Note this is why I was complaining about the RD-150's thrust a while back, as an upper stage engine it really was not up to the task of burning to orbit. RL10b-2 is not a final burn to orbit engine either. On many of my rockets I just use the RL68-A to get to orbit.
 

[regex]

33 minutes ago, PB666 said:

RL10b-2 is not a final burn to orbit engine either.

DCSS pushes 28 tons to 200km orbit with a single RL-10b-2. It's a perfectly serviceable upper stage to-orbit engine. RD-150, IIRC, was more than capable of that.

[PB666]

8 hours ago, regex said:

DCSS pushes 28 tons to 200km orbit with a single RL-10b-2. It's a perfectly serviceable upper stage to-orbit engine. RD-150, IIRC, was more than capable of that.

28 tons 28,000*9.8  274400. 110000/274400 = 0.404. u = 398600441800000. Lets see what the laws of physics say. 6371000 + 200000 = 6571000. Vorb = 7788 m/s and g = 9.23

1125 sec x 110000 = 12375000/(462*9.8) = 27332 kg of fuel, empty mass is 3042 kg =30710. If the payload is 28 tons then 58710 at the beginning of the burn would have a TWR of 0.198.

if r = 6571000 and a = 1.98 m/s then our ship is only capable of fighting gravity when 9.23 - 1.9404 < centripedal acceleration = 7.29 a_c

SQRT(ar) = V .  V = 6291 So in terms of burning to orbit, its could only burn for the last 1500 horizontal dV and do so rather inefficiently because of a huge vertical component.

Which . . . .is what I was saying. I have a working model of the RL10b-2, I generally have 3 or 4 of these on the backside when I need to push, the problem is that the rocket has a fairly wide nozzle. On a 5 meter diameter rocket they get packed close together and its difficult to put fairings around them. 3 engine mount is on a F4 fuel tank (5 meters) and there is a modified 38 stack decoupler that goes under. I generally give it the roscosmos treatment  with structs  and no shroud (swinging in the breeze all the way up).

 

[regex]

2 hours ago, PB666 said:

at the beginning of the burn would have a TWR of 0.198.

And at the end would have about 0.34.

2 hours ago, PB666 said:

So in terms of burning to orbit, its could only burn for the last 1500 horizontal dV and do so rather inefficiently because of a huge vertical component.

The 5m DCSS has a single RL-10b-2 with some 2.77km/s of delta-V when pushing 28.79 tons and burns for 1125 seconds. For some reason I don't really think ULA/Boeing is going to waste 1.27km/s of delta-V putting something that massive into orbit.

2 hours ago, PB666 said:

Which . . . .is what I was saying.

I have no idea what you're saying. Clearly ULA knows how to use a single RL-10b-2 to get a payload to orbit efficiently or they would have searched for a better solution for their upper stage.

[PB666]

Just now, regex said:

And at the end would have about 0.34.

The 5m DCSS has a single RL-10b-2 with some 2.77km/s of delta-V when pushing 28.79 tons and burns for 1125 seconds. For some reason I don't really think ULA/Boeing is going to waste 1.27km/s of delta-V putting something that massive into orbit.

I have no idea what you're saying. Clearly ULA knows how to use a single RL-10b-2 to get a payload to orbit efficiently or they would have searched for a better solution for their upper stage.

There are two other smaller options, you can use half the fuel but you would not reach orbit or the circularization burn starts at a higher velocity, you could use 70% of the fuel but your rocket would still be sitting pretty vertical for a long period of time. You could burn to a higher apogee and then allow the pitched burn to bring the orbit back to 200.

As a circularization engine 28t payload to LEO200k is about the limit of RL10b-2s performance.

[regex]

4 minutes ago, PB666 said:

As a circularization engine 28t payload to LEO200k is about the limit of RL10b-2s performance.

That's not what you said though:

13 hours ago, PB666 said:

RL10b-2 is not a final burn to orbit engine either.

^ This is what prompted the argument, especially given clear, IRL evidence against that statement. I have no doubts the RD-150 as given in the design we discussed in another thread would serve just as admirably.

Edited December 11, 2017 by regex

[PB666]

Anything can be a final burn to orbit, including an ION drive, however it silly to devote a whole stage to 1k m/s of dV (out of a total of 9500) just because its the most efficient engine you have, particularly when its burning 35' to prograde most of the time. Im am testing their model right now, shot up to 250 km and now correcting with an off angle RL10b-2 thats overloaded back to 200 because of the radial velocity. If you want to argue thats good, go ahead, I design rockets to get Kt in orbit, not 28k limping to orbit on its last stage to orbit.

Let me clarify this so I don't get a pedantic response. In an orbit where the target orbital velocity is 7788 m/s an final stage to orbit can be any size, including 1 m/s dV if the previous stage leaves you at 7787 m/s and all you need it one. In terms of limit, it is literally the limit as thrust decreases the previous stage velocity at a given that an engine can place a payload in orbit. This is why we talk about burning vertically, you can burn your rocket vertically to keep up long enough to lose some fuel and gain a fraction of that thrust in horizontal dV, but that is unwise, the engine should be more powerful or the fuel should be smaller, if the fuel is to small to complete the task then the problem really is the engine. An ION drive in a 200 x 140 orbit can, if more powerful that the force of drag push a payload to orbit, but that is not a wise use of an ION drive since the time it takes to get the craft out of the muck it probably could push the payload out to Mars. Once the engine falls below the ability to push the craft to a viable periapsis and increase that periapsis it has reached its limit.

Maybe ULA doesn't want to redesign a stage and things its just cheaper to Jimmy it into orbit with non-optimal equipment.

Edited December 11, 2017 by PB666