Rocket ascent profile (again)

[Warzouz]

I'm reworking my launchers and as I tested them, I always have hard time to follow a proper ascent path.

I do a very classic ascent

- Full throttle all the way up

- Starts turning at 60m/s

- Cross 45° at 10km

- continue to flat it as much as possible.

But on that ascent, I always end at 30°, I can't flat it more, except bruning with AoA much higher than 5° under the prograde vector which looks inefficient. My usual orbit insersion is now 700 to 1000m/s. I think it was much lower in 1.0.2 (400 to 700m/s, IIRC).

Do you think the last part of the flight (after 10km) changes a lot on the dV expense ?

Is really necessary to flat it more than 30° ?

Is is better to let the rocket fly by itself or force it to flat ? (cosinus loss, is that it ?)

Thx for you input.

Edited June 27, 2015 by Warzouz

[SebLavK]

I've found that what really matters is having the proper acceleration while ascending through the thicker part of the atmosphere. Fly too slow, and you'll waste energy fighting against gravity. Too fast and you'll waste it fighting against drag, which will increase exponentially with your speed.

I usually try to keep the G-force meter around 2Gs until I reach 12-14km, while flattening the rocket's trajectory as much as I can without having a big AoA (Angle of Attack, the angle between your attitude and your prograde marker).

So, to answer your questions

Do you think the last part of the flight (after 10km) changes a lot on the dV expense ?

I've noticed that the last part of the flight does matter, and if done efficiently you'll be left with more dV to use on orbital maneuvers.

Is really necessary to flat it more than 30° ?

A fine balance between your attitude/prograde marker and your altitude, flattening your trajectory as you get higher is very important. Yes, it's better to flat it more than 30º, but don't do it too soon or you'll be wasting energy again.

Is is better to let the rocket fly by itself or force it to flat ? (cosinus loss, is that it ?)

You might need to force it a little. Too much and you'll have a high AoA, which will mean you'll have a lot of drag, and even danger of your rocket flipping over. This does not matter in the upper atmosphere so much, if by 25-30km you're not flat, forcing your rocket to flat will save you some fuel.

Basically, getting into orbit can be compared to driving between two opposite points of a square, you can follow the sides (making a 90º turn), or go through the diagonal, covering less distance. This is the difference between burning up, then eastwards when you reach the Ap, or flattening your trajectory as you go, reaching the Ap as you circularise the orbit.

[Warzouz]

Maybe I could try to cross the 45° under 10km, maybe 8km. That would help to flat the trajectory. I'll try that.

Because even if I force he AoA at 5° as soon as I cross 10km/45°, I can't turn the rocket enough

BTW : the rocket is a really heavy (240T Paylod, around 1500T total), but it seems the lighter rockets behaves the same way.

[SebLavK]

Can you provide a picture of the rocket?

[maculator]

I turn a few degrees over right after the start, then arround 10km (~50sec to AP) i start pointing at 45° and when the AP is over 70km and about 60sec ahead of me i throttle a bit back and try to hold it there.

[Rickenbacker]

I turn over a few degrees right off the pad, then leave it. It usually ends up pretty flat by 45K on its own, if it's got a few fins at the back and isn't too bottom-heavy.

[Warzouz]

Can you provide a picture of the rocket?

Here it is

Control comes from reaction wheels (weak for the rocket mass) and the 100% gimbal of the 6 mammoth.

More on the rocket series

http://forum.kerbalspaceprogram.com/threads/123195-Cygnus-family-Recoverable-SSTO-Rockets-from-12-to-240-tons

[Warzouz]

BTW, I forgot to tell how efficient my actual ascent is :

- fuel start = 3650m/s

- fuel left at LKO (before payload separation) = 300 to 400m/s

-> Net fuel to LKO = 3250 to 3400m/s

It's already quite nice, but my ascent seems no natural as I've to force the rocket all the way up.

I'm sure I'm waisting some fuel. I'm trying to evaluate how much I can gain by having a better ascent. If it's over 100m/s, it's worth improoving.

[SebLavK]

How much TWR does that first stage provide? You shouldn't need more than a TWR of 2 at launch pad. If it's bigger than that, you can limit the thrust of the outside mammoths, or use different engines.

I see the wings you use as fins have no control surfaces, those could make a difference. Maybe add some vernor engines as well to the bottom to provide more pitch authority, instead of reaction wheels, which add more weight. The fuel used to power these is negligible compared to the mammoths. Try doing a manual ascent.

I like to keep my orientation pointing just outside of the prograde marker, but I don't usually launch rockets as big as yours. It's ok for the rocket to fight against your control, meaning that you always have to make little corrections.

Starting to turn as soon as lift off usually makes the rocket tip over slowly while it goes up. Play with this, see if it tips over too quickly or to slowly, and help it along the way.

[Nich]

If you have to force the rocket over after 10k that means you are going to fast at 10k. What is your TWR? However going slower at 10k will increase your gravity losses, lowing thrust at 45 will increase your gravity losses. Your best bet for improving the efficiency might be to increase payload weight. or switch 2 or your engines to rhinos and tanks and stage off the mammoths before the rhinos since they will run out of fuel first. Can you change the thrust limiter on 2 engines at once? You could always just toggle the rhinos on/off to maintain the speed you want. Also having the 2 rhinos at the end should greatly increase your dv of the craft as they will be more efficient after 15ish km. you might use more dV but you will have more for the same mass.

Also I do hate how concerned ppl are with dV efficiency's. Payload fraction is a much better measure i.e. you can get to LKO for only 2900 dV but you only have a 1% payload fraction or you can get to LKO using 4700 dV and have a 22% payload fraction.

Edited June 26, 2015 by Nich

[A_name]

OP: I was too lazy to read through the whole thread so maybe this has been discussed before, but from the sound of it you might be suffering from one or both of these issues:

-Too much thrust. Again, this is from persona experience only, but what I have found is that lower thrust rockets tip over more and faster. Throttle down a bit.

-Not enough drag near the bottom of the rocket. I know nothing about physics but from my personal experience more wings on the bottom help the rocket naturally tip over without any control input (after the initial pith-over maneuver, that is).

Also, it seems like you're already doing this, but in case you're not, you should be aiming for a true gravity turn, which requires zero control input after the initial pitch-over maneuver, letting gravity (and drag) do the rest of the turn for you. I turn off SAS entirely. As mentioned above, you can influence the rate at which your rocket continues to turn by adjusting throttle.

Edited June 26, 2015 by A_name

[SebLavK]

Also I do hate how concerned ppl are with dV efficiency's. Payload fraction is a much better measure i.e. you can get to LKO for only 2900 dV but you only have a 1% payload fraction or you can get to LKO using 4700 dV and have a 22% payload fraction.

dV efficiency is measured in the dV you need to get into orbit, but only the dV that the injection stage has, the lifter. You can get any amount of payload into orbit, as long as the lifter's dV is around 2900m/s.

The way you put it, using 4700m/s dV with a .22 payload fraction, you wouldn't be using all that dV... Or at least shouldn't need to

[Jouni]

Your rocket has a very high TWR for an SSTO, probably around 1.5 to 5.0. The launch TWR isn't that bad, but the final TWR is so high that it's practically impossible to do the gravity turn well. Even at low angles, the vertical component of the thrust is high enough that the climb rate keeps increasing.

The rocket should work better with 4 Mammoths and a Rhino. The radial stacks with Mammoths should burn out when the gravity turn is well underway (at 30 km?), while the central stack with a Rhino would last long enough to complete the mission. You can of course achieve similar results by throttling down or shutting down some engines, but a rocket that doesn't need so much babysitting is always a better design.

[Warzouz]

I'm currently testing the 240T payload. TWR is 1.49 ASL, and 1.59 VAC. But it seems all rockets behaves the same.

My rockets aren't tipping. They are stable. They can slide a little but that only mean I won't be on a perfect 0° orbit. I'll have few m/s.

First test : turning hard at 50m/s and cross 45° at 8km. Ended at 15°. 570m/s insertion burn -> 500m/s left to LKO (I even miss 75km and ended at 79km). It means I sent a 240T payload for a 3150m/s. My gravity turn was better, I didn't need to force the rocket.

Second test : turning harder at launchpad and cross 45° at 5km, ended at 5°, 240m/s burn -> 530m/s left to LKO (76km). It's technically a 3100m/s (as my initial dV is 3630m/s VAC)!

Please note that my test payload is quite streamlined.

EDIT : I increased the payload from 240T to 300T : Success to LKO

Edited June 26, 2015 by Warzouz

[kzauner]

I'm reworking my launchers and as I tested them, I always have hard time to follow a proper ascent path.

I do a very classic ascent

- Full throttle all the way up

- Starts turning at 60m/s

- Cross 45° at 10km

- continue to flat it as much as possible.

Keep an Eye on the VSI. It is OK, when your get off the ground fast, but you should get your climb-rate under control. I assume that your Craft maybe has quite an TWR when the tanks empty. If your G-Meter displays more than 3 G you basically have two options: Turn earlier, flatten out more, or throttle down. The above rules are more-or less a general advise, for typical launchers.

Basically what you are observing is that your climb rate is relatively high regarding your horizontal speed. As a rule of thumb: as soon as you have passed transsonic try to keep the ascend-speed somewhere between 200 and 300 m (thats a Time to Apoapsis somewhere in the range of 20 to 35 seconds). The VSI on top has a logarithmic scaled analog readout, then there are also mods like KER or MJ giving our digital display. I personally prefer using thottle

to adjust the climb-rate - once you are used to it, a 'late gravity-kick' can easily circumvented.

With a little practice you can gravity turn by using throttle alone (after the kick), only correcting heading with steering-inputs. If you find yourself mainly correcting the angle with steering input, than your TWR is off for the current phase of the flight (too high, or too low). In your case - IMO to high. I assume that somewhere within you flight path a throttle-down to 2/3 for about 10-20 seconds would solve your problems; probably somewhere between 10 and 20k as most are going supersonic there.

[ajburges]

As long as you don't exceed a 55 km Ap before 30km altitude, just cut/reduce throttle until prograde is where you want it. The only losses you suffer is some of the Oberth effect since the coast will have you do more of your burn at a lower velocity. That velocity loss likely has less of an effect on energy efficiency than cosine losses.

Efficiency of atmospheric launch of a nuanced thing. You need to balance drag, thrust, and mass against each other. Low (or single) stage launchers complicate this by reducing the flexibility in thrust and mass profiles. At what point does engine mass outweigh TRW gains? Faster acceleration gets you out of the air faster reducing dV losses, but needs more fuel per dV. Taking to the extreme, very rapid exit from the soup could require suboptimal burns to orbital velocity. Is the reduced time in the soup worth it?

The mechanism of your gravity turn is gravity applying acceleration towards the surface. This alters the velocity direction via vector addition. The turn rate has an inversely proportional relationship with engine acceleration and a porpotionate relationship with the sine of the angle from vertical of your heading. High TWR designs need an aggressive turn to account for a shorter burn and the reduced work of gravity. OTOH, a shallower accent increases drag losses. Especially since you are flying with greater speed. The important thing is to grain horizontal momentum. That controls your angle from horizon at a given time from Ap.

Try fidiling with your initial pitch mauver while trying for that 45° at 10 km. This part of the flight is where you need acceleration the most. If you exceed 110% atmospheric efficiency getting to a 10 km angle of 45° or more, reduce your engine array for mass savings. Don't use throttling if you have too much power down low. Reduced engine weight reduces other requirements. Be warned, due to the sinusoidal nature of the gravity turn, the greater your pitch-over, the more accurate it needs to be. It could be worthwhile to note both the initial pitch-over and the desired angle a minute later so you can correct before you deviate from flight plan too much.

Disclaimer: I tend to be bad with high TWR gravity turns in atmo. I much prefer space planes. More active control.

Edited June 27, 2015 by ajburges

[Warzouz]

OK, I tested with throttle reduction past 10km. Gain was around 40m/s. I think it's under the measurement precision (too many variables).

Finally for a streamlined payload and TWR around 1.5 VAC, a good ascent is to cross 45° at 8km and flatten to 10° (maybe reducing throttle a bit).

This ascent to LKO is 3150 - 3200m/s