Question about rocket thrust throttling during ascent

[Screeno]

I would like to know, is it better/ more efficient to let the rocket stay at full throttle the entire journey, therefore letting the thrust to weight ratio go higher as the specific impulse increases and fuel drains? or is it better to throttle down as you go higher and faster, to keep TWR at the same level throughout the flight?

[Superfluous J]

You should go as fast as possible without exploding, generally. If you can control the rocket and aren't actively exploding, you're not going fast enough. Well, okay maybe not "not going fast enough" but going faster won't hurt you. Until you explode or your rocket flips.

If you need to throttle down because your rocket's getting squirrely or all the heat gauges are at 90%, then you have too much engine and should redesign the rocket to not be capable of going that fast.

I frequently see flames all over my rocket on launch. I'm also usually mostly sideways and going over 1km/s under 30km. Sometimes, my Pe is over the ground when My Ap gets to 100km, and those are the launches that I feel I launched the best.

[Lucius]

Now someone else will explain it for you better i'm sure, but i can tell you in brief, there is a chart out there somewhere of the TWR of the apollo missions, and they would deliberately cut engines as they ascended for 2 reasons:

1) was to maintain lower g forces on the nauts

2) was honestly to maintain an efficient TWR so that they were following their projected flight path (i believe)

I think how it worked was, if the SV was a 5 engine craft, they would cut the center engine, then 2 side engines so you were, at the end, only running on 2 engines, never going above a 3:1 TWR.

I'm gonna presumptuously go out on a limb here and say if NASA did it, its probably the more "efficient" way to go about it.

Hank Hill Edit: BWaaaaaaaa -> http://www.braeunig.us/apollo/pics/SVsim-fig3.gif

Edited September 2, 2015 by Lucius

[Geher]

Now someone else will explain it for you better i'm sure, but i can tell you in brief, there is a chart out there somewhere of the TWR of the apollo missions, and they would deliberately cut engines as they ascended for 2 reasons:

1) was to maintain lower g forces on the nauts

2) was honestly to maintain an efficient TWR so that they were following their projected flight path (i believe)

I think how it worked was, if the SV was a 5 engine craft, they would cut the center engine, then 2 side engines so you were, at the end, only running on 2 engines, never going above a 3:1 TWR.

I'm gonna presumptuously go out on a limb here and say if NASA did it, its probably the more "efficient" way to go about it.

As far as I know they only cut out the center engine. And the only reason was because of astronaut comfort.

Generally is a higher thrust always more efficient than a lower one (when you ignore the atmosphere). Imagine a rocket which TWR is ALWAYS 1 (even when it gets lighter). It would just hover above the ground burning all its fuel and never getting anywhere.

[Lucius]

Yes sorry i peeked through the diagram myself and you're right they only ever cut out the center engine.

[Just Jim]

I'm not sure about the efficiency one way or another, but a couple things I might factor into this before gunning the engines.

I've found after 1.0.4 came out I have much better luck cutting back the throttle to keep my rocket under 300 m/s until I'm out of the lower atmosphere to keep the drag from sending my ship into a death flip. Drag can be a nightmare...

And after I'm in space I seldom throttle up past 2/3rds, but this again isn't for efficiency, it's to prevent any possible overheating on longer burns.

My point is, full throttle may be more efficient, but in certain situations it can also get you killed.

[Clipperride]

I haven't done the comparison flights to test this but I (admittedly a new player) thought that seeing atmospheric effects meant that you were going too fast and fighting drag therefore using more fuel. As soon as I got liquid fuel engines I've kept my speed under 300m/s below 10,000m and not breaking the 1.5 km/s till 30,000 m. I can see pen, paper and test rocket in my immediate future :-). I may have been doing things totally wrong.

Lucius - I'm a bit of an Apollo fan. You're right about the centre engine cutting out first, followed by the outer 4. I believe it was, as you say, to reduce the G force on the astronauts as well as the load on the vechile when it "unloaded" the forces built up during acceleration. The engines didn't have a throttle as such, but the ratio of fuel/ox changed during ascent. It was at ignition that the engines started in opposing pairs, again to reduce the forces involved in lighting such a monster engine.

[Screeno]

My launchers are always stable, I've never had a flip problem or overheat, maybe I was playing it to safe with the throttle, so perhaps I should just let it go full power. I'll see which ends up with more dv in orbit

[Nich]

http://forum.kerbalspaceprogram.com/threads/128509-Goddard-Problem-Maximal-launch-altitude-Challenge

600 m/s is the first speed limit at about 6000 m. This increases as you get higher. If you hitting are these speeds your twr is probably too high.

[Geher]

You can also install a mod called Mechjeb. It's Ascent Guidance has a checkbox which says Limit to terminal velocity. When this is checked you can always fully throttle up and when it would be more efficient Mechjeb will automatically throttle back. But for me this rarely ever happens, because my rockets mostly have a SLT between 1.1 and 1.6

[ForScience6686]

Full throttle all the way or you are wasting mass with engines you don't need. I have not had flipping issues even on my rapid rescue which can launch and rescue from lko in under 25 minutes. This thing is glowing by the time I hit 20km, and is not the most aerodynamic as it holds a glider in a big fairing on a short rocket. If you are having flipping issues, it's your lack of fins or an improper aoa, not your engines.

[cantab]

Since 1.0 added realistic aerodynamics you are generally OK to go full throttle all the way. The exception is if you have excessive liftoff TWR and don't want to remove engines for whatever reason. If you are having stability issues you're arguably better off improving your rocket's aerodynamics, a few fins at the back are a good start.

[Red Iron Crown]

Throttle pinned until I am fine tuning the apoapsis. If it burns up I either did the gravity turn poorly or the design needs revision.

[Stoney3K]

My usual strategy is to keep a constant acceleration of about 15m/s until I'm out of the lower atmosphere, because accelerating any faster means you're only fighting drag, and wasting fuel in the engine's inefficient atmospheric regime.

Also, with a low acceleration in your spacecraft's direction of movement, it's easier to turn. A high acceleration means you need more engine gimbal to change direction.

From 23km up, you can turn (surface)prograde and slam the throttle flat out.

[Pecan]

My usual strategy is to keep a constant acceleration of about 15m/s until I'm out of the lower atmosphere, because accelerating any faster means you're only fighting drag...

Sounds slow; don't forget all the time you're trying to gain height you're fighting gravity, and gravity always wins.

...thought that seeing atmospheric effects meant that you were going too fast and fighting drag therefore using more fuel...

...Limit to terminal velocity...

As cantab said above, that advice is outdated. Terminal velocity is so high now that you're unlikely to be able to reach it unless you are only climbing very slowly but accelerating quickly horizontally.

Full-throttle or you're carrying more engines than you need. Stage them off or get them working!

On Apollo: note that they cut the centre engine one minute before jettisoning the whole first stage. Given how much more complex staging is in real-life than KSP, think of the extra work it would have taken to just jettison that one engine, better to hang on to it for a minute even if it isn't helping.

I've now seen a few people though who are saying they are getting good results with an almost-continuous low-thrust circularisation burn while climbing to Ap instead of cutting throttle and cruising to it. With this technique Pe is already above-ground, orbit possibly almost circular, at an altitude of 50km which leaves very little for the final circularisation at Ap. Whether this does indeed require less dV and, if so, what TWR regime is best, is not something I've investigated much because MJ does most of my launches these days, I'm busy.

[ETA: and I MJ because I test my designs by hand first, and hand-over once they launch well enough to autopilot boringly well]

Edited September 2, 2015 by Pecan

[cantab]

I don't understand how on earth you guys are launching anything AND gravity turning with the throttle pinned.

My generic ascent profile: Vertical until 100 m/s. Pitchover to somewhere between 75 and 85 degrees, depending on the rocket. Let the prograde marker catch up with the heading. Then follow prograde until I'm up to at least 30 km. Provided I hold it right on that prograde marker once I'm at any appreciable speed even a mildly unstable rocket will fly fine. A severely unstable rocket with an oversized payload up top will still flip, but I just try not to build rockets like that!

And throttling down is also wasting fuel. If you're ascending vertically with 2.0 TWR then for every 1 unit of fuel that accelerates you you waste 1 unit fighting gravity. Throttle back to 1.25 TWR and suddenly for every 1 unit of fuel that accelerates you you wast 4 units fighting gravity.

(As for why not have a huge liftoff TWR, partly because carrying excess weight in engines is another way to waste fuel, and partly because is still such a thing as too much speed too low it's just much faster than in the .90 days.)

Edited September 2, 2015 by cantab

[jwenting]

I tend to throttle back to stay outside of sonic shock or heating effects.

Also, some rockets have a tendency to flip if you empty the tanks too fast, so rising more slowly keeps them stable.

[Kelderek]

Don't forget that you can also thrust limit your engines and then still go full throttle. When designing your rockets it is not always a perfect fit to go from one engine type to another, thus if a particular engine is slightly too powerful, then you can thrust limit it in the VAB, but then go full throttle in flight for a smooth ride. In general though, I think if you are thrust limiting an engine down below 80% then you are probably using the wrong engine, but above that can be useful to ensure that you don't go too fast too low in the atmosphere.

A good example is if you use an orange tank + mainsail and your TWR is too low, so you switch to a twin boar for roughly the same mass but now you might have a little too much TWR, so you thrust limit it a little.

[severedsolo]

I ... thought that seeing atmospheric effects meant that you were going too fast and fighting drag therefore using more fuel. As soon as I got liquid fuel engines I've kept my speed under 300m/s below 10,000m and not breaking the 1.5 km/s till 30,000 m. I can see pen, paper and test rocket in my immediate future :-). I may have been doing things totally wrong.

In previous versions you would be correct. If you are seeing mach effects, you are too fast. In my experience though, in 1.0.4 mach effects aren't too bad. Actually, if you're seeing them and not flipping out/overheating, you're probably around about where you need to be TWR wise.

I also agree with 5th, on the really efficient launches (the ones where I'm almost horizontal at 36k, when the navball switches over to orbital mode) - heating effects are the norm.

[cantab]

I think we're confused on some terms, tbh... A gravity turn requires almost no actual steering, if it's done just right. Honestly, it sounds like you're not turning enough... guideline being around 45 degrees before 15km, and 10-20 degrees by 35km or so.

Those pitch attitudes at those altitudes are about what I get when I do a good ascent. A gravity turn still needs an initial pitchover to get it started, unless you opt to tilt the rocket on the launchpad.

Also, rocket-flipping comes from not enough weight up top, rather than too much. Think of throwing a lawn dart backwards.

Quite, hence why I spoke of the problem being oversize payloads not overweight payloads. I try and ensure my fairings are flush with the stage below, or at most slightly wider - if they're bulging out a lot that's asking for trouble.

I'm never ascending vertically beyond 100m/s or so (during which time throttle is pinned). After 100m/s, the throttle has to be moderated to allow the rocket to turn without flipping - it's sort of a balancing act. Drag and gravity are both pulling the nose down - too much drag (created by speed) over balances it onto its head (rocket flip).

In angled flight the consideration about losing fuel fighting gravity still applies, but to the vertical component of your thrust.

But I just don't find my rockets as flip-prone as yours seem to be. Like I said, provided I follow that prograde marker above about 150-200 m/s I'm fine. To which I can suggest a few explanations:

1) Newstock aero makes the rockets too flippy. I use FAR myself. But I thought newstock was supposed to be more forgiving than FAR, not less.

2) You're starting off with too much TWR, which lets you hit problematic speeds lower down in thicker atmosphere. I usually design for 1.3-1.4 TWR at launch.

3) Your rockets are unaerodynamic, which I was talking about earlier. I really do think that if you can sort the aerodynamics out so you can go full throttle you will get better performance.

4) You have lousy control authority. Likely if you use solids a lot, but then you wouldn't be throttling down.

[ForScience6686]

I don't understand how on earth you guys are launching anything AND gravity turning with the throttle pinned. For the majority of the game, you are limited to a few possible engines for a given payload size, unless you duplicate small ones (expensive), or throttle down an overpowered one. If I happen to hit JUST the right TWR, I'll pin the throttle, but it's always based on the speed in lower atmo. Generally 250 m/s below 10k m, 550 until 35k, then floor it. Anything faster just leads to flips with even the simplest rocket designs. This is probably due to the capsules/payloads being too light for the rocket, but such is KSP.

Yes, I'm aware of how to use fins to help, but these are essentially putting a brake on the entire rocket, wasting fuel, when you could just as easily slow down. At most, I'll slap a few basic rocket fins at the bottom. No, I do not MechJeb, nor will I ever

I always go to throttle otherwise you're packing unnecessary engine weight. Fins add control authority as well if you use the right ones, and why worry about the added drag if you are throttling down anyways then you have the power to compensate. In real life, they have to be cautious about the effects of g forces on people and equipment, we don't need to worry about that. The way I look at it, my time is more important than fuel efficiency, and why go slow and spend more time in the thick atmosphere? I also usually use srb first to gain altitude before igniting main engine for better ISP. My gravity turns also start immediately off the pad.

[Clipperride]

So many methods for getting into orbit. As soon as time permits I'll try some out to see which leaves the most fuel. As this is my first game of KSP and I'm playing science mode (money? Who's that?) I may just be really over engineering my launcher's.

[xtoro]

I keep below 300 under 10k, after that it's full throttle. If my AP is where I want it and I'm getting close to 1900m/s orbital, I throttle back to save fuel. When burning at AP I'm not fighting gravity anymore so there's no sense it wasting vertical dV at that point.

[MarvinKitFox]

About the only time throttling down would be a good idea is if your rocket is very close to the speed of sound, and you have a low-ish TWR.

Then it is more efficient to dawdle around at mach 0.90 for a bit before accelerating again to pass the trans-sonic region, rather than bulling your way through mach 1.0 with too low surplus thrust.

With planes, this is quite noticeable.

With rockets.... not so much

[ajburges]

First of gravity losses:

Mathematically, you are in a orbit while landed. It is just an elliptic orbit with you getting 1g of radial acceleration at Ap (orbital velocity equal to surface). Gravity losses are off prograde thrust (also known as cosine losses) and thrust inefficiencies when not burning at Pe (Oberth effect) or Ap (Kepler observations).

All you need to do to minimize gravity losses is to efficiently transfer from that elliptic (sub-)orbit to a higher, circular one. The complication is that you have atmosphere and a relatively low TWR craft for the dV the manuver needs.

Low atmo is full throttle unless your aerodynamics are a drag. Drag will sap around 600 m/s on a good rocket ascent! That's more than any reasonable cosine loss. When you GTFO of the low atmo like that, you tend to be on an overly high gravity turn trajectory. You want to reduce TWR so that your trajectory allows you to circularize your orbit without losing much speed from coasting. The best way is to stage to a lower TWR configuration, but SSTOs are stuck reducing throttle.

My most efficient SSTO to date is a nuclear space plane. It has only enough oxidizer for the RAPIERs to bring Pe to approximately -200 km. That kick is just enough for the .4 TWR nuke to circularize orbit by the time it reaches Ap (60 km) while burning at full throttle (at which point the low TWR is efficient)