Shaking off the rust, trying to get into orbit

[Streetwind]

@Landwalker Alright, I managed to noob my way through recording video. The quality isn't the greatest, probably due to blindly guessing recording settings... but the UI text is still readable, so it should serve its purpose. You can get it from my Dropbox here. (rightclick the link -> save if you don't want the web player screwing up quality even further)

It's a fairly aggressive ascent, in part thanks to the way the TWR of this craft ramps up sharply as fuel is burned. A two-stage craft would have a slightly steeper initial ascent, a longer second stage burn, and less of a coast phase at the end. Also, you'll see what I mean with this being an exercise in manual steering. Launch vehicles can be built so that they gravity turn on their own, as @Spricigo says; this is intentionally not such a vehicle. Mind you, if the pilot is able to use SAS Hold Prograde, even this craft becomes very comfortable to fly.

I had 478 m/s vacuum dV left on this attempt, out of an initial 3,739 m/s vacuum dV. Means I achieved orbit using 3,261 m/s. Plenty of margin for you, I'd think

Edited September 25, 2017 by Streetwind

[Spricigo]

53 minutes ago, capi3101 said:

With stock air, a good launch TWR is 1.5 or so (at least it was last I checked; the old 1.6-1.7 rule may still apply)

You mean this thread:

I'd point out that drag losses itself are not that a big deal people tend to assume. 100-200m/s if you can keep the rocket pointed close to prograde is about a common ballpark (even for some more powerfull rocket) .Deviating from prograde tend to be a more serious problem.

[MajorMushroom]

45 minutes ago, Spricigo said:

the issue he is facing is not that he is not reaching orbit but that he is using too much fuel to reach orbit.

That method would makes his problem worse.

Also, I strongly recommend that you stop using it immediately since you can be much more efficient even with a badly performed gravity turn.

 

Sorry, but the idea that a gravity turn is in some form worrisome to pull out is just wrong.  I can tell from my experience , doing a reasonable gravity turn is not only easy but trivial to perform.

About 60% of my gravity turn are done like: 1. press space(to launch) 2. press space(to drop booster) 3.wait for apoapsis

The rest are like: 1. press space(to launch) 2.set SAS to follow prograde 3. press space(to drop booster) 4.wait for apoapsis

That's it, 2-3 key-presses/clicks and you go to orbit for less than 3,5km/s.

It can be made even more efficient with some extra care for details, but most of the advantage can be achieved with a 'lazy' gravity turn like that. The only thing that is missing in your firsts attempt are the knowledge to know how much of a initial turn is required (/muscular memory for those that prefer the 'nudge' method.)

I actually tried out a turn, and it worked well. It wasn't as frustrating as I expected. I guess because before I used some oddly shaped rockets which are a bit clunky. 

With a bit of work though I'll get better  sorry to be a bother!

[Spricigo]

 

8 minutes ago, MajorMushroom said:

  sorry to be a bother!

 

There is nothing to apologize.

The idea of this forum is exactly to share our collective knowledge. Sometimes we can explain the bits we know better, other times we learn the bit we didn't grasped very well yet.

 

Mind you that the way I write is a bit blunt at times but by no means meant to be aggressive.

[Kryxal]

The funny thing is, if you were to add some more fuel to stage 3, this would likely be a very nice craft to do a Mun flyby ... or maybe even orbit and return.

[Zeiss Ikon]

21 hours ago, Landwalker said:

But the second stage basically came cartwheeling out of the decoupler with nothing to stabilize it—the same problem I ran into originally—and ruined the whole launch.  This occurred multiple times, so I have no doubt that there's a fundamental flaw in the second stage.

Okay, I'm going to ask a question along the lines of "did you put gas in the tank" -- please, don't take it personally, every one of us has forgotten parachutes, ladders, decouplers, and heat shields on multiple occasions.  But: are your probe controller and reaction wheel both in the upper stage, or are one or both staging away with the booster?

[Landwalker]

Man, I've got a lot of comments to catch up on.  Here we go.

 

5 hours ago, Streetwind said:

@Landwalker Alright, I managed to noob my way through recording video. The quality isn't the greatest, probably due to blindly guessing recording settings... but the UI text is still readable, so it should serve its purpose. You can get it from my Dropbox here. (rightclick the link -> save if you don't want the web player screwing up quality even further)

It's a fairly aggressive ascent, in part thanks to the way the TWR of this craft ramps up sharply as fuel is burned. A two-stage craft would have a slightly steeper initial ascent, a longer second stage burn, and less of a coast phase at the end. Also, you'll see what I mean with this being an exercise in manual steering. Launch vehicles can be built so that they gravity turn on their own, as @Spricigo says; this is intentionally not such a vehicle. Mind you, if the pilot is able to use SAS Hold Prograde, even this craft becomes very comfortable to fly.

I had 478 m/s vacuum dV left on this attempt, out of an initial 3,739 m/s vacuum dV. Means I achieved orbit using 3,261 m/s. Plenty of margin for you, I'd think

This was great.  Thanks a ton for doing this!  I noticed some big differences between your profile and mine—for example, by the time you reach 10km, you're going twice as fast as I am and are about half the angle to the horizon that I am. 

It seems like from there, you close in on the horizon angle more slowly than I do.  I assume that's a function of going much faster much earlier "holding" your angle better, where as I tend to be going slower throughout the ascent.  I've been driving my rockets more like a freight train.

For practice, and because I learn best through imitation, I slapped together the same craft to see how I could do trying to recreate your launch profile.  On my first attempt, I came in too steep (again), but still managed to end up getting things circularish (102x94km) with around 200m/s left.  For some reason (I think related to RealChute altering the mass of the nose parachute), my start dV was a little lower than yours, but I still chewed through probably 3,500m/s of dV to get up there.  Which is still a huge improvement over where I was at, so that's a good starting point for further practice.

5 hours ago, capi3101 said:

Okay...taking a look at the Circe, and some of the advice you've already received on this thread.

Just at a glance, @HvP is spot-on when he pointed out you're not in atmo mode on KER (the button that says "Atmospheric"). Turn it on and leave it on - KER these days is smart enough to know which bodies have atmo and which ones don't. You may find that you don't have nearly as much delta-V as you think you have.

Yeah, that was my initial inclination when I was first running into problems.  But as I mentioned before, the dV map I'm referring to is clear that its figures are vacuum dV, so my understanding is that if a craft on the launchpad has a hypothetical vacuum dV of (let's say) 3,800m/s, that should be more than enough to get into orbit and get back down.

Still, I think a lot of my problem is boiling down to "I'm going too slow and spending too much time at low altitudes."  So I end up wasting a lot of potential vacuum dV by dickering around with things like "gravity." 

5 hours ago, capi3101 said:

The launch TWR on that craft is nuts; you're going to lose a lot of your first-stage delta-V to drag. I use FAR; with FAR, a good launch TWR is 1.2-1.3. With stock air, a good launch TWR is 1.5 or so (at least it was last I checked; the old 1.6-1.7 rule may still apply). As you're flying along, watch your gee meter - if it climbs out of the green at any point, that's a cue to throttle the engine back. You don't need to change out the engine - just play with the throttle limiter until you get a TWR in that range. You can always swap out the engine later if you need to, and you can always adjust the throttle limiter more in flight if you feel it necessary to do so.

The launch TWR in that readout is deceptive, because it's assuming the Reliant will be firing at full power—which is not how I was operating it.  With the Circe, I was launching just on the SRBs, and only engaging the main LFO engine when the SRBs ran out.  I just happen to put my main LFO in the same stage as the SRBs as a "just in case" option just in case I decide I need to engage it before they finish burning out.  The actual launch TWR, with just the SRBs, was 2.16.

5 hours ago, capi3101 said:

The second-stage TWR is not high enough; KER is telling you that much. You've got a 25 kN engine trying to push a 3.5 tonne craft when that stage lights - a TWR of 0.7 at that point. You're losing delta-V to gravity at that point. Your second craft has rectified this somewhat, but you're saying it's giving you control headaches. Would you mind taking a screenie of just that stage with your control balls on (CoM, CoL, CoT)? Wagering your CoM is too far aft on that stage. TWR is still a tad on the low side for that stage of the ascent.

Yeah, I'm thinking that the Pug is too weak for what it's being asked to do.  The upper stage is too heavy for it to handle when gravity is still at issue.

 

For giggles, I rebuilt Circe by replacing the Reliant with a Swivel and losing the SRBs:

Spoiler

Unfortunately, this tended to get pretty unruly when I tried to turn it.  I slapped some basic fins on the back and it handled a little better, but still got kind of wacky after a bit—I got the nose pointed too far away from prograde and it just started fish-tailing.

I rebuilt my second craft from earlier in the thread, as requested, and here are its control balls:

Spoiler

 

Whew, lots of stuff.  With that out of the way, it's time to get back into practicing.

 

 

 

6 minutes ago, Zeiss Ikon said:

Okay, I'm going to ask a question along the lines of "did you put gas in the tank" -- please, don't take it personally, every one of us has forgotten parachutes, ladders, decouplers, and heat shields on multiple occasions.  But: are your probe controller and reaction wheel both in the upper stage, or are one or both staging away with the booster?

Yes, the probe core and reaction wheel are both in the uppermost stage.  Don't worry, I've made that mistake with the reaction wheel before...

Edited September 25, 2017 by Landwalker

[Landwalker]

After some practice, trial, error, and blatant imitation, I've been able to mostly-replicated Streetwind's launch profile.  Not as efficiently—I seem to be between 3,400 and 3,500m/s of dV burn, which isn't exactly 3,261m/s, but it also isn't 4,200m/s, so there's that.

The next challenge, of course, is taking my launch profile knowledge and building a rocket capable of actually, you know... following it without flipping out (figuratively and literally).

[Palaceviking]

Your col is in front of your com in your second stage and will get further forward as you burn fuel which creates a lever arm around your com.

[Landwalker]

42 minutes ago, Palaceviking said:

Your col is in front of your com in your second stage and will get further forward as you burn fuel which creates a lever arm around your com.

Okay... So how do I make that not be a thing?

It also doesn't explain why this rocket also has a tendency to reach a certain point and then cartwheel uncontrollably:

Spoiler

My assumption when I had trouble with this particular design was similar to the issue you point out:  As it burns fuel, the CoM moves further back until eventually it catapults the back of the rocket forwards as soon as it gets the chance (due to minute deviations from prograde).  But I can only find ways to move the CoM so far forwards on the rocket.

Edited September 26, 2017 by Landwalker

[Landwalker]

Well, after spending the whole evening doing exhaustive testing, I was able to come to a few conclusions:

1. The leading cause for my single-stage craft cartwheeling was not a CoM issue, it was a drag issue.  Namely, I didn't have any, and the rocket was big enough so that if it got even slightly off-center from prograde it could lose control entirely as the air "grabbed" the front of the rocket and threw it backwards.
2. Extensive testing identified the Science Bay parts from DMagic's Orbital Science mod as the culprits.  All designs without these parts did not suffer from cartwheeling unless I really janked up the piloting.
3. Even without those parts, the craft was unstable in a turn (probably a secondary CoM issue).

So my solutions:

Spoiler

1. Per the suggestion of @Aegolius13, rather than adding stabilizing fins / winglets to the rocket, I added Elevon 0's.  This immediately solved my control problems on the "no-science" craft design and allowed me to basically waltz into an orbit for a hair over 3,300m/s.
2. Even with the elevons, the "yes-science" craft suffered from very fussy controls.  It handled better, but slight errors could doom it.  As a result, I encased the "probe stage" in fairings in an attempt to hide it from the janky aerodynamics.  This worked marveously.

Spoiler

There were probably more efficiencies to be had, but that particular orbit weighed in at exactly 3,350m/s dV.  Relative to Streetwind's launch, that was about 97% efficiency, which is well within my "acceptable margin" range.

Edited September 26, 2017 by Landwalker

[Streetwind]

8 hours ago, Landwalker said:

This was great.  Thanks a ton for doing this!  I noticed some big differences between your profile and mine—for example, by the time you reach 10km, you're going twice as fast as I am and are about half the angle to the horizon that I am. 

It seems like from there, you close in on the horizon angle more slowly than I do.  I assume that's a function of going much faster much earlier "holding" your angle better, where as I tend to be going slower throughout the ascent.  I've been driving my rockets more like a freight train.

Yes. Always remember that going to space is not about gaining altitude - it's about going sideways so fast that you miss the ground every time you fall back towards it. (This is the layman's definition of being in orbit.) Altitude is only useful for two things: one, avoiding the atmosphere; two, giving you time to accelerate sideways before you hit the ground again.

This is why I tilt over so aggressively. I know this craft has lots of thrust, so I know I can get away with it. Even when turning over sharply, a part of that large amount of thrust is still providing a vertical component. The higher my thrust output, the less of an angle I need to have "enough" vertical component to keep me from falling back down. Additionally, the higher my thrust output, the less time I need to achieve orbital velocity - the amount of speed I need to miss the ground once in freefall. That, in turn, means I need even less vertical component, and can turn over even harder. This is also why the rocket "holds its angle", as you put it.

You of course cannot see my eyes in that video, but they constantly go back and forth between five things: the navball, the speed readout, the altimeter, the current apopasis height, and the time to apoapsis. These are the five main things I use to judge how to steer at any given moment. For most of the flight, as long as time to apoapsis is stable or rising slowly between around 30 to around 45 seconds, I know that my flight profile is good and I can keep turning over at my current rate. If it is rising quickly, I know I made a mistake and the ascent isn't going to be very good; if it is decreasing, I screwed up and may in fact fall out of the sky unless I can find a way to reverse the trend through high thrust and angling upwards. Towards the end, I allow time to apoapsis to run away from me, because I know my apoapsis altitude is approaching 70 km, my craft's speed is approaching 2 km/s, and my craft's current altitude isn't very high yet. This is expected of the aggressive single-stage launch profile. I put my apoapsis where I want it, then cut the engines and coast until I can make a moderate orbit insertion burn of around 400-500 m/s at apoapsis.

In a more conventional two-stage launcher, the first stage would push the time to apoapsis away a bit (to say, 50-60 seconds), and then the lower-thrust upper stage would see time to apoapsis decrease during its burn. An ideal launch profile would see your craft's time to apoapsis decrease to exactly 0 seconds at exactly the time you achieve orbital velocity at exactly your target altitude, using a single continuous burn. But this is pretty much only achievable by computer guidance. A human can come close, with experience and a well-designed launch vehicle that the human has practised with often. But most of the time you will end up with leftover time to apoapsis during your upper stage burn, and have a small coast phase before making an orbital insertion burn. It's the safer way to do it, and assuming your coast phase is short and the insertion burn small, the loss in efficiency is negligible.

(This means you can use the magnitude of your orbital insertion burn as another indicator of how well you flew and how well your rocket is designed. The smaller, the better!)

 

4 hours ago, Landwalker said:

The leading cause for my single-stage craft cartwheeling was not a CoM issue, it was a drag issue.  Namely, I didn't have any, and the rocket was big enough so that if it got even slightly off-center from prograde it could lose control entirely as the air "grabbed" the front of the rocket and threw it backwards.

Well, it is both a CoM and a drag issue.  

The center of mass can be thought of as the axis around which the craft turns. The center of pressure is the location of the compounded average of aerodynamic force application. Now, it is important to remember that as long as a rocket flies straight prograde, all of the aerodynamic force is on the nose, and the rest of the rocket receives zero force. Only when the craft deviates from prograde - meaning the flanks of the rocket dip outside the nose's slipstream - does the center of pressure actually come into play.

If the center of pressure is behind the center of mass, that means that anytime the the rocket deviates from prograde, the force exerted on it by aerodynamics creates a lever arm around the center of mass that instantly pushes the rear back into the slipstream of the nose. The rocket is naturally self-stabilizing, like an arrow.

If the center of pressure is in front of the center of mass, the situation is reversed. Anytime the rocket deviates from prograde, a lever arm is created that tries to twist the rocket further away from prograde. The craft is naturally self-destabilizing, seeking to reverse itself in midair, like trying to fire an arrow with the fletching pointing forwards.

 

Thus is the golden rule of rocketry: "heavy parts at the front, draggy parts at the rear".  

 

And yes, this can be hard to pull off, because you generally have payload at the front, and that payload tends to be less dense than pure fuel. The easiest solution to this issue are fins at the bottom of the launch stage. But in general, it helps if you build tall, narrow rockets instead of flat pancakes. The mental image of an arrow is really helpful, as it exemplifies the ideal mass and drag distribution you want to achieve: a heavy head right at the front, and the draggy fletching right at the back, connected by a long and narrow shaft.

[Pand5461]

@Landwalker glad you figured out how to fight your problem.

As a side note, I would like to point at the few things you can learn from the @Streetwind's video.

  1) The ideal on-the-pad TWR is usually about 1.5-1.7. Real-life designs usually have it closer to 1.2-1.3 but that's because mass difference between fully fueled rocket and rocket at lower stage burnout IRL is much greater than in KSP, so that pad TWR of 1.5 would result in unacceptable g-loads midflight.

  2) Start turning as soon as you reach TWR > 1.2, meaning you can start pitching right off the pad with 1.5.

  3) Aero effect animations are misleading. You may have pretty impressive plasma effects while having in fact very little drag, so don't be afraid to go red. The overheat gauges is when you need to start worrying.

  4) You absolutely don't need to throttle down before reaching TWR of 3 (and maybe even more if you're pretty high in atmosphere at that point). Trying to keep constant TWR is just going to increase gravity losses. If you're worried about high vertical speed, pitch down more aggressively. You don't need to throttle down at all while at vacuum (other than for the sake of precision).

  5) Cutting off engine, coasting to apoapsis and circularizing at full thrust is more efficient than a long low-thrust burn.

And a couple more things regarding multistage rockets.

  6) CoM shifts up as you burn the fuel, so first stage generally becomes more stable as time goes. When you stage, positions of CoM and CoL change very suddenly and that may cause instability. Therefore, it's better to turn surface prograde before staging, especially in low atmosphere. And try not to design things that need staging at 300 - 600 m/s (the typical maxQ region).

  7) Do not jettison anything in unpowered atmospheric flight (e.g. when you coast to the apoapsis). Lower stage or a fairing don't affect the cross-section (and drag force) much but increase mass, hence with them attached you get lower acceleration (or, rather, deceleration) from drag.

And a note on this thing

16 hours ago, Spricigo said:

I'd point out that drag losses itself are not that a big deal people tend to assume. 100-200m/s if you can keep the rocket pointed close to prograde is about a common ballpark (even for some more powerfull rocket) .Deviating from prograde tend to be a more serious problem.

Some may infer that this is untrue because Tylo and Eve have massively different dV to orbit while having similar size and gravity with Kerbin. But don't fall to that common misconception, drag losses on Kerbin are truly that low. The difference is mainly to the fact that you can't follow a good launch profile on atmospheric planets and have more gravity loss due to need to go upwards at launch for quite some time (and even worse, you need to do it when the engines have the worst performance).

[Landwalker]

25 minutes ago, Pand5461 said:

And a couple more things regarding multistage rockets.

  6) CoM shifts up as you burn the fuel, so first stage generally becomes more stable as time goes. When you stage, positions of CoM and CoL change very suddenly and that may cause instability. Therefore, it's better to turn surface prograde before staging, especially in low atmosphere. And try not to design things that need staging at 300 - 600 m/s (the typical maxQ region).

  7) Do not jettison anything in unpowered atmospheric flight (e.g. when you coast to the apoapsis). Lower stage or a fairing don't affect the cross-section (and drag force) much but increase mass, hence with them attached you get lower acceleration (or, rather, deceleration) from drag.

Yeah, one of the things I concluded (but have not yet tested) from last night's experiments was that if I'm multi-staging to LKO, as I was originally trying to, I need to have a much smaller and later-activated upper stage that's primarily for the circularization and de-orbit burns, and a lifting stage capable of shoving that orbital stage at least close to the desired apoapsis (and therefore needing to burn longer than it was in my original design).  A scheme for testing tonight, perhaps...

[Streetwind]

55 minutes ago, Pand5461 said:

  1) The ideal on-the-pad TWR is usually about 1.5-1.7. Real-life designs usually have it closer to 1.2-1.3 but that's because mass difference between fully fueled rocket and rocket at lower stage burnout IRL is much greater than in KSP, so that pad TWR of 1.5 would result in unacceptable g-loads midflight.

*snip*

  4) You absolutely don't need to throttle down before reaching TWR of 3 (and maybe even more if you're pretty high in atmosphere at that point). Trying to keep constant TWR is just going to increase gravity losses. If you're worried about high vertical speed, pitch down more aggressively. You don't need to throttle down at all while at vacuum (other than for the sake of precision).

  5) Cutting off engine, coasting to apoapsis and circularizing at full thrust is more efficient than a long low-thrust burn.

I' m not sure I agree 100%. Time to discuss

Plenty of real life rockets have higher TWR than you think. Examples (before adding payload, error ~2%):
- Falcon 9 Block 4: 1.41
- Electron: 1.57
- Vega: 1.68
- Ariane 5 ECA:  1.98 (rumble in the jungle)
- PSLV XL: 2.26 (moar boosters!)

Reason for this large spread is that everyone does things a little differently. Depending on what a launcher is optimized for, how its individual stages look, what it's flight profile is supposed to look like, what technology is available for use and so on and so forth, the correct answer can vary. And often the answer just has to be "close enough" if it hits a sweet spot of price/performance. The PSLV XL even ignites another two additional SRBs 25 seconds into the flight, while all of its liftoff engines are still going full bore! India definitely subscribes to the bat-out-of-hell style of rocketry.

(Though if you ever wanted to know why the PSLV pitches over so crazy hard right off the pad, now you know why! )

 

Regarding point 4, I'd heavily disclaim that advice. Theoretically it may be sound, but in practical application it has little to do with what's really going on. Your throttle setting should be chosen in response to how your actual launch is going. If throttling down will put you on a better trajectory, then absolutely throttle down! (But I agree that there are no aerodynamic reasons to limit TWR to a specific number.)

 

Point 5, could you expand on that? I've been led to believe the exact opposite in the past, and it has worked quite well for me. Additionally, in RL orbital rocketry, coast phases before establishing at least a minimal parking orbit are nearly unheard of, unless you want to count brief moments during staging.

If your claim can be mathematically supported, I would be very interested in seeing it!

Edited September 26, 2017 by Streetwind

[Pand5461]

3 minutes ago, Streetwind said:

Plenty of real life rockets have higher TWR than you think.

And you did not mention Soviet designs.

But concerning the actual advice, it rather supports it than disproves. No need to worry or throttle down if starting TWR is higher than 1.3 or something. More TWR = less gravity losses.

20 minutes ago, Streetwind said:

Rgearding point 4, I'd heavily disclaim that advice. Theoretically it may be sound, but in practical application it has little to do with what's really going on. Your throttle setting should be chosen in response to how your actual launch is going. If throttling down will put you on a better trajectory, then absolutely throttle down!

I watched the numbers carefully, you went full thrust up to 3.1 SLTWR. It's just you have to have a really unorthodox design to make throttling down before TWR = 3 put you into a better trajectory.

25 minutes ago, Streetwind said:

Point 5, could you expand on that? I've been led to believe the exact opposite in the past, and it has worked quite well for me. Additionally, in RL orbital rocketry, coast phases before establishing at least a minimal parking orbit are nearly unheard of, unless you want to count brief moments during staging.

If your claim can be mathematically supported, I would be very interested in seeing it!

Ah, by low-thrust burn I meant "not at max throttle". That's a consequence of control theory. The full sentence should be "for a fuel-optimal trajectory, engine must always be either at max throttle or min throttle". In KSP, that means either go full throttle or coast. In RL, rockets usually don't coast and don't throttle engines, so rule is satisfied. Also, RL rockets typically go without coasting to much lower (compared to planet radius) orbits than KSP ones do. 75 km Kerbin orbit transforms into roughly 800 km Earth orbit, and spacecrafts do coast to reach that orbits.

Finally, the control theory does not say exactly how many coast periods is optimal, so more than one might in fact be needed, and it does not say if coast is a lot or just barely better.

[Streetwind]

41 minutes ago, Pand5461 said:

And you did not mention Soviet designs.

Yeah, I considered grabbing Proton or something, but then I remembered that it has like five gazillion confusingly named subvariants and promptly came down with an acute case of "can't be arsed".  

 

43 minutes ago, Pand5461 said:

I watched the numbers carefully, you went full thrust up to 3.1 SLTWR. It's just you have to have a really unorthodox design to make throttling down before TWR = 3 put you into a better trajectory.

That's more of a coincidence though. I did not watch my TWR at all. I made the call to throttle back based on the five variables I mentioned to Landwalker above. On the run I recorded, I had a fairly flat trajectory, so I stayed at high throttle fairly long. In an unrecorded test before that, I had a steeper initial ascent and ended up throttling down earlier to avoid grossly overshooting. Since this craft does not have fins, strong pitch corrections during max-Q are not exactly the best idea.  

 

48 minutes ago, Pand5461 said:

Ah, by low-thrust burn I meant "not at max throttle". That's a consequence of control theory. *snip*

Ah, yes. I am in full agreement with you now.