Efficient Delivery of heavy equipment

[Northstar1989]

Since this is a "heavy equipment" thread I'm assuming that 'this' load is the heavy one and anything else is therefore lighter. It is always preferable to manoeuvre the lightest vehicle(s) during any operation as they will need least fuel for any acceleration, unless they have very inefficient propulsion. As such a heavy load should be dropped as soon as possible - 70-75km and everything else should do the work of getting to it. However, everything depends on how many other vehicles are going to have to go to that extra effort, in which case you might want to put the heavy load into a a higher parking orbit, as you say (generally, I use 250km for exactly that reason), but you don't launch into that orbit!

Indeed, everything depends on the relative sizes and numbers of craft being launched... For instance, if you're launching several equal-mass payloads to assemble a multi-part ship in orbit (I'm working on something like this right now in my Career save, in fact), you DON'T want to park at a 70 km orbit (the first craft should be launched into the higher orbit, as it will have a lower speed over land that way, and thus will be easier to time additional launches to require less phasing to reach...)

You SHOULD launch straight into a 250 km orbit, however...

Every launch (rules are made to be broken) should be circularised at 70-75km, then perform a Hohmann transfer to the final parking orbit. Using this method you're minimising the flight requirements in thick, draggy, atmosphere and maximising them at optimal orbital-adjustment points. Like a lot of things, the difference isn't much with light vehicles over a short distance but can quickly become significant with large loads across large distances.

That flight profile (circularizing at 70 km and then transferring to a 250 km orbit) is less efficient than launching straight to a 250 km orbit in the first place.

The reason for this is the Oberth Effect. I'm sure you're aware of what it is (more energy is obtained from thrust when traveling at higher speeds), and that it is the reason you always want to raise your apoapsis from periapsis, rather than, say performing a burn at apoapsis until it becomes your new periapsis, and then burning from there...

Yet, this is precisely what you're doing if you circularize at 70 km first. You're burning at the apoapsis of a sub-orbital trajectory until that apoapsis becomes your new periapsis (or very close to it- your orbit won't be very elliptical), and THEN burning at the nee periapsis to raise your apoapsis... That simply isn't an efficient way to get your rocket to 250 km...

Why? You will be moving more slowly at 70 km than you will at, say, 50 km- if your engines are powerful enough that you have to cut throttle and cruise to the edge of the atmosphere at that point to avoid overshooting 70 km (they should be, or else you're wasting fuel on a low TWR and the high gravity-losses that result...) You still experience drag at 50 km, but it is extremely low- in fact so little that your terminal velocity is actually quite a bit higher than orbital speed at that point...

As a result, you are losing a lot more energy to not taking full advantage of the Oberth Effect than you are to increased drag in the uppermost atmosphere, if you don't continue firing your engines on the prograde vector until your apoapsis lies at 250 km... (you should then straight-out circularize at 250 km).

I've performed numerous experiments in the past with launching to 250 or 250 km using MechJeb (to eliminate steering losses) directly vs. first to 70 km in order to confirm this... Try it out for yourself if you don't believe me.

Regards,

Northstar

[Northstar1989]

I agree with this. A high-TWR rocket like Northstar1989 describes may use a few dozen less m/s to attain orbit than a lower one, but it will consume more fuel to do so because of its higher dry mass. I've found that an initial TWR between 1.3 and 1.6 maximizes payload fraction and consumes less fuel to reach orbit for a given payload size.

Everything depends on the particulars, once again. I actually benefit from the dry mass in those engines, because I re-use it via my orbital salvaging/scrapping/recycling operations (what I basically do is recycle the spent upper stages into RocketParts using specialized salvage ships). Even if one doesn't use mods, one can make the engines detachable using docking ports, and slip out empty fuel tanks from between them and the payload before reattachment- therefore reusing the upper stage engines on the payload... Generally if you can get a rocket to orbit for less Delta-V, it's worth it- even if this ends up costing you more fuel at first...

[Citation needed] Jet engines are mass produced, the cost of a jet engine is at least an order of magnitude less than a comparable rocket engine. We don't use jets for getting to orbit because they don't work all that well at the speeds required, and they don't work at all at higher altitudes (can't air hog like in KSP).

Maybe I should have clarified: I'm talking about jet engines of the highest quality and most demanding engineering standards, the kind you'd likely see in use on a spaceplane or air-breathing launch platform (if one existed)- NOT the kind of (slightly) cheaper and lower-quality jet engines you see used for commercial airliners and such... (which are hand-made in small batches, NOT mass-produced, by the way) But even those jet engines cost more than the rocket engines used on actual launch vehicles.

Jet engines have a lot more moving parts than rocket engines, and therefore have a much higher base manufacturing cost. Combined with the fact that virtually NOTHING of space-quality is actually mass-produced (not rocket OR jet engines), the cost difference is likely to be even more drastic if they're being used for a space program.

Jet engines used for large commercial airliners (such as a 747) typically cost over $2,000,000 dollars: https://answers.yahoo.com/question/index?qid=20080922115617AAe2XCD

Whereas a rocket engine of the same type used to power the Saturn-I costs about $200,000 today: http://www.dailymail.co.uk/sciencetech/article-2068224/Up-sale-online-A-NASA-rocket-engine-used-build-missiles.html

Regards,

Northstar

P.S. Those aren't the best sources, but they'll do... I could find nothing that did not confirm the high pricing of large jet engines at over $2 million, whereas it was difficult to find reliable numbers on rocket engine prices. It's also worth noting that large jet engines are often sold *at a loss* by manufacturers, even at a $2 million pricetag- who make back the loss on overpriced spare parts for the engines...

Edited May 18, 2014 by Northstar1989

[TeeGee]

I see detached boosters flying off in every direction in the final screenshot. How can that possibly be reusable?

Regards,

Northstar

All of those boosters reenter back into atmosphere and parachute down for a soft landing. Every one of them is recoverable, thus the booster system is reusable.

[Jouni]

Maybe I should have clarified: I'm talking about jet engines of the highest quality and most demanding engineering standards, the kind you'd likely see in use on a spaceplane or air-breathing launch platform (if one existed)- NOT the kind of (slightly) cheaper and lower-quality jet engines you see used for commercial airliners and such... (which are hand-made in small batches, NOT mass-produced, by the way)

Actually, the highest quality jet engines are those used in passenger planes. They are silent, fuel efficient and extremely reliable. A typical engine used in a long-haul jet may run well over 50% of time for years, before it's taken off the wing and rebuilt.

The engines used in fighter jets are of lower quality. In part this is a design choice: fighter jets only fly a small fraction of time, so efficiency and reliability are traded for raw power. On the other hand, nobody is buying that many fighter jets, so the development budgets of their engines are lower than for passenger jets.

The same reasons that make fighter jet engines to be of lower quality apply even more for (hypothetical) spaceplane engines.

[Red Iron Crown]

Everything depends on the particulars, once again. I actually benefit from the dry mass in those engines, because I re-use it via my orbital salvaging/scrapping/recycling operations (what I basically do is recycle the spent upper stages into RocketParts using specialized salvage ships). Even if one doesn't use mods, one can make the engines detachable using docking ports, and slip out empty fuel tanks from between them and the payload before reattachment- therefore reusing the upper stage engines on the payload... Generally if you can get a rocket to orbit for less Delta-V, it's worth it- even if this ends up costing you more fuel at first...

The equation does change if you're considering your upper engines as part of the payload, true. But the point remains that it is more cost efficient to optimize for payload fraction than for delta-V, all other things being equal.

Maybe I should have clarified: I'm talking about jet engines of the highest quality and most demanding engineering standards, the kind you'd likely see in use on a spaceplane or air-breathing launch platform (if one existed)- NOT the kind of (slightly) cheaper and lower-quality jet engines you see used for commercial airliners and such... (which are hand-made in small batches, NOT mass-produced, by the way) But even those jet engines cost more than the rocket engines used on actual launch vehicles.

Jet engines have a lot more moving parts than rocket engines, and therefore have a much higher base manufacturing cost. Combined with the fact that virtually NOTHING of space-quality is actually mass-produced (not rocket OR jet engines), the cost difference is likely to be even more drastic if they're being used for a space program.

Jet engines used for large commercial airliners (such as a 747) typically cost over $2,000,000 dollars: https://answers.yahoo.com/question/index?qid=20080922115617AAe2XCD

Whereas a rocket engine of the same type used to power the Saturn-I costs about $200,000 today: http://www.dailymail.co.uk/sciencetech/article-2068224/Up-sale-online-A-NASA-rocket-engine-used-build-missiles.html

Jet engines are high volume compared to rocket engines, I don't think that's debatable. Even if they are assembled by hand, there are still economies of scale in the manufacture of the components. The number of moving parts is not a good metric for cost; a piston engine has more moving parts than a jet or rocket, yet they are far, far cheaper.

The price you quote for the Saturn engine is a "surplus" price, I suspect it is far lower than what a new, modern rocket engine costs, especially as the R&D costs are amortized over fewer engines. Hard to say about "space-grade" jet engines, as none exist yet and the ones that are in development are in the high-cost prototype stage.

[Pecan]

You SHOULD launch straight into a 250 km orbit, however...

That flight profile (circularizing at 70 km and then transferring to a 250 km orbit) is less efficient than launching straight to a 250 km orbit in the first place.

The reason for this is the Oberth Effect. I'm sure you're aware of what it is (more energy is obtained from thrust when traveling at higher speeds), and that it is the reason you always want to raise your apoapsis from periapsis, rather than, say performing a burn at apoapsis until it becomes your new periapsis, and then burning from there...

Yet, this is precisely what you're doing if you circularize at 70 km first. You're burning at the apoapsis of a sub-orbital trajectory until that apoapsis becomes your new periapsis (or very close to it- your orbit won't be very elliptical), and THEN burning at the nee periapsis to raise your apoapsis... That simply isn't an efficient way to get your rocket to 250 km...

Why? You will be moving more slowly at 70 km than you will at, say, 50 km- if your engines are powerful enough that you have to cut throttle and cruise to the edge of the atmosphere at that point to avoid overshooting 70 km (they should be, or else you're wasting fuel on a low TWR and the high gravity-losses that result...) You still experience drag at 50 km, but it is extremely low- in fact so little that your terminal velocity is actually quite a bit higher than orbital speed at that point...

...

I've performed numerous experiments in the past with launching to 250 or 250 km using MechJeb (to eliminate steering losses) directly vs. first to 70 km in order to confirm this... Try it out for yourself if you don't believe me.

I didn't believe you so I did experiments, again.

Over several flights with four different vehicles launching direct to 250km had 0.88 - 0.89 (round to 10% less) of the fuel remaining than first circularising at 75km. The (only) exception to this was with a very-high TWR vehicle (A 40t lifter without payload, TWR 3.5 - 5+) where MJ had a lot of trouble chasing the fine-adjustments at circularisation, both at 75km and 250km. In this case circularisation at 75km used 1.001 - 1.004 more fuel (say 0.3% worse on average). (MJ was playing-up a bit and stopped showing deltaV, which is why I'm only giving fuel stats).

• "you always want to raise your apoapsis from periapsis" - yes, exactly - while you're climbing during launch you're nowhere near periapsis.
  
• "You're burning at the apoapsis of a sub-orbital trajectory until that apoapsis becomes your new periapsis (or very close to it- your orbit won't be very elliptical), and THEN burning at the nee periapsis to raise your apoapsis... That simply isn't an efficient way to get your rocket to 250 km" - except that it is, because it's most efficient to raise apoapsis at periapsis, as just stated, AND most efficient to raise periapsis at apoapsis - which is what the circularisation burn does. The burn to raise apoapsis to 250km is therefore performed at the optimum position/speed - periapsis, which itself was established in the most efficient manner.
  
• "You will be moving more slowly at 70 km than you will at, say, 50 km" - True, you will have slowed somewhat during cruise but after the circularisation burn you're faster in orbit than 50km, otherwise you'd fall back into the atmosphere. The orbit-change to 250km, apart from being done at periapsis anyway, is therefore done from a higher initial velocity.
  
• "You still experience drag at 50 km, but it is extremely low" - yes, I think drag's immaterial to the argument one way or another.
  

I recommend you try launching to 750km and higher. The effects of first circularising at 75km are much easier to see.

[Northstar1989]

All of those boosters reenter back into atmosphere and parachute down for a soft landing. Every one of them is recoverable, thus the booster system is reusable.

Errr, not really. From a realism perspective, it doesn't work, as parachutes are a lot weaker and bulkier in real life- meaning that the boosters would sustain damage with this strategy. From a gameplay perspective, it doesn't work either, as the moment one of those boosters exits physics loading-range in the lower atmosphere, it will cease to exist and not be recoverable...

Regards,

Northstar

[Northstar1989]

The equation does change if you're considering your upper engines as part of the payload, true. But the point remains that it is more cost efficient to optimize for payload fraction than for delta-V, all other things being equal.

But if you find a way to re-use or recycle all the mass you get to orbit, then the equations become almost identical... It's a useful simplification to find the way to get a rocket to orbit that costs the least Delta-V, rather than the least fuel...

v

Jet engines are high volume compared to rocket engines, I don't think that's debatable. Even if they are assembled by hand, there are still economies of scale in the manufacture of the components. The number of moving parts is not a good metric for cost; a piston engine has more moving parts than a jet or rocket, yet they are far, far cheaper.

The price you quote for the Saturn engine is a "surplus" price, I suspect it is far lower than what a new, modern rocket engine costs, especially as the R&D costs are amortized over fewer engines. Hard to say about "space-grade" jet engines, as none exist yet and the ones that are in development are in the high-cost prototype stage.

That is part of the point, that no space-grade jet engines currently exist. Therefore, developing them requires an extensive R&D cycle, adding to their cost- whereas rocket engines have already been in use for over 50 years to carry payloads to orbit. Therefore, the most realistic cost-comparison really would be between a surplus rocket engine and a state-of-the-art jet engine, as this best mimics the relationship between a state-of-the-art rocket engine and a only-theoretical jet engine in terms of cost...

The number of moving parts is indeed a major factor in manufacturing and design cost. It is true, it is not the only factor, but it is one of the main reasons jet engines cost so much more than rocket engines...

Regards,

Northstar

[Northstar1989]

I didn't believe you so I did experiments, again.

Over several flights with four different vehicles launching direct to 250km had 0.88 - 0.89 (round to 10% less) of the fuel remaining than first circularising at 75km. The (only) exception to this was with a very-high TWR vehicle (A 40t lifter without payload, TWR 3.5 - 5+) where MJ had a lot of trouble chasing the fine-adjustments at circularisation, both at 75km and 250km. In this case circularisation at 75km used 1.001 - 1.004 more fuel (say 0.3% worse on average). (MJ was playing-up a bit and stopped showing deltaV, which is why I'm only giving fuel stats).

There *might* be some differences in the ascent profiles of our rockets, even when circularizing at the same points. Don't forget that the MechJeb default launch pattern is usually not the most efficient flight plan to get your rocket to orbit. While it works well as a general approximation, you need to adjust your flight path based on factors like launchpad TWR and staging, which cause you ideal flight plan to deviate from the default/generalized MechJeb plan...

• "you always want to raise your apoapsis from periapsis" - yes, exactly - while you're climbing during launch you're nowhere near periapsis.
  
• "You're burning at the apoapsis of a sub-orbital trajectory until that apoapsis becomes your new periapsis (or very close to it- your orbit won't be very elliptical), and THEN burning at the nee periapsis to raise your apoapsis... That simply isn't an efficient way to get your rocket to 250 km" - except that it is, because it's most efficient to raise apoapsis at periapsis, as just stated, AND most efficient to raise periapsis at apoapsis - which is what the circularisation burn does. The burn to raise apoapsis to 250km is therefore performed at the optimum position/speed - periapsis, which itself was established in the most efficient manner.
  
• "You will be moving more slowly at 70 km than you will at, say, 50 km" - True, you will have slowed somewhat during cruise but after the circularisation burn you're faster in orbit than 50km, otherwise you'd fall back into the atmosphere. The orbit-change to 250km, apart from being done at periapsis anyway, is therefore done from a higher initial velocity.
  
• "You still experience drag at 50 km, but it is extremely low" - yes, I think drag's immaterial to the argument one way or another.
  

I recommend you try launching to 750km and higher. The effects of first circularising at 75km are much easier to see.

I've repeatedly performed these experiments pre-0.23.5 The result was always the same- the direct launch saved fuel after I figured out the most fuel-optimal flight pattern for the direct launch...

Try tweaking the flight pattern on the direct launch to 250 km, or even performing it manually using the MechJeb ASAS (in "Surface Mode"), to see if you can't out-perform your 2-part launch with a bit of optimization. My bet is, you can (I could). It might also be helpful to see screenshots, with a lot of emphasis on the gravity-turn portion, to see how each ascent could be improved.

Also, what mods were you running during your ascent? Running FAR, for instance, changes the relationship between a direct launch and a 2-part launch, by making drag less of an issue with properly-designed rockets...

Regards,

Northstar

Edited May 18, 2014 by Northstar1989

[Northstar1989]

I also feel the need to more carefully address these parts of your post in particular:

yes, exactly - while you're climbing during launch you're nowhere near periapsis.

Yes, but you're *closer* to periapsis than when you're at apoapsis. The worst possible place to raise your apoapsis is from another apoapsis, by burning there until it becomes your new periapsis, as the average speed you are moving during your burn (which increases throughout the burn) is the lowest there. If you burn closer to periapsis, you will have a more fuel-efficient increase in your apoapsis... This increases the closer you are to periapsis, until you reach the ideal point of burning *at* periapsis...

(I need to double-check whether you want to burn at prograde, or parallel to the horizon, or something in-between at a point that is between apoapsis and periapsis, however... I remember the answer not being 100% intuitive, and may have incorrectly stated it before when I said to simply burn prograde...)

except that it is, because it's most efficient to raise apoapsis at periapsis, as just stated, AND most efficient to raise periapsis at apoapsis - which is what the circularisation burn does. The burn to raise apoapsis to 250km is therefore performed at the optimum position/speed - periapsis, which itself was established in the most efficient manner.

As I stated before, in the post to which you were responding, this point only becomes your periapsis through an apoapsis-burn. This is less efficient than burning at a point closer to your original periapsis...

True, you will have slowed somewhat during cruise but after the circularisation burn you're faster in orbit than 50km, otherwise you'd fall back into the atmosphere. The orbit-change to 250km, apart from being done at periapsis anyway, is therefore done from a higher initial velocity.

You CANNOT discuss two sections of a continuous burn separately like that, and ignore only one portion, focusing on the other. Yes, your speed is higher AFTER the circularization is complete- but that is already well into your burn to raise your apoapsis to 250 km. The average speed at which you are moving during that burn, starting from the beginning of your circularization burn, and ending at the end of your burn when your apoapsis has reached 250 km, is LESS than during a burn started at 50km in a sub-orbital trajectory, that raises your apoapsis to 250 km.

Even if you actually perform the burn in two parts, by first circularizing, and then raising the apoapsis at the new periapsis the next orbit around, the 2-part burn can be treated the same mathematically as an idealized (instantaneous) 1-part burn at the apoapsis of your sub-orbital trajectory. And, the average speed of that burn is less than in an idealized (instantaneous) burn at 50 km to raise your apoapsis to 250 km.

yes, I think drag's immaterial to the argument one way or another.

At least we can agree on this.

I recommend you try launching to 750km and higher. The effects of first circularising at 75km are much easier to see.

I've tested once or twice going to 750 km or even higher, and the effects didn't seem to change. The problem with that experiement, however, is that the transfer orbit from a 70 km circular orbit to a 750 km orbit, is highly elliptical- so there is *VERY LITTLE* difference between its eccentricity and that of a direct transfer to 750 km. Therefore, Delta-V savings, while not decreasing in absolute size, become relatively much smaller- and thus *MUCH* harder to actually measure...

The *BETTER* experiment is actually to compare a direct launch to 100 km vs. a 2-part launch to 100 km (to 70 km and then a transfer to 100 km), where the fuel-savings of the direct launch should be MUCH more noticeable. Try performing *that* experiment a few times, and see how your results differ...

Regards,

Northstar

Edited May 18, 2014 by Northstar1989

[Pecan]

At least we can agree on this.

So the rest is simple - your statistics please.

I have no idea how you think "you're *closer* to periapsis than when you're at apoapsis" makes raising periapsis during ascent better - since the whole point is NOT to try to change it close-to. As you say "The worst possible place to raise your apoapsis is from another apoapsis" and the worst place to change periapsis is at periapsis, or close to it. I change Pe at Ap, not Ap at Ap - obviously they do get close in circularisation, as you say.

Are you being deliberately obtuse with "If you burn closer to periapsis, you will have a more fuel-efficient increase in your apoapsis"? I specifically said I increase Ap at Pe - a Pe I've just incremented, most efficiently, at Ap. The whole thing is four burns 1) Ascend to 75km Ap, 2) Circularise - increase Pe at Ap, 3) Increase Ap at Pe, 4) Circularise - increase Pe at Ap. I am NOT saying you should combine 2 & 3 to go straight from a 75km Ap to a 250km one.

"As I stated before, in the post to which you were responding, this point only becomes your periapsis through an apoapsis-burn. This is less efficient than burning at a point closer to your original periapsis..." (You may want to check the post I'm responding to, Ap/Pe is getting a bit convoluted.) This sentence says 'burn at periapsis to increase periapsis' and I assume you typed it in error - because it is completely incorrect.

... comments removed ...

Having looked at the rest of your post in detail I think that's where your problem is actually - you seem to think burning at apoapsis is always wrong. It isn't; it's the most efficient place to raise periapsis.

Get into orbit first - adjustments, especially as small as 75km -> 250km take hardly any deltaV and/or fuel. To be honest I don't even understand your comments about 750km and other orbits unless, again, you're not circularising but trying to burn straight from a 75km apoapsis to a 750km one. All results I've seen mean the greater the orbital change the greater the difference between doing it right and doing it wrong. If you're finding the results instead get closer ... ?

ETA: All but 2 mods I use are info-ony (MJ/KER/KAC, etc.) The only parts-mods I'm using are SCANSat and PF. While PF may have an impact it was only part of one of the four vehicles I tested with, this time around. In stock KSP I believe it's only effect is to make performance worse, however (increased mass, no streamlining benefit). I used MJ's default ascent-path for all launches as I wanted to test against your assertion that, using it for consistency, you'd got different results. Obviously, by tweaking things and flying manually *you* may be better at direct-ascent than 75km circularisation - that's a different test. As I pointed out, in one of my models MJ was sufficiently bad at circularisation that direct-ascent became more efficient. All that means is that if you aren't good at circularising you'll waste so much fuel you'll lose any benefit you *should* have got.

Edit 2: I'm not trying pick on you; every 'you' in that last sentence, now italicised, is in the sense of 'one'. Essentially the difference is very small and the simplicity and reliability of a direct-ascent may more than compensate for any potential gains one chooses to forego by not first circularising at 75km (or lower).

Edited May 19, 2014 by Pecan

[Northstar1989]

So the rest is simple - your statistics please.

I have no idea how you think "you're *closer* to periapsis than when you're at apoapsis" makes raising periapsis during ascent better - since the whole point is NOT to try to change it close-to. As you say "The worst possible place to raise your apoapsis is from another apoapsis" and the worst place to change periapsis is at periapsis, or close to it. I change Pe at Ap, not Ap at Ap - obviously they do get close in circularisation, as you say.

Are you being deliberately obtuse with "If you burn closer to periapsis, you will have a more fuel-efficient increase in your apoapsis"? I specifically said I increase Ap at Pe - a Pe I've just incremented, most efficiently, at Ap. The whole thing is four burns 1) Ascend to 75km Ap, 2) Circularise - increase Pe at Ap, 3) Increase Ap at Pe, 4) Circularise - increase Pe at Ap. I am NOT saying you should combine 2 & 3 to go straight from a 75km Ap to a 250km one.

"As I stated before, in the post to which you were responding, this point only becomes your periapsis through an apoapsis-burn. This is less efficient than burning at a point closer to your original periapsis..." (You may want to check the post I'm responding to, Ap/Pe is getting a bit convoluted.) This sentence says 'burn at periapsis to increase periapsis' and I assume you typed it in error - because it is completely incorrect.

... comments removed ...

Having looked at the rest of your post in detail I think that's where your problem is actually - you seem to think burning at apoapsis is always wrong. It isn't; it's the most efficient place to raise periapsis.

Get into orbit first - adjustments, especially as small as 75km -> 250km take hardly any deltaV and/or fuel. To be honest I don't even understand your comments about 750km and other orbits unless, again, you're not circularising but trying to burn straight from a 75km apoapsis to a 750km one. All results I've seen mean the greater the orbital change the greater the difference between doing it right and doing it wrong. If you're finding the results instead get closer ... ?

ETA: All but 2 mods I use are info-ony (MJ/KER/KAC, etc.) The only parts-mods I'm using are SCANSat and PF. While PF may have an impact it was only part of one of the four vehicles I tested with, this time around. In stock KSP I believe it's only effect is to make performance worse, however (increased mass, no streamlining benefit). I used MJ's default ascent-path for all launches as I wanted to test against your assertion that, using it for consistency, you'd got different results. Obviously, by tweaking things and flying manually *you* may be better at direct-ascent than 75km circularisation - that's a different test. As I pointed out, in one of my models MJ was sufficiently bad at circularisation that direct-ascent became more efficient. All that means is that if you aren't good at circularising you'll waste so much fuel you'll lose any benefit you *should* have got.

Edit 2: I'm not trying pick on you; every 'you' in that last sentence, now italicised, is in the sense of 'one'. Essentially the difference is very small and the simplicity and reliability of a direct-ascent may more than compensate for any potential gains one chooses to forego by not first circularising at 75km (or lower).

I'm not understanding how you're getting your own language so convoluted. Or mine. What I said was very clear.

In a sub-orbital trajectory to 70 km, your apoapsis is at 70 km. You can model a sub-orbital trajectory as a highly elliptical orbit, if you forget the presence of the atmosphere for a second (it ceases to have any truly significant effect by the altitudes we are discussing), and forget about the terrain (model the planet as a point-mass at its Center of Mass- which is basically what KSP does- adding a collision mesh at the planetary radius). Your "periapsis" from a mathematical standpoint, then lies within the planet itself.

It is *more efficient* to burn before you reach 70 km (at 50 km) in order to set a sub-orbital trajectory with an apoapsis at 250 km, and then circularize at 250 km, than to first circularize, and then raise your apoapsis at 250 km.

For simplicity's sake, I assume that if you have an apoapsis at 70 km, you will burn until this point is your periapsis (and just above the atmosphere) and your new apoapsis is 71 or 72 km on the opposite side of the planet. It's almost exactly the same mathematically as if you set a sub-orbital trajectory at 71 or 72 km, and then raise your periapsis to 70 km in your circularization burn, but is easier to imagine/visualize.

Once gain, you are moving faster at 50 km when in a sub-orbital (ballistic) trajectory with an apoapsis at 70 or 72 km, than you are when you reach 70 or 72 km. While you are much closer to apoapsis than to "periapsis" (which is actually a mathematical point close to the core of the planet, which you would crash long before reaching) at 50 km, you are still closer to the imaginary (negative-altitude) periapsis than you will be at 70 km. It is a more efficient point to raise your apoapsis further than at 70 km.

What I've tried to point out repeatedly, and you don't seem to understand, is that first circularizing, and then raising your apoapsis, is exactly the same mathematically as setting a highly elliptical orbit, and then burning at the apoapsis of that elliptical orbit until that altitude is your new periapsis, and continuing to burn at that altitude until your apoapsis is even higher. So actually, you ARE "burning at apoapsis to raise you apoapsis", which is inefficient.

This is why I repeatedly state you will save Delta-V going for a straight launch to a higher orbit, and then circularizing when your reach apoapsis at the altitude of that higher orbit.

Regards,

Northstar

Edited May 19, 2014 by Northstar1989

[Northstar1989]

Also, it makes almost no difference whether you perfectly circularize at 70 (or 75 km), and then burn at another point of the exactly circular orbit until your apoapsis is 250 km; or whether you were to burn at apoapsis of a sub-orbital trajectory, until you reached a perfectly circular orbit, and then continue burning at the same point (assume either an infinite TWR, or that you wait until your orbit takes you back to the exact same point- the two are essentially the same) until your apoapsis is at 250 km.

Either one is mathematically identical. The real deviations between the two are due to real-world limits due to TWR, and make little real difference in the discussion here.

A continuous burn at 70 km, to transform that point from the apoapsis of a sub-orbital trajectory to the periapsis of an elliptical orbit with an apoapsis at 250 km; is less efficient than an ascent at which you continuously burn your engines in the upper atmosphere until your apoapsis lies at 250 km, and then coast to that point, and circularize there.

I would like to point out that I *NEVER* said you are "closer to periapsis at apoapsis". I said you are closer to periapsis of a sub-orbital trajectory at 50 km than at 70 km. This assumed you could understand the point that sub-orbital trajectories can be modelled as elliptical orbits.

Real deviations between the two are due to TWR limitations of vessels, steering losses, and drag.

Once again, try the one-part vs. 2-part ascent to 100 km. You should see the differences in fuel consumption more clearly there...

Regards,

Northstar

Edited May 19, 2014 by Northstar1989

[Northstar1989]

Having looked at the rest of your post in detail I think that's where your problem is actually - you seem to think burning at apoapsis is always wrong. It isn't; it's the most efficient place to raise periapsis.

I think much of my point can be summed up as this: it's most efficient to FIRST raise your apoapsis to the desired altitude, and THEN raise your periapsis at that altitude so as to circularize; than it is to FIRST raise your apoapsis to an intermediate point between where you are (even being landed on the surface at sea level can be modeled as a 600 km x 600 km orbit around Kerbin's Center of Mass, with the normal force with the ground making up for the rotational velocity being less than the orbital velocity at that altitude, if you ignore the atmosphere for a moment...) and where you want to be, and THEN circularize at that intermiedate orbit only to have to raise your orbit again to your desired altitude.

A thought question for you: which of the following do you think would be more efficient, starting from a 70 km perfectly circular orbit?

(1) Performing a burn until you are in an elliptical orbit with an apoapsis at 350 km, performing another burn at 320 km (*NOT* 350 km) until your apoapsis lies at 700 km, and THEN circularizing at 700 km.

(2) Performing a burn until you are in an elliptical orbit with an apoapsis at 350 km, circularizing at 350 km, and then raising your orbit AGAIN to reach 700 x 700 km.

The FIRST of these is an excellent analogy for a direct launch to a higher orbit that I am advocating, the SECOND of these is an excellent analogy for the "circularize first" pattern of reaching higher orbits that you are espousing.

It can be mathematically shown that the first of these requires slightly less Delta-V.

Regards,

Northstar

Edited May 19, 2014 by Northstar1989

[Pecan]

"In theory, theory and practice are the same. In practice, they're not"

Whatever you may be able to show mathematically is contradicted by actual results, which consistently show it is more efficient to circularise just outside the atmosphere (subject to being able to cicularise efficiently).

[InF3RNo]

Hello, you talk about finding uranium and all that, is that coming from a mod or game itself now have fuels obtainable from planets or moons?

Thanks!

InF3RNo.

[Pecan]

"Kethane" fuel is from the Kethane mod. Uranium, etc. from Interstellar.

[TeeGee]

Errr, not really. From a realism perspective, it doesn't work, as parachutes are a lot weaker and bulkier in real life- meaning that the boosters would sustain damage with this strategy. From a gameplay perspective, it doesn't work either, as the moment one of those boosters exits physics loading-range in the lower atmosphere, it will cease to exist and not be recoverable...

Regards,

Northstar

Hunh? Then how do I recover them? I have DR AND a OKTD probe pod on them, as well as reaction control wheels for reentry management in DR. It works.

In stock, it DEFINITELY works, because I've done it dozens of times. Go into map mode, select debris, fly debris and follow it to the ground. The chutes arm during staging and deploy when they hit thick enough air.

[Northstar1989]

"In theory, theory and practice are the same. In practice, they're not"

Whatever you may be able to show mathematically is contradicted by actual results, which consistently show it is more efficient to circularise just outside the atmosphere (subject to being able to cicularise efficiently).

That's precisely the problem.

While I can definitively state that the mathematics prove that a direct launch to higher orbits should be more efficient than circularizing first at 70km, your own experience may be somewhat different. I can envision two possible explanations for this. Either:

(A) Your execution of a two-part launch is significantly more efficient than your execution of a direct launch (that is, it has fewer steering errors and more optimization of the gravity turn)

OR

( My two-part launches are highly inefficient, leading to my consuming significantly more Delta-V when carrying them out than in a direct launch- even if it should theoretically be easier to perform a perfect two-part launch than a perfect direct launch.

All I can say is this. A rocket with infinite TWR, perfect steering, and no significant aerodynamic perturbations of the ascent (basically, stock aerodynamics without lifting surfaces) would be able to reach a higher altitude for less fuel and Delta-V with a direct launch than with a two-part launch.

How theory and practice actually line up most of the time, with most players, I can't say- I can only speak from my own experiences, where I have always found direct launches more efficient.

Regards,

Northstar

[Northstar1989]

Hunh? Then how do I recover them? I have DR AND a OKTD probe pod on them, as well as reaction control wheels for reentry management in DR. It works.

In stock, it DEFINITELY works, because I've done it dozens of times. Go into map mode, select debris, fly debris and follow it to the ground. The chutes arm during staging and deploy when they hit thick enough air.

Read what I said- it doesn't work because of their exiting physics loading range. If you manage to get them all to the ground within a 2.3 kilometer radius circle, that's great- but most of the time, such precise simultaneous landings of multiple pieces of debris is impossible...

If you're in a high enough orbit, even the force of a decoupler will often force the landing sites of the different engines more than 2.3 km apart...

From a realism perspective, it works even less, because jet engines are OP'd in the first place. Try building that launch vehicle with Advanced Jet Engines mod (which is a realism mod that basically converts the stock jet engines to their Apollo Era equivalents- the stock engines being far beyond what can be accomplished even with current technology), and Real Solar Ssytem, and see how it works out... (spoiler: it won't)

Regards,

Northstar

P.S. I'm not a fan of Advanced Jet Engines mod myself- but only because the technology is nerfed back to the Apollo Era, and I prefer playing with modern and near-future technology. It also doesn't include any pre-cooler technology, even though it introduces jet compressor and engine core overheating- making jet engines effectively worthless for spaceplanes...

Edited May 22, 2014 by Northstar1989

[Pecan]

Then I would appreciate something from you that I can replicate - I mean this; I'm not denying the possibility of what you say but you've not given me anything I can test. Every way I fly to orbit, manually or with MJ, beats what I think you mean, which may be very different to what you actually mean. If you're right I would like to know but I think we've hijacked this thread more than we should already. PM me if you wish (or however best you think to continue).

[Northstar1989]

Then I would appreciate something from you that I can replicate - I mean this; I'm not denying the possibility of what you say but you've not given me anything I can test. Every way I fly to orbit, manually or with MJ, beats what I think you mean, which may be very different to what you actually mean. If you're right I would like to know but I think we've hijacked this thread more than we should already. PM me if you wish (or however best you think to continue).

It might be best if I post something if I get around to it...

Like I said, the math works out- but it's been a long time since I actually tested this. It's even possible (though unlikely) something actually changed since my flights to test this...

I've just been putting off performing more test flights, because I've a rather lot to do with my Career save besides pulling off test flights of a concept I already tested in the past (I *COULD* go and find the screenshots from my test flights- but they're buried deep in the archived images of my Imgur account...)

New post when I've got some screenshots to show of what I mean, I guess... Might be a while, if I remember it at all, and don't get distracted by real life...

Regards,

Northstar

[Northstar1989]

This was going to bug me and bug me and bug me if I didn't do it, so.... DATA!

Javascript is disabled. View full album

Javascript is disabled. View full album

I meant to do this with MechJeb's Ascent Guidance feature- but then I remembered (after a couple failed launches) oh yeah, that thing sucks! Actually, I just had a rocket that was too realistic for it- very little torque, so almost all turning had to come from thrust vectoring, and MechJeb just couldn't handle that without my doing at least some of the piloting... (I still mostly used its ASAS function in Surface mode, rather than manual steering, to minimize steering losses)

I think it's pretty clear which rocket has more fuel in the end...

Note that due to the rocket equation, that 500 or so LiquidFuel (and corresponding Oxidizer) actually only represents a very small amount of Delta-V.

Regards,

Northstar

P.S. I'm aware that running the engine at low power in the upper atmosphere like that seems a little funny, but it's actually the most fuel-efficient way to keep apoapsis up and counteract drag at that altitude while finishing your climb above the atmosphere (at anything much over 32000 meters, terminal velocity is greater than orbital velocity.) Of course, it's also unique to attempting to attain such a low orbit from launch- anything higher usually requires that you don't shut off or throttle down your engines until significantly later...

Edited May 22, 2014 by Northstar1989

[Northstar1989]

Also, here was the one successful attempt I managed before I decided that (1) MechJeb's Ascent Guidance still sucks, and (2) It would be better to aim for an 170 km orbit than a 120 km orbit with this rocket, as it would at least make the direct launch significantly easier (I have a tendency to overshoot my target apoapsis if I'm not careful)

Javascript is disabled. View full album

As you'll notice, due to my reliance on the MechJeb Ascent Guidance, it has quite a bit less fuel left at the end than either of the vessels that I manually guided to 170 km- despite being at a lower orbit...

These are all to be used in my Career save as fuel tankers by the way- hence their lack of any non-fuel payload.

Regards,

Northstar

Edited May 22, 2014 by Northstar1989