Transfering to Mun: First to LKO or Directly to Mun?

[numerobis]

Horizontal is significantly more efficient, unless TWR is high.

So, it seems the Kerbals were right when they said to add MOAR BOOSTERS!

Now code it up for unbounded TWR, but maintain terminal velocity the whole way (until terminal velocity exceeds your target speed), and add the atmospheric losses.

[Empiro]

Yeah I think that one thing to keep in mind is that delta-V is not the only measure of efficiency out there. Total mass, total fuel burned, and total cost are all things that can be taken into account as well.

Where the the standard LKO -> Mun transfer approach shines is that your upper stage can be a single light, efficient engine. This means that everything below it can be made much lighter and smaller.

[numerobis]

At TWR 2 the graph shows 50% loss, which means you're burning 3300 m/s more than the competing approach.

That's quite pessimistic. The full analysis, including atmosphere, should only show about 1 km/s loss compared to doing a gravity turn and direct transfer.

[arkie87]

At TWR 2 the graph shows 50% loss, which means you're burning 3300 m/s more than the competing approach.

%_Loss is calculated as (dV_vertical/dV_horizontal - 1)*100%, so 50% loss is about 1650 m/s if 3300 is required total.

That's quite pessimistic. The full analysis, including atmosphere, should only show about 1 km/s loss compared to doing a gravity turn and direct transfer.

i might code it better accounting for the fact that before doing a gravity turn, one has to jump up to a certain height first before starting the turn.

If that height is 20 km, then 632 m/s is wasted to jump up that high, which is about 20% of total deltaV...

[GoSlash27]

FWIW, I tested out this proposition using one of my "pathfinder" probes to get some empirical data.

Using the prograde "gravity turn" direct injection (yeah, I know. save the hate mail), I was able to hit a munar apoapsis with 10.01 units of fuel remaining.

Using the direct vertical ascent, I had 4.46 units remaining. Ironically, this actually got me a rendezvous due to the timing, though I wasn't shooting for one. Just blind luck.

As a check, I ran the prograde to LKO, then transfer prograde burn to Munar apoapsis.

9.66 units remaining.

Result:

The direct prograde injection was the most efficient.

The prograde LKO/ TMI burn was next.

The vertical TMI was by far the least efficient of the 3 approaches.

I can run the numbers and get you DV figures tomorrow if you'd like them.

Best,

-Slashy

Edited December 12, 2014 by GoSlash27

[Superfluous J]

At TWR 2 the graph shows 50% loss, which means you're burning 3300 m/s more than the competing approach.

That's quite pessimistic. The full analysis, including atmosphere, should only show about 1 km/s loss compared to doing a gravity turn and direct transfer.

In my testing 1.1km/s was what I saved by going to orbit first. That's a big "only" though. You can do a lot with 1 km/s, and you can do a lot by not bringing 1 km/s worth of useless fuel with you.

[arkie87]

FWIW, I tested out this proposition using one of my "pathfinder" probes to get some empirical data.

Using the prograde "gravity turn" direct injection (yeah, I know. save the hate mail), I was able to hit a munar apoapsis with 10.01 units of fuel remaining.

Using the direct vertical ascent, I had 4.46 units remaining. Ironically, this actually got me a rendezvous due to the timing, though I wasn't shooting for one. Just blind luck.

As a check, I ran the prograde to LKO, then transfer prograde burn to Munar apoapsis.

9.66 units remaining.

Result:

The direct prograde injection was the most efficient.

The prograde LKO/ TMI burn was next.

The vertical TMI was by far the least efficient of the 3 approaches.

I can run the numbers and get you DV figures tomorrow if you'd like them.

Best,

-Slashy

10 Units of fuel out of how many? If it's hundreds, then they are a lot closer than i would have thought...

What was your TWR?

[FleshJeb]

When you burn diagonally, your gravity losses are lower. For example, to experience 1g acceleration straight up, you'd need a TWR of 2. To get 1g when burning at a 45 degree angle, you'd only need a TWR of 1.41 (square-root of 2). This is because gravity has no effect on the horizontal component of your burn.

That's not the same comparison, but I get your point.

[GoSlash27]

10 Units of fuel out of how many? If it's hundreds, then they are a lot closer than i would have thought...

What was your TWR?

10 units out of 45. This is a very minimalist craft, and 10 units accounts for an awful lot of DV.

TWR for all cases was a maximum of 2G. I held the TWR to the standard for typical prograde profiles* and left it running hot for the vertical ascent at 2G for a reverse terminal velocity fall into space for the vertical.

* Prograde profile was gravity kick at 7KM, 68* pitch at 15KM, 45* pitch at 25KM, and 23* pitch at 35KM with a constant, linear reduction throughout. 1 -> 2G acceleration in the vertical boost phase, .75G acceleration in the transstage phase, and .5G in the injection phase.

I'll get you exact DV numbers (+/- .1%) tomorrow, but trust me... it's a lot. I usually arrive on station outside Kerbin's SOI after correction burns with more than 6 units remaining, so we're looking at at least 500 M/sec DV above the usual 4,900 for a munar mission.

Best,

-Slashy

Edited December 12, 2014 by GoSlash27

[arkie87]

FWIW, I tested out this proposition using one of my "pathfinder" probes to get some empirical data.

Using the prograde "gravity turn" direct injection (yeah, I know. save the hate mail), I was able to hit a munar apoapsis with 10.01 units of fuel remaining.

Using the direct vertical ascent, I had 4.46 units remaining. Ironically, this actually got me a rendezvous due to the timing, though I wasn't shooting for one. Just blind luck.

As a check, I ran the prograde to LKO, then transfer prograde burn to Munar apoapsis.

9.66 units remaining.

Result:

The direct prograde injection was the most efficient.

The prograde LKO/ TMI burn was next.

The vertical TMI was by far the least efficient of the 3 approaches.

I can run the numbers and get you DV figures tomorrow if you'd like them.

Best,

-Slashy

I did the test and recorded it (many discarded trials). Results were very close but direct "gravity turn" and immediate injection to mun apoapsis wins. Please check my video and tell me if im doing something less efficiently that i could be...

But i think the results are close enough such that its the decision between carefully executing the gravity-turn and then transfer vs. just burning straight up and getting a snack is not a trivial one (assuming you have a high TWR)-- after all, kerbals like (read: NEED) snacks. The difference was not anywhere near a km/s (though i didnt include circularization burn).

Will post video soon.

[Superfluous J]

Super-high TWR craft spend a lot of money on engines (and the fuel to carry them), and you can't effectively use those engines doing a proper gravity turn because you will lose so much to drag.

It'd been so long since I'd done this that I forgot that wrinkle originally in this thread. Yes. It is possible to construct a horrid behemoth that is capable of getting to Mun going straight up "more efficiently" than doing a gravity turn. However, if you want to get a payload to Mun orbit and want to do it with the least money spent on both the rocket and the fuel then getting to LKO first is the best way.

That isn't as flashy and simple to see as "less dV always!" which is sad but it's still a fact: Straight up is not and never will be the best method.

[arkie87]

[arkie87]

Straight up is not and never will be the best method.

I'm not saying it's the best method. All i'm saying is:

(1) it's not necessarily as bad a deltaV killer as people say it is

(2) given how nominally worse it is but how drastically easier it is to perform, it should be considered as an option

(3) It's definitely something to consider in career mode early on in the tech tree --> see below

Super-high TWR craft spend a lot of money on engines (and the fuel to carry them), and you can't effectively use those engines doing a proper gravity turn because you will lose so much to drag.

It'd been so long since I'd done this that I forgot that wrinkle originally in this thread. Yes. It is possible to construct a horrid behemoth that is capable of getting to Mun going straight up "more efficiently" than doing a gravity turn. However, if you want to get a payload to Mun orbit and want to do it with the least money spent on both the rocket and the fuel then getting to LKO first is the best way.That isn't as flashy and simple to see as "less dV always!" which is sad but it's still a fact:

Yes. I agree. It's not always about dV. In campaign it's about what parts you have available and cost. In early campaign, it's hard to make a gravity turn since you have no torque or control surfaces as per original post. Furthermore, boosters are great at increasing TWR during launch and are CHEAP in campaign. Thus, boosters and vertical launch go well together.

[Wanderfound]

Requiring high TWR is a ÃŽâ€V killer all by itself, however. You need more engine mass and less fuel capacity in order to achieve it.

[Superfluous J]

I'm not saying it's the best method. All i'm saying is:

(1) it's not necessarily as bad a deltaV killer as people say it is

(2) given how nominally worse it is but how drastically easier it is to perform, it should be considered as an option

(3) It's definitely something to consider in career mode early on in the tech tree --> see below

(1) Yes it is. 20% more fuel is not marginal. It's not as bad as some poor methods of flying, sure, but it's still significantly suboptimal.

(2) If you don't care how much fuel you waste and/or extra engines you carry, they go for it

(3) I have never, ever had trouble turning a rocket in the early career, unless I literally forgot to think about it in the design phase. And then it's usually just while the engines aren't running.

I'm on a poor connection right now so can't watch the video. Is that a mainsail on the bottom of that orange tank? I'll build the same rocket and see how it goes.

[arkie87]

(1) Yes it is. 20% more fuel is not marginal. It's not as bad as some poor methods of flying, sure, but it's still significantly suboptimal.

(2) If you don't care how much fuel you waste and/or extra engines you carry, they go for it

(3) I have never, ever had trouble turning a rocket in the early career, unless I literally forgot to think about it in the design phase. And then it's usually just while the engines aren't running.

I'm on a poor connection right now so can't watch the video. Is that a mainsail on the bottom of that orange tank? I'll build the same rocket and see how it goes.

(1) From my video, the increase in deltaV was 70 m/s on a 3000+ m/s burn, which is negligible... I agree though in theory -- 20% is too much, but from my video, its less than 2%. And I'm sure you would agree that 2% is negligible...

(2) I only care about cost. And SRB's are CHEAP!

(3) How do you turn a rocket with only a command pod to provide torque (no control surfaces or separate reaction wheels)? You could replace SRB's with gimboling rocket engine (LV-T45) but its more than double the price...

It's a parachute, mk1 command pod, decoupler, medium size fuel tank, LV-T30 (i think), decoupler, adapter, orange fuel tank, mainsail with 5264 m/s deltaV at launch

Let me know how it goes!

[arkie87]

Requiring high TWR is a ÃŽâ€V killer all by itself, however. You need more engine mass and less fuel capacity in order to achieve it.

That is true, but we are comparing the performance of the same rocket being flow differently (one with high TWR, in this case).

While it's true that high TWR is a deltaV killer, it might be cheaper in terms of Kerbucks, since SRB's are much cheaper and can be used to increase TWR at launch.

The career definitely tries to push SRB's at present, since they are about the cost of a liquid engine (not including fuel), but its been hard for me to steer them up into LKO efficiently, since they dont stop burning and dont have thrust vectoring, so this entire post has been about that observation and one potential way to deal with it

[arkie87]

That is true, but we are comparing the performance of the same rocket being flow differently (one with high TWR, in this case).

While it's true that high TWR is a deltaV killer, it might be cheaper in terms of Kerbucks, since SRB's are much cheaper and can be used to increase TWR at launch.

The career definitely tries to push SRB's at present, since they are about the cost of a liquid engine (not including fuel), but its been hard for me to steer them up into LKO efficiently, since they dont stop burning and dont have thrust vectoring, so this entire post has been about that observation and one potential way to deal with it

Maybe I should built two craft that cost roughly the same amount of Kerbucks, and have the same upper stage lander, but the booster stages will be different:

(1) Booster stage made from LV-T45 engines for thrust vectoring-assisted gravity-turn into LKO and then horizontal transfer to the Mun

(2) Booster stage made from SRB's using vertical take off

And we can see which one can get farther

[Wanderfound]

SRBs are certainly the wise budget choice, but they don't imply high TWR to me. Instead I use the SRB thrust limiter to set launch TWR at 1.7 or so. As well as allowing the main liquid engine to be run at full blast from launch, it extends the SRB burn to the time when the main engine has burnt enough fuel that a relatively low thrust engine can maintain the ascent. Makes turning a bit easier, too.

While I certainly appreciate the appeal of overpowered ships...

...I don't see much use for them apart from giggles. Which is a perfectly good reason for them in itself.

[Superfluous J]

It is very possible to build a ship that can get to Mun straight up with about the same dV as getting to orbit first.

It is also very possible to build a ship that can get to Mun via getting into orbit first with significant savings over going the straight-up method.

It is NOT possible for the first ship to be cheaper, smaller, or use less fuel than the second ship (for the same payload).

Also, as the payload increases, the savings tilts towards the orbit-first craft.

[arkie87]

It is very possible to build a ship that can get to Mun straight up with about the same dV as getting to orbit first.

It is also very possible to build a ship that can get to Mun via getting into orbit first with significant savings over going the straight-up method.

It is NOT possible for the first ship to be cheaper, smaller, or use less fuel than the second ship (for the same payload).

Also, as the payload increases, the savings tilts towards the orbit-first craft.

Craft #1: Cost: 9972 Kerbucks; deltaV: 5412 m/s

Craft #2: Cost: 9842 Kerbucks; deltaV: 6685 m/s

I launched craft #1, into LKO then to mun, and arrived there with enough fuel to land, but not sure about come back, since i already used some of my upper stage fuel

I launched craft #2 vertically upwards and escaped Kerbin SoI before SRB's burnt out..., thus, still had full fuel in upper stage

Craft #2 is cheaper than Craft #1... so i dont know where you are getting your rules from...I also think it would be very difficult to fly Craft #2 into LKO first. But you are more than welcome to try!

[Superfluous J]

Craft #1: Cost: 9972 Kerbucks; deltaV: 5412 m/s

http://i.imgur.com/lvSXd5s.jpg

Craft #2: Cost: 9842 Kerbucks; deltaV: 6685 m/s

http://i.imgur.com/fHLNZcV.jpg

I launched craft #1, into LKO then to mun, and arrived there with enough fuel to land, but not sure about come back, since i already used some of my upper stage fuel

I launched craft #2 vertically upwards and escaped Kerbin SoI before SRB's burnt out..., thus, still had full fuel in upper stage

Craft #2 is cheaper than Craft #1... so i dont know where you are getting your rules from...I also think it would be very difficult to fly Craft #2 into LKO first. But you are more than welcome to try!

Escaping Kerbin's SOI is *NOT* landing on the Mun. Landing on Mun takes roughly 800 more dV than escaping Kerbin's SOI. You need to do the same thing to have a valid test.

Also, you're using FAR right? I'm using Stock right now.

I made a craft that cost 7782 kerbucks and has 6162 dV, and can just match orbits with Mun (I didn't bother getting an encounter) before running out of fuel. Launching that rocket straight up, it ran out of fuel before its apoapsis reached Mun's orbit. And both of these are with the more restrictive stock aerodynamics.

And as a bonus, you almost can't make a simpler rocket

Edited December 12, 2014 by 5thHorseman
Changed "Periapsis" to "Apoapsis"

[arkie87]

Escaping Kerbin's SOI is *NOT* landing on the Mun. Landing on Mun takes roughly 800 more dV than escaping Kerbin's SOI. You need to do the same thing to have a valid test.

I think you misunderstood what i was saying. I didnt actually land on the Mun in either case. What I meant is that with the first craft, i had to tap into upper stage just to get up to the Mun's altitude, whereas with the second craft, i left Kerbin SoI before even tapping into upper stage. Thus, second craft has more deltaV (this is clear from KER anyway so it's pretty obvious).

I made a craft that cost 7782 kerbucks and has 6162 dV, and can just match orbits with Mun (I didn't bother getting an encounter) before running out of fuel. Launching that rocket straight up, it ran out of fuel before its apoapsis reached Mun's orbit. And both of these are with the more restrictive stock aerodynamics.

Did you limit thrust of SRB for both cases? Clearly, the vertical approach will perform better without the thrust being limited, but to get to LKO, you might have to limit thrust.

Also, using FAR is very important, since spaceships will not spin around violently with stock aerodynamics if you perform too-aggressive a gravity turn.

Once again, the main point of the thread isnt to claim vertical ascent is more fuel efficient, but rather, in career mode, before you unlock reaction wheels and winglets/fins with control surfaces, it is very difficult to perform gravity turn using torque from command pod alone (i think maybe you have yet to experience this since you use stock aerodynamics rather than FAR).

You said before that you "design your rockets with this [ability to turn] in mind". I would like to ask how? If you use LV-T45, the it quickly becomes more expensive in terms of Kerbucks than simply strapping MOAR BOOSTERS and using vertical ascent. And you are not allowed to use control surfaces or reaction wheels...

Do me a favor: remove the winglets from your ship and try to get it into LKO efficiently using FAR (you may reduce thrust)

Then launch vertically with no thrust limit and see which has more fuel left.

I think that is a good test.

[Jarin]

I have never had trouble with a gravity turn in FAR, even early-tech-tree. One LV-45 surrounded by thrust-limited SRBs, run the 45 at very low thrust for vectoring purposes and shut it off otherwise. THe pod's torque is enough for minor corrections and course following most of the time.

[Superfluous J]

I think you misunderstood what i was saying. I didnt actually land on the Mun in either case. What I meant is that with the first craft, i had to tap into upper stage just to get up to the Mun's altitude, whereas with the second craft, i left Kerbin SoI before even tapping into upper stage. Thus, second craft has more deltaV (this is clear from KER anyway so it's pretty obvious).

I did misunderstand you, but you did say you could have landed one on the mun and the other not. And in a Mun craft it's important to not overshoot your target

Did you limit thrust of SRB for both cases? Clearly, the vertical approach will perform better without the thrust being limited, but to get to LKO, you might have to limit thrust.

I limited the thrust to 54 in the orbiter, and set it to 100 for the straight up craft. I was going to try it limited in the straight up craft but never did.

Also, using FAR is very important, since spaceships will not spin around violently with stock aerodynamics if you perform too-aggressive a gravity turn.

It's also important because you can essentially go as fast as you want in the atmosphere and not experience the severe drag you get in stock.

You said before that you "design your rockets with this [ability to turn] in mind". I would like to ask how? If you use LV-T45, the it quickly becomes more expensive in terms of Kerbucks than simply strapping MOAR BOOSTERS and using vertical ascent. And you are not allowed to use control surfaces or reaction wheels...

I've found that one LV-T45 can turn just fine in stock. And the basic non-controlling fins hold a (real) gravity turn in FAR just fine if you do it just right. It may take a few tries but once you know how the rocket climbs it's pretty easy.

Do me a favor: remove the winglets from your ship and try to get it into LKO efficiently using FAR (you may reduce thrust)

I don't understand why you're allowing the largest SRB but not the simplest controlling fins, but if I can use the non-controlling ones sure, I can do that.