What am I doing wrong?

[TubaHorse]

So, I've always known that KSP isn't an easy game, and I'm not the best at it. I kinda picked it up and fiddled until I learned it. However, I feel I'm lacking something. If I ever try to go to any other body that isn't the Mun or Minmus, the mission never works out. I've watched countless videos seeing people do things that seem impossible, and yet I can't get to Duna with an overpowered rocket without cheating. Just recently I installed Space Y and some near future mods. I made a rocket to send to Duna, waited for a window, and went on my way. Things went great for the most part. Getting into orbit was easy, circularising was no problem, and the Duna encounter was a piece of cake. However, when I got there (after a 6 min. burn with 7 nuke engines), the game told me it would take a 15 minute burn to get into Duna's orbit.

 

So, what am I doing wrong? I really want to enjoy this game and eventually land on every planet/moon, but it seems currently it's just a $30 Kerbin orbit simulator.

[Streetwind]

Welcome to the forums!

It's difficult to say just from that description what exactly is going wrong. But just to make sure I understand you correctly: you prepared a spacecraft for a Duna transfer window, planned and executed a maneuver node at the proper time, successfully got an encounter, and travelled all the way to Duna. Then upon arriving, you discovered that the capture burn exceeds your remaining fuel, and in fact exceeds the initial transfer burn by a factor of 2.5. Right?

You can help us help you by providing some extra information. For example, specifying burn magnitude via the time your engines are lit isn't very useful. It depends 100% on the specific construction of your spacecraft. Instead, could you give us the delta-V figures? That's the number (in m/s) you see when you make a maneuver node. Delta-V (a shorthand for "change in velocity") is the universal term to describe the reach of a rocket, and how much effort is required to go places in space. How much did you spend on your initial transfer? How much would it cost you to capture at Duna?

Additionally: could you tell us the ingame date (as seen on the space center screen) when you departed, and the date when you arrived?

And if you want to be really fancy: could we see your arriving trajectory through Duna's SOI in a screenshot, ideally showing the node you made for your capture burn?

 

Edited March 27, 2017 by Streetwind

[Dark Lion]

42 minutes ago, TubaHorse said:

So, I've always known that KSP isn't an easy game, and I'm not the best at it. I kinda picked it up and fiddled until I learned it. However, I feel I'm lacking something.

By the sound of things, you've still more yet to learn! Meaning no insult, what you're doing wrong is asking the wrong questions. You know what went wrong with your injection burn at Duna... you ran out of fuel!  The real question, my musical, equestrian friend, is how to take more fuel along with you!

Our friend above, Streetwind is right though... if you'd like some solid, viable advice in your rocketry designs, we need to see the subject in question. Did you know you can screenshot KSP at any time with the F1 key? Just make sure you don't press F2 first, so we can see all your HUD information too.

If you have Kerbal Engineer Redux, even MORE useful information could be shown, provided it's properly set up... this would help us help you in ways we couldn't with a stock KSP. Or to simplify things for yourself the most, just record your attempt and post a link to the video here so we can properly scrutinize it. If you'd just rather figure it out on your own, I couldn't suggest Kerbal Engineer Redux any more strongly. It doesn't just give you deltaV approximations like you see in the stock KSP manuever nodes. There are a myriad of calculations going on in simulation that aren't shown to you. KER gives you a heads-up display readout or pop-out window for ease of access, and even offers a great deal of in-editor calculations as you build, such as your total deltaV, thrust offset torque, thrust-to-weight ratio and many more, most of which can be modified by the specific celestial body you wish to encounter. That's really more helpful for complex designs, but it makes rocket-building easy-- well... a lot easier than just "eyeballing it" and hoping for the best. An educated guess is better. Less Kerbal, but more survivable. 

By the way, mods like Space Y and the Near-Future set of mods aren't intended to make the game easier in any way, they only make certain objectives easier to obtain. So don't think you're terrible at KSP... I've been playing a little over a year almost non-stop and still haven't successfully visited every moon and planet. Hey, even NASA sucks at making spaceplanes! LOL don't hurt me Realism players, I know Earth is much bigger than Kerbin. I'm just saying the shuttle program didn't work out very well, but they finally found another, better way, with time and testing.

Good luck TubaHorse! Hope to see your work at Duna soon!

[TubaHorse]

I really appreciate the responses. I actually did put together a video of a second attempt that (to no surprise) ended in failure. I, as you said, have a lot to learn and have literally no clue how to calculate and use ΔV in any way. I have Kerbal Engineer redux installed and I hope that I have the right window open for info. But I hope you can understand the frustration and disappointment I feel when I see people like Matt Lowne doing amazing things and I try to do similar things and end up never being able to get the simple stuff done.

The video will be uploaded sometime in the next few hours. My upload speeds are horrid, even with ethernet. It's not sped up or edited, my apologies, but hopefully you should be able to hop though and find the information you need. Also, have a good time laughing at some of the mishaps as well.

[bewing]

Oh yeah, we all understand the frustration to some degree.

Some people are incredible pilots on EVA. Some people really understand how to minimize drag, and build rockets that have incredible performance. Some people just know when the best launch windows are.

Most of the rest of us just do things by the seat of our pants and carry some extra fuel.

 

[Streetwind]

5 hours ago, TubaHorse said:

...literally no clue how to calculate and use ΔV in any way

Then while we wait for your upload, why not tackle this subject in the meantime? It is surprisingly simple, you will find.

The reason why we use delta-V at all is because we need some sort of measurement for "the distance" between two objects in space. Like you'd have in a car, where you know the distance between two places and decide whether you can do it on your current fuel level or not. But in space, distance as such doesn't work, because distances constantly keep changing. All things in space are always moving; it is physically impossible to not move in space. So instead, people came up with the idea that if you change your current speed and direction (AKA velocity) so that you have the same direction and the same speed as the target object, you are essentially "already there" - provided you performed this maneuver at the right time and at the right point in space. And voila: change in velocity is now the only performance characteristic you need to know of a spacecraft to figure out if it can go somewhere. The "when" and "where" is independent of that, and can be determined separately (see: transfer window).

The calculation is done via the Rocket Equation: dV = 9.80665 * Isp * ln(mass fraction), where the mass fraction in turn is the wet mass of a rocket stage divided by the dry mass of the same rocket stage. If you look closely, you'll notice that this complicated looking formula is in fact just a constant multiplied by a quasi-constant multiplied by something your calculator gives you. If you get comfortable with that fact, you'll quickly learn to do it on the fly. And now comes the point where I once more link my most popular wall of text of all time, in which you can learn about how mass fractions work.

Edited March 27, 2017 by Streetwind

[bewing]

You also don't really need to do the math if you don't want to. Just use SetOrbit to put the ship in a very high orbit around something (Minmus at 1Mm works fine). Make a note of the initial velocity. Light your rocket and burn until your fuel is all gone. Note ending velocity. Subtract.

 

[Kryxal]

Align prograde before that burn, since off-angle thrust will just add vector geometry to it all.

[Aegolius13]

If your burn to inject at Duna was too big,  that probably means you didn't get the transfer just right.   More specifically, you probably burned too much leaving Kerbin, and did not have the ejection angle quite right, meaning your ship was rocketing past Duna during the encounter.   

It may be helpful to look at diagrams for a Hohmann transfer.  You want your transfer orbit to be to exactly tangent to that of Duna.  If your orbit goes further out than that,  it means you're burning excess fuel on the way out of Kerbin, and will have to burn excess fuel coming into Duna since you're passing by at higher speed.  The ideal launch window is generally where Duna will be right at that tangent point when your orbit hits it.

One other thing that might be an issue: as you approach Duna, you want to set your periapsis very close to the planet.  This is the most efficient way to capture,  due to the Oberth effect.  

Duna's thin atmosphere is also good for aerobraking, which can often save you from having to spend any fuel on the capture.

[TubaHorse]

6 hours ago, Streetwind said:

Then while we wait for your upload, why not tackle this subject in the meantime?

The video is up! 

 

 

[FancyMouse]

Looks not terribly bad - the only missing step is a mid-course correction burn to bring your Duna periapsis down to low altitude when you're still far away from Duna. You would want to burn in low altitude to take advantage of Oberth effect, and that's what every delta V map assumes. If you do your capture burn from high altitude, it's no surprise that you'll take much more fuel than expected.

[FloppyRocket]

I'm not likely to watch an hour of video, but I wanted to say that's a gorgeous rocket on liftoff.

[TubaHorse]

27 minutes ago, FloppyRocket said:

I'm not likely to watch an hour of video, but I wanted to say that's a gorgeous rocket on liftoff.

Thanks, I appreciate it! I can design things well enough, but it's a bummer they never make it anywhere lol

[W. Kerman]

1 hour ago, FancyMouse said:

Looks not terribly bad - the only missing step is a mid-course correction burn to bring your Duna periapsis down to low altitude when you're still far away from Duna. You would want to burn in low altitude to take advantage of Oberth effect, and that's what every delta V map assumes. If you do your capture burn from high altitude, it's no surprise that you'll take much more fuel than expected.

That's a trick I use, because it works really well- However, I only realized how to do this after watching Wexus on YouTube. 

You can 

1. Zoom in on the planet you're visiting  in map view

2. Turn down thrust to prevent "accidents"

3. Set your direction to one on the navball (If you can't see it, it's hidden with a pull tab at the bottom of the screen

4. Burn, and learn how to adjust your orbit

5. PROFIT

[tater]

You're approaching Duna from a higher apoapsis it looks like (I skipped around the video). As a result, you are not just kissing Duna's orbit, you're moving past at a high closure.

 

Here's an image on the net that shows the rough geometry to aim at for when you leave Kerbin:

You'll notice that if you draw a rough tangent from Kerbin orbit, it's a almost pointing at Duna.

This Earth->Mars example shows similar geometry:

 

That's the first bit to try. You'll not have to warp more than about a year to try it, if you are on the wrong side to start. Stretch your maneuver node around until you get a Duna encounter. Don't worry about the specifics of it, honestly. Your orbit should not really exceed Duna's, or not by much.

OK, once en route, make a new node as a course correction.

As others have said, make a mid-course correction. After placing a node, click on Duna to "ficus view." You will now see where your encounter will actually pass Duna. Now select the node in the distance (since your POV is now Duna, not your ship), and adjust the node to move the flyby path to where you want it (closer to Duna, a lot closer). You can then execute that burn in Duna focus, and see where you will pass. If you put the periduna just above the atmosphere, you will have an optimal braking burn to achieve orbit.

 

 

[bewing]

1. You can probably gain another several hundred m/s of dV by improving your gravity turn technique on launch.

2. You can probably gain another 300 m/s of Oberth dV by leaving from a very low (73km) Kerbin orbit, rather than your 200 km orbit.

3. You threw away almost 500 m/s of dV when you chucked your booster stage with fuel left in it.

4. You can probably get a bit more efficiency out of your Kerbin ejection burn by burning prograde, rather than pointing at the maneuver node.

5. and like the others have said, a mid-course correction saves a lot of fuel at the end.

But in general, your technique doesn't suck at all.

 

Edited March 28, 2017 by bewing

[FloppyRocket]

1 hour ago, TubaHorse said:

Thanks, I appreciate it! I can design things well enough, but it's a bummer they never make it anywhere lol

From what the other posters here have said, I think you're being too hard on yourself. A few tweaks in the methods you're using, and you'll be flying successful missions.

[FancyMouse]

1 hour ago, bewing said:

4. You can probably get a bit more efficiency out of your Kerbin ejection burn by burning prograde, rather than pointing at the maneuver node.

I disagree with this one. It gives minimal benefit, but it can screw your transfer in a noticable way, especially when your burn time takes long - it will tend to make your ejection angle less than what you want (intuitively it's because it brings periapsis earlier than maneuver node). So unless you know what you're doing (e.g. you write your own small program to simulate long burn and solve an optimal maneuver), otherwise I wouldn't recommend pointing prograde to burn to escape.

[bewing]

6 minutes ago, FancyMouse said:

... but it can screw your transfer in a noticable way, especially when your burn time takes long

Well, if he takes my other advice, he's going to be burning to duna using his main booster stage at 10Gs, so it's gonna be a really short burn.

 

[Streetwind]

4 hours ago, bewing said:

4. You can probably get a bit more efficiency out of your Kerbin ejection burn by burning prograde, rather than pointing at the maneuver node.

I also have to disagree with this one unfortunately. I actually brought this up for discussion ages ago, and someone did some math wizardry. It showed that in order to have the smallest possible deviation from the target orbit (defined by the maneuver node), the best method is to point directly at the node as opposed to pointing prograde. The latter option more efficiently raises your orbit around Kerbin, but introduces an error into your solar trajectory that requires you to spend just as much as you saved - or even slightly more - on a correction burn. I can try and find that old thread again if you're interested, but I'm honestly not sure what it was called or when it was...

 

9 hours ago, TubaHorse said:

The video is up! 

Alright, let's see... starting with the positive:

- Your rocket is beautiful.
- The transfer is pretty good. 1100 m/s is usually what people budget for trans-Duna burns, and that's exactly what you paid. Even the most perfect transfer would probably only shave off another 50 m/s here, and honestly, that's peanuts.

Now, the point where it all goes wrong... that's your Duna capture burn. As others have already said, you're missing one crucial step, and that is the act of lowering your flyby periapsis with a correction burn en-route. Because you're capturing at the edge of the SOI, you're required to propulsively cancel out a majority of your speed around the Sun, only to let Duna's gravity accelerate you again as you fall towards it. Instead, you want to burn deep down in Duna's gravity well, and only cancel as much travel speed as you must in order to match speed with Duna. Basically, by burning where you did, you spent fuel for several days worth of acceleration through Duna's gravity that you didn't need to pay if you burned much closer to Duna. It's a weird concept to wrap your head around, but unfortunately I'm not finding the words to explain it better right now. What matters though is that a capture + full circularization in low Duna orbit should generally cost only about 600 m/s; you paid over 1,500 m/s. The difference is really that drastic.

Three quarters of the way towards Duna, make a node, and try to lower your flyby periapsis to something like 75 km. The light blue "radial in" and "radial out" markers are very useful for that kind of thing - usually more useful than the green prograde/retrograde markers.

 

Other than that - the only thing that hasn't already been said by others is that your rocket feels way too heavy. I don't know what you're trying to land on Duna, so I can't really tell if it's appropriate or not, but I don't think I've ever flown a Duna mission this heavy. Something that you learn as your player skill improves over time is the art of "downsizing" - of making your missions just as small as they need to be. This has a massive effect, due to something called "the tyranny of the rocket equation". That fancy term describes the fact that in order to lift more mass to orbit (or push it to Duna), you need more fuel, which itself has mass, so you need more fuel to lift that fuel, which itself has mass, so again you need more fuel... Long story short, rocket size grows exponentially with payload mass. So downsizing is good practice no matter where you intend to go.

 

Edited March 28, 2017 by Streetwind

[bewing]

6 minutes ago, Streetwind said:

It showed that in order to have the smallest possible deviation from the target orbit (defined by the maneuver node), the best method is to point directly at the node as opposed to pointing prograde.

This assumes that the maneuver node was created perfectly. Which it never is in stock KSP. The KSP reality is that a prograde burn has just as much chance of being more efficient than a manuever node burn, as it does of being less efficient -- when defined in terms of deltaV getting to your destination.

 

Edited March 28, 2017 by bewing

[FancyMouse]

7 minutes ago, bewing said:

This assumes that the maneuver node was created perfectly. Which it never is in stock KSP. The KSP reality is that a prograde burn has just as much chance of being more efficient than a manuever node burn, as it does of being less efficient -- when defined in terms of deltaV getting to your destination.

 

Nope - it was an agreement that burning prograde is minimally more efficient but you get an orbit with a larger error. For most practical purposes burning maneuver wins.

 

[Streetwind]

Sweet, you found the thread I meant!

[bewing]

5 minutes ago, FancyMouse said:

larger error

This is a subjective term. It's not necessarily larger if you just time it better than the maneuver node burn is timed. The thing you guys are refusing to acknowledge is that the manuever node burn itself is only an approximation. You can always beat an approximation if you just know what you're doing better than the approximation.

And if you watch the video for the amount of fine tuning done on the maneuver node for the Kerbin ejection burn, it was pretty minimal.

 

Edited March 28, 2017 by bewing

[Streetwind]

Now now, no need to get hot-blooded over such a trivial detail. Remember, we're trying to give TubaHorse advice on how to improve his trip to Duna. Arguing over which way he should execute his node not only relies on assumptions on our end, but it also ends up being completely irrelevant in the grand scheme of things. Even assuming that his node was badly planned (the dV cost disagrees, unless you further make assumptions about Duna's precise position), and further assuming that he (as a complete newbie to interplanetary transfers!) somehow has the ability necessary to blindly time a transfer by gut feeling better than when using a maneuver node, what does it save him? 5 m/s? 20 m/s? Remember: he's losing literally 1,000 m/s during his Duna injection, among other things. This whole node-following-or-not thing doesn't even matter in the slightest. So let's drop it and worry about more important things.

Such as: why is there an MX-F.L.A.T. reactor on the transfer stage? That thing weighs like *checks spreadsheet* 10 tons with radiators, and it never performs any function! No wonder the rocket felt much heavier than it should be...

Edited March 28, 2017 by Streetwind