Over 400 hours in KSP... and apparently I have never orbited correctly.

[Snark]

57 minutes ago, SpaceCommunism said:

http://imgur.com/a/U7YZL

http://imgur.com/a/AyJNK
This would be a good example of the problem I had to work on this one for a long time the other night just to get it up into orbit so that it wouldn't flip. COM is obviously closer to the bottom around the engine for the no atmos stage. For reference those solids on the side are from interstellar and have gims (also set to 40% so they last till I clear atmos), the attached AV-R8 are scaled up to 200% and the booster are set in a separate group to trigger when I start to have difficulties with control. The central thruster is a t30 tweaked to 3.75 on an x200-8 tweaked the same and the tanks around it are 2 stage asparagus with regular t45s (just for the gim thrust) The 4 nacelles on the top are just for the orbital stage, they are fuel tanks setup in a 2 stage asparagus with no engines feeding to an x200-8 tank that has a single terrier scaled to 2.5.

Even though I was able to use this rocket to achieve the mission this is the best working model of all the ones I had that kept tipping over I could never do the gradual side shift you guys are talking about or even keep it going without flipping most of the time, Clearly the COM is on the bottom but always above the drag however as I drop the solids and the first liquid stage I have a massive shift that causes me to loose control and usually flip.

The reason why I am asking general tips is because this is a very typical design style for me so if you have advice it may be generally helpful.

A few things.

First, and overwhelmingly, the rocket is way too short and squat.  You want it tall and skinny.  Making it short and squat like that is far too draggy, and it also will hurt your stability because the things you have available for controlling your direction (fins, engine gimbal) aren't very far behind your CoM and therefore don't have much lever arm to work with.  "Fins on the back" only helps you if your CoM is a lot higher than the fins.

(Are you sure those SRBs are from Interstellar?  They look like the SpaceY radial boosters to me.)

The other thing is that your TWR looks rather on the high side-- 1.83 is really high, especially for a very squat draggy design like that.  You'll go too fast too soon.  Lower, like around 1.5, would serve you better and save mass on dead weight.

The fact that your payload up front is super draggy is just going to make matters worse.

Can you describe just what it is that you're trying to accomplish with the ship?  i.e. what's the actual payload and what are you trying to do with it?

[foamyesque]

3 hours ago, SpaceCommunism said:

typically that works ok until I drop boosters then it flips same with when I go to drop depleted asparagus staging. I am thinking this is because of reduction of drag from the bottom portion but if I have to not use boosters or asparagus staging, I am definitely not saving on delta-V.

 

Asparagus staging does not, as such, save you deltaV. It lets you get the same amount of deltaV for less mass, which can be useful, but usually the penalties in design stability and construction hassle aren't worth it. 90%+ of launches can be done with a 2.5 stage approach: Payload -> upper stage (Terrier, Poodle, or Rhino depending on size) -> Sustainer stage (Reliant/Swivel, Skipper/Mainsail, Mammoth) + cofiring boosters for launch TWR. Generally only a pair of boosters is needed to bring up the TWR for launch, and by the time they detach your sustainer stage has enough TWR to put your upper stage into the upper atmo and take full advantage of its higher Isp without worrying about the lower TWR. It's simple, easy to build, easy to fly, and effective.

[Provisional Name 12]

The purpose of that one was to launch those satellites (comsat for LKO on remotetech mod) to a 2kkm orbit, take a tourist to LKO and do a few rescue missions. As such the payload has to have the satellites on it, I tried putting a pharynx over it but it would smash off of the fuel nacelles and break my solar panels.

I think your right, boosters from space y.

These are some helpful hints, for some reason my response to having tipping problems was to try to make it shorter and stalkier because it seemed to give more control and prevent tipping. I hadn't payed much attention to TWR because I frequently just controlled my throttle instead, however it seems like maybe this is less ideal as the aero forces are not nearly as bad as the gravity forces. You guys have given some interesting advice and I will look at how I can use it to make more stable designs and perhaps get this stuff more under control. Thank you.

Edited July 20, 2016 by SpaceCommunism

[Snark]

4 hours ago, SpaceCommunism said:

The purpose of that one was to launch those satellites (comsat for LKO on remotetech mod) to a 2kkm orbit, take a tourist to LKO and do a few rescue missions. As such the payload has to have the satellites on it, I tried putting a pharynx over it but it would smash off of the fuel nacelles and break my solar panels.

Okay then.  So the payload's not doing a Mun landing or something.  In that case, you really want to lose those radial 4-ton fuel tanks on the top payload section, they're killing you with drag.

Here's what you do:

1. Start by designing your top payload section.  Keep it slim and trim.
2. Put a fairing around it.  You mention that this has a problem with smashing stuff?  Well, let's figure out why and fix that.    The fairing will take care of the aero on those satellites.
3. Under that payload (i.e. under the fairing), have a single stack, all the way to the bottom.  The stack can grow to whatever diameter you need, based on payload mass.
4. Some radial SRBs around the center stack ought to do just fine.

A design that goes something like:  radial SRBs off the pad, then a large central engine on a big fuel tank, then a 3rd stage to circularize the orbit. That should be all you need.  Tall, skinny, probably a small fraction the mass of what you've got there.

4 hours ago, SpaceCommunism said:

I hadn't payed much attention to TWR because I frequently just controlled my throttle instead

One thing to bear in mind:  Ideally, you'll be at full throttle the whole time.  Why?  Because if you're running at less than full throttle, it means you're carrying more engine than you need.  Which, in turn, means you're lugging unnecessary dead weight to space, which is a good thing to avoid.

 

[Edit] ...So, just as an example, here's a sample rocket I banged out.  It uses SpaceY parts, since you mentioned you have that.

The payload inside the fairing is 19.25 tons.  This rocket can deliver that payload to 90 km circular orbit, with 600 m/s of dV still remaining in the big orange tank.

From the bottom up:  That's the 2nd biggest 2.5m SRB (SpaceY), with four Delta Wings around it.  (No ailerons or other control surfaces, it just uses the small gimbal on the SRB.)  Atop that is a 2.5m decoupler, then the SpaceY engine that's intermediate between a Skipper and Mainsail in power. Then the stock big orange tank, and the fairing.  The SRB is set to 75% thrust to give it a launchpad TWR of just over 1.5.

It's steady as a rock, practically flies itself.  Just tip it slightly eastwards almost immediately upon takeoff, and set SAS to "hold prograde".  Don't have to do anything but stage when the SRB burns out.

If 19.25 tons isn't enough payload for you, you could always go up to a 3.75m stack, but you get the general idea. 

[Norcalplanner]

Thanks for linking to the other thread.  Just to echo and expand on what Snark said, the need for massive and complicated staging schemes is a thing of the past.  It's entirely possible to make practical rockets using stock parts which fly nicely and still have lots of delta V.  Here are two examples I did half a year ago, but the aero has remained essentially the same:

Both of the above examples were before my recent revelations regarding TWR and minimizing gravity losses.  If I had to redo these rockets today, I wouldn't be thrust limiting the kickbacks.

One thing to keep in mind with the stock aero model is that radially-mounted parts now add significantly to drag, as do parts of mismatched diameters (such as putting a 1.25m engine beneath a 2.5m fuel tank.  Struts and fuel lines now have a lot of drag, and the blunt nose cones are a lot worse than the advanced ones.

If you're launching a bunch of unaerodynamic satellites, then the solution is to put them all beneath a fairing.  It's also possible to stack satellites using jr. docking ports or stack separators - here's an example I used in the guide posted in my sig.

  

Edited July 21, 2016 by Norcalplanner
clarified things

[ag3nt108]

force of habit I almost always wait until I hit around 100m/s to start gravity turn, except on very early career craft that are super light and simple.  These craft I aim at 90 right of the pad.  Some very heavy craft with absurd pay loads (for instance when I'm launching station parts) I might wait until around the 7km mark as otherwise I find the AOA can be too much for a top heavy load (rocket flips).  Ive seen other players engineer around that problem but for me a little extra DV works just fine.

 

 

[Zamolxes77]

On 7/20/2016 at 8:55 PM, tjt said:

How much energy are you losing by pushing your way through all the additional atmosphere? I've been trying various ascent profiles too and always wondered at the tradeoff between the benefits of having a smooth ascent curve against the benefits of getting out of thicker atmosphere.

Negligible. This is 1.1.3 not 0.25. Soupy atmosphere is gone and you don't go that fast low in the atmosphere anyway. At 10 km you can expect some 300-400 m/s and at 40 km perhaps 1200-1500 m/s. With the ascent profile I posted, you might get some heating effects around 30-35 km, but you save a crapload of delta V. Try it out, the more vertical you go to "escape quicker the thick atmosphere" the more fuel you waste.

[FyunchClick]

23 hours ago, Clubbavich said:

 It's this one! 

Lol, now to post in the other thread directing here again.

[Kar]

On 07/07/2016 at 4:24 PM, Snark said:

You can make your life easier by designing your rockets so that they always have the same launchpad TWR.  Just pick a number that works for you (I like to use 1.5, myself), and religiously keep your designs to that launchpad TWR.  That way, they all have a similarly shaped gravity curve and the piloting for that critical, initial eastward nudge will be similar.  (You may want to adjust that for extreme discrepancies of scale:  really big ships are less affected by aero drag, so you may want to run them at higher TWR, which in turn would be a different shape to the curve.)

1.5 TWR err... ok.... lets assume that's the measure for the 1st stage (boosters, orange tanks, asparagus setups, whatever) but what about the payload? 

How you keep it if you launch a satellite with 10 parts or a big huge landing module with 50 parts?

What about TWR in stage 2, 3 or how many you need? How to keep efficiency with the upper stages?

Design each stage individually? I also, use a lot of mods, some of them that turns the game a bit hard (Remote tech, life support, scansat, and many others) and a couple for info and control, MJ and KER.

[Snark]

3 minutes ago, Kar said:

1.5 TWR err... ok.... lets assume that's the measure for the 1st stage (boosters, orange tanks, asparagus setups, whatever)

Yes.  That's when you need the highest TWR.

3 minutes ago, Kar said:

but what about the payload? 

What about TWR in stage 2, 3 or how many you need? How to keep efficiency with the upper stages?

For upper stages, you need a lot less TWR.  For example:  the orbital insertion stage can have a TWR of significantly less than 1... but if your launchpad TWR was <1, you wouldn't even get off the ground. 

Upper stages can get by with lower TWR because, 1. they're going mostly sideways rather than up, so they're not fighting gravity, and 2. they generally have more time available to thrust, so they can spread the dV out more.

Your 1st stage should always be your highest TWR.  Your 2nd stage's ideal TWR will depend on what part of the flight it gets used on-- i.e. does it kick in when you're already going mostly sideways, or does your 1st stage burn out while you're still going pretty close to vertical?  The former would need a lower TWR than the latter.

So it's a lot harder to give a general rule of thumb, more specific than "it depends".  But in general, each successive stage should have a lower TWR than the one before it, and once you're in orbit, a really low TWR is fine.  Both because it needs less TWR, and also because having more TWR than you need means you're lugging unnecessary dead weight in the form of extra engines, and the higher those are up your stage stack, the bigger the price you pay in terms of overall ship mass.

[Kar]

My main problem in my staging TWR was (is) always in the 2nd or 3rd stage. I already know I do not make many efficient ships (or landers) simple because I like to have a multi-purpouse lander that can hop around in the mun or mimnus. So my 1st stage is kinda to get all the thing of the ground due to the massive stuff I put on my rockets (most of the times I can get around 300m/s around 10k) but then my 2nd stage is the one that really needs to burn more time since usually, is the one that take all that stuff in to space. I like to use a 3rd stage to orbit, and keep a bit of DV to initiate a transfer burn, then it separates in the middle of the transfer and the (big) lander does the rest of the mission alone.

I play also with stage recover, so I need to recover most of the stuff I can except the 3rd stage. Usually I use for a 3rd stage a rockomax x200 32 fuel tank with a Poodle.

I know this is not an efficient design, but it serves me. Sometimes I play around with the designs to try new things, but I always played KSP in career mode, with the mods I use I need to recover everything, grind some boring missions for cash, and "re-use" (read: stage recover) almost 100% of what I fly.

My 1st Stage TWR is always 2 or a bit more, just because I have (probably the bad habit) to try fly the thing a bit more while controlling my speed with the throttle.

God.... too many variables in my game.

[Zild]

On 07/07/2016 at 8:48 PM, JoeNapalm said:

 

Any orbit you can walk away from is a good orbit.

 

-Jn-

Surely you mean any orbit you can spacewalk away from...

[flatbear]

while I do do gravity turns, I have to turn at around 15-20km or else my ship will flip from aero

[Hevak]

This will be a bit long sorry, but I feel it must be said. 

@MacLeod-Industries first and foremost, don't ever let anyone tell you that you are doing anything wrong, incorrectly or that there is a correct way to do anything in ksp ever.  There simply is no wrong or correct way. Ksp is a game, and on top of that it is a sandbox game. The most important thing is that you enjoy playing and have fun, and that you achieve what you want. This is evidenced by the  fact that one user here almost only makes trains, others rarely go to space and simply make atmospheric aircraft, but you know what that is what brings them fun.

now with that being said, there is a best, or better ways to achieve things. And there are ways to make some things easier, depending of course on your skill piloting your skill building craft. But keep in mind many of the people telling you the most efficient way to get to orbit are building craft made to fly the way the describe. 

There is not one single, perfect, "correct" ascent profile, unless you build your craft a certain way. Depending on what and how you build craft, your ascent profile may need to change. Of course this is dependent on your skills as mentioned above. 

Now if you want to push for the best, most efficient way to get to orbit, then I would definitely pay attention to a few of the posts discussing it in terms of efficiency, I find this difficult to achieve, and not fun as well. I play a game to have fun, not to do complex calculations, build a specific vehicle with specific numbers or twr and such. That doesn't bring me enjoyment, some people enjoy that aspect of the game though, and they thrive on doing the real math and stuff and achieving these results. 

If you were to experiment with various gravity turns and ascent profiles, you will find that the delta v difference is very minimal overall between them. Now this delta v could matter depending what you are doing sure, it also makes things easier to either have more delta V after orbiting, or even building a craft with only the needed amount of delta v vs. over doing it. Now your original way of getting to orbit would probably show a bigger difference in delta v usage compared to an ascent with even a sub-optimal gravity turn, but the difference between when you start a gravity turn will tend to be minimal.

as one person mentioned here, the more important aspect is horizontal speed, and the reason I feel you want to worry about this more than anything is not because of delta v usage, but because it will make your circulization burn much much easier if you get some horizontal velocity going. the closer you get to orbital velocity horizontally, the shorter and easier your circulization burn will be, and the easier it will be to have your Pe and Ap close in altitude which is a much nicer orbit in most cases.

my usually gravity turns start anywhere from the 4,000 meter, to 8,000 meter range depending on the craft I build and the orbit I'm going for with said craft. I try to be close to 90 deg flight well before my Ap is near what I want while balancing it enough that I'm not generating too much heating effects in atmosphere. How gradual my turns are also depend on my craft as well, but you always want to try not to get your nose marker too far  from your prograde marker. Think of it more as just slightly pulling that prograde marker over, don't just go yanking on it. I have never once been able to make a craft the can execute its own gravity turn as some people say they do. But that is due to my skill as a builder and I know this, 

Now some people will slam this as incorrect, which it isn't because it works for me and I get to have fun doing what I want in ksp. I got into orbit and completed my mission I intended, so it worked. Now it certainly isn't the most efficient way, but as I said I'm not worried about that. I am not great at building spacecraft in ksp, and I especially have trouble when trying to launch any kind of payload into orbit. 

So my advice is to just take any information you get from the forums with a grain of salt, or a pile of it even. Try some things out that people mention of you want to and experiment a bit. The best way is to try and build a fairly decent and controllable craft and use that one over and over again with various ascent profiles and see what works for that craft. Then maybe try a few more different craft and see if it requires a slightly different approach. And also note the amount of delta v left over after trying the different ascent profiles. And find what works best for you overall. I can't achieve these perfect or so called "correct" gravity turns people tout, but I have found what works for me and is easily and consistently achievable. 

Also decide how serious you want to take things in ksp. Do you want to push for high efficiency, best ways of doing things? Or do you want to just have some fun, build what you want and do what you want with Ksp? The best or perfect efficiency in doing things is in no way required to achieve anything in ksp. But achieving a decently efficient even if sub-optimal way to orbit will certainly help a bit of course.  I'm pretty bad at building and flying things at times. But I have no problem getting around to most of the kerbin system and even building bases on mun and minmus as well as giant space stations in orbit using part mods. 

So certainly don't disregard some of the things said here, but don't ever feel you have to do things the way people say are the best, or real way, or the way the math says it should be done. First and foremost find what works for you, and then if needed or wanted you can always refine your techniques and/or look into more efficient ways to do it if you want to worry about that. 

Above all, just find what makes you happy and allows you to do what you want in ksp. And never, ever let someone tell you that you're doing something the wrong or incorrect way. There simply is no such thing in this game.

Edited July 25, 2016 by Hevak
Errors

[W. Kerman]

Sees this, "OOO, gravity turn", flips over rocket in 15 seconds, ragequits.

[Norcalplanner]

I'd be happy to help if you provide a picture of the rocket.

[numerobis]

One fun trick: use launch clamps to hold your rocket one shift-tick away from vertical (5 degrees off vertical) and launch with three SRBs and maybe a few fins as your first stage -- not even any vectored thrust. On launch, just stay prograde. You won't flip, because you have mass up front and  drag out back.

That will get you a bit too steep of a first stage ascent, but for very cheap!

[JAFO]

On 08/07/2016 at 8:21 AM, Clubbavich said:

Are there any mods out there that track gravity losses and aero losses during ascent? Would be interesting to see so I could refine my designs based upon first-hand experimental results. It would probably make efficient ascents more instinctual as well. Would be nice to see an accumulated total of dV lost to gravity and drag, as well as an "instantaneous" output of how much loss there was in the last second.

The GravityTurn mod does this, in addition to making your launches as dV-efficient as possible.. (see pic below)

On 20/07/2016 at 10:21 PM, Zamolxes77 said:

Try this. Do your normal launch, tip slightly and make sure by 10 km you're at 45 degrees. You can hold it there for a while, until your TIME to Apo is over 1 minute. Once there, hold prograde while at same time, reduce throttle so your time to Apo stays the same, aprox. 1 minute. Once you reached your target Apo height, tip to 90 degrees and continue burning horizontally, while maintaining the time to Apo constant. Obviously, once you hit 70 km mark, you can shut off engines and do next burn 5s to Apo.

What will happen is basically combine the circularize burn and take off burn INSIDE the athmosphere, so you take full advantage of Oberth effect, will have to expend very little dV to finish off circularization. Using this technique, you can take crafts into orbit that seemed impossible before. 3400 dV, it can be done with 3300 dV after a bit of practice, even less.

This really is the key to efficient launches.. using the throttle to maintain a constant time to Ap throughout the launch. This is how the GravityTurn mod works. By using GravityTurn set to maintain a constant 40s to Ap, my personal best to a fully circularised 80km orbit is 2924.8 m/s. (see pic below)

Whilst I'm not entirely sure (it was several months ago), I seem to recall that this also unintentionally became my first ever rocket-SSTO, in that the entire ship made it to orbit.

On 21/07/2016 at 3:55 AM, tjt said:

How much energy are you losing by pushing your way through all the additional atmosphere? I've been trying various ascent profiles too and always wondered at the tradeoff between the benefits of having a smooth ascent curve against the benefits of getting out of thicker atmosphere.

As has been mentioned several times above, a smooth ascent curve is far more important than quickly getting out of the lower atmosphere. For much of the launch, this ship was a huge fireball pushing through the atmosphere, and some parts got seriously hot (temperature bars popping up all over the place, several of them getting up to around 80% of the bar), which was very scary, but nothing exploded, and the end results speak for themselves. 2888.7 m/s to reach 70km, at which point the orbit was almost circular already, needing just 36.1 m/s to circularise. Total loss to gravity at Ap was 566.64 m/s, whilst loss to atmospheric drag was just 103.7 m/s, despite more than 10 minutes spent in the atmosphere all-told.

Edited July 29, 2016 by JAFO

[magnemoe]

On 21.7.2016 at 10:51 PM, Zamolxes77 said:

Negligible. This is 1.1.3 not 0.25. Soupy atmosphere is gone and you don't go that fast low in the atmosphere anyway. At 10 km you can expect some 300-400 m/s and at 40 km perhaps 1200-1500 m/s. With the ascent profile I posted, you might get some heating effects around 30-35 km, but you save a crapload of delta V. Try it out, the more vertical you go to "escape quicker the thick atmosphere" the more fuel you waste.

main issue is more flipping over than drag even if launching insane stuff like this, notice the large aircraft tails as fins. 

I uses an 2.5 meter stack on purpose to make the rocket longer, 4 long SRB for TWR, mechjeb launch with turn angle 60. the base has an TWR of 0.6 with the fuel it need to complete orbit. yes it uses over 4km/s to reach orbit, too much gravity loss and high drag on this brute force solution to problem

[Zamolxes77]

1 hour ago, magnemoe said:

main issue is more flipping over than drag even if launching insane stuff like this, notice the large aircraft tails as fins. 
 

So find other solutions to the problem. The reason you have problems with flipping is having the thrust so far back and during ascent, as the fuel gets consumed, your CoM shift forward, way way too forward. A slight bump and you introduce a torque in your forward vector, instantly flipping the thing.

Instead of adding fins, try to build your ascent stage AROUND your payload, so CoM doesn't shift much during ascent. Like this for example:

The white rockomax tanks with all the engines are in fact the payload, while the orange tanks and the SRB's are ascent stage. Not only I use the engines of the payload for ascent, but by shifting the main stage around the payload I don't even need fins, Vector's gimbal is enough to steer AND I only ditch empty tanks, not expensive engines.

I put that craft into orbit easy enough, albeit bingo fuel, so it will require a refueling mission, but its a long range refinery able to refuel itself once landed on a body. Note the parachutes, it can land safely even on Kerbin but won't be able to take off again, in Kerbin's case. Parachutes are there to safely bring back a 250,000 kerbalbucks piece of machinery, when eventually I design a better model.

Edited July 29, 2016 by Zamolxes77

[pandaman]

4 hours ago, Zamolxes77 said:

So find other solutions to the problem. The reason you have problems with flipping is having the thrust so far back and during ascent, as the fuel gets consumed, your CoM shift forward, way way too forward. A slight bump and you introduce a torque in your forward vector, instantly flipping the thing.

Instead of adding fins, try to build your ascent stage AROUND your payload, so CoM doesn't shift much during ascent. Like this for example:

Actually that's incorrect. 

Ideally your COM needs to be forwards and your COL at the back.  What causes flipping is the COM moving back towards the COL as the fuel gets used.

The vessel 'pivots' around the COM and as it moves back the fins and gimbals have to work harder to maintain stability then reach a point where they don't have enough leverage to do it anymore and the rocket flips. 

The 'optimum' configuration for a rocket is long and thin with the weight at the front and the drag at the back (darts fly with the heavy bit at the front and the feathery bit at the back).

[GarrisonChisholm]

On 7/7/2016 at 10:40 AM, MacLeod-Industries said:

Oh I know, I just didn't have the right altitude offhand...

and I accidentally jettisoned the retro-rocket before I could change it

Here is the rocket:

 

 

My friend (dude!), you're calling that crap, but I actually *like* your usage of fairings!  I have always felt married to capsule diameter, and so my orbiters always end up looking like Apollo, even if I don't need anywhere near that much fuel or engine.

The next game I start, I think I will remember this look for my early attempts!

[GarrisonChisholm]

On 7/7/2016 at 11:35 PM, Dfthu said:

Oh. Was somebody inspired my thread?

Just remember, henceforward, if I *ever* like your posts, it is because I am hearing them in Ahnold's voice.  :]

(...unless you say something of divine brilliance or wit, in which case it is for that reason first, & *then* because I hear it in Arnold's voice)