1.0.2 - Rocket ascent profile and orbit delta-V

[eviator]

Especially now that thrust also varies with atm. pressure, and mostly because of the rather large difference in engine specific impulse now, you really can't just measure delta-v accurately by looking at the vac. delta-v before and after, but you have to actually measure the aggregate expended delta-v throughout the launch. There are mods for this.

This also goes for TWR, now more than ever before. I wish KER made the switch for atm. sea level quickly available in the compact view now that it's so important.

I wonder if this is the major confusion. Doing the measurements differently offers the possibility that my profile is just fine, and that I'm just not comparing apples to apples. I use the non-atmosphere numbers KER provides in the VAB and the left-over dV in orbit.

Unfortunately, this doesn't help me solve my original problem, which is trying to find good general rules for construction and ascent profile.

[icedown]

Here's my basic rules for rocket building and ascent profile. I do use mechjeb. I have done a little testing, but not to much.

First Stage - Minimum 1.3SLTWR

Stages after first 2500Dv - 1.2VTWR

Launch to orbit - 3900VDv - I usually use about 3700 of it.

MJ Limit acceleration: 18m/s for entire launch

MJ Acent Profile:

Turn Start: 1k

Turn End: 45k

Final Flight path Angle: 0

Turn Shape: 70 %

This follows the prograde vector fairly closely in the lower atmosphere, but up around 35k starts going below it.

This may not be the most efficient, but it produces consistent launches.

[NikkyD]

the whole "following prograde" greatly depends on the weight distribution of your rocket.

If you have a very long rocket the fuel is used from top to bottom so the weight shifts, you have a heavy head (payload etc) and a heavy rear (engine and fuel left).

Some designs just cant turn as quickly as others

Edit:

we should all use the same design, a simple few parts rocket with the same engine and same dV.

Edited May 10, 2015 by NikkyD

[xendelaar]

Great idea NikkyD! I will try to perform several experiments tonight. If anybody would able to design a general rocket, that would be great. Else i will design one myself.

I would suggest a rocket with a 5 tonnes payload with fairing, a fuel budget of 4000 delta v, 3 stages with 1.5 twr and an initial srb stage to get things going? Or would you guys recommend a different setup?

[Foxster]

How about this for a "Standard Rocket"?

Part list, top down:

Aerodynamic nose cone

Mk1-2 command pod

Service Bay (2.5m) containing 2 PB-NUK

X200-16 tank (payload)

Rockomax decoupler

X200-16 tank

Poodle

Rockomax decoupler

X200-32 tank with 4 Advanced Canards at the bottom

Skipper

Rockomax decoupler

Jumbo-64 tank with 4 Advanced Canards at the bottom

Mainsail

MJ says it has 4014 vac dV with useful TWRs at all the stages.

Dead easy to get to a 100km orbit and has a few hundred dV left.

Edited May 13, 2015 by Foxster

[andruszkow]

I generally go by this rule:

Surface-to-orbit on atmospheric bodies now requires 85% of the delta-v pre-1.0. That gives 3850 for Kerbin and it seems to be pretty spot on.

[problemecium]

I'm not in favor of a single standard rocket on account of the issue of different numbers of stages in different people's play styles. I, for example, almost exclusively use SSTO or almost-SSTO launch vehicles so that I can recover them SpaceX style. A standard test vehicle would yield very different results with one or two stages than it would with four, and those results would have varying levels of usefulness for people used to different setups.

We could, perhaps, have a few "standard" vehicles for different stage counts, e.g. an SSTO test vehicle, a two-stage test vehicle, and a three-stage test vehicle with SRBs.

Edited May 12, 2015 by parameciumkid

[eviator]

I'm not in favor of a single standard rocket on account of the issue of different numbers of stages in different people's play styles.

While rocket design is, of course, part of the equation, finding an optimal ascent profile with only one design is a good first step. It removes that as a variable, which is part of the problem with how this thread has progressed so far.

[godefroi]

Are you (all) talking atmo or vac dV, cause i only look at vac

It's a kind of nonsense measurement, because it has as much to do with the engines used (and the engines still on the ship) as it does with the ascent profile.

While it's meaningful as a sort of "orbit golf," it's not generally useful for comparing ascent profiles across different ships (and different payloads).

[SanderB]

I've managed 3,157m/s vacuum dV to LKO with 1.0.2 pitching to 40° and maintaing vertical velocity at 350m/s until apoapsis is 75km. I think the solution is mostly high TWR, high mass, long elongated rockets and shallow launch profiles.

Edited May 13, 2015 by SanderB

[cybersol]

How about this for a "Standard Rocket"?

Part list, top down:

Aerodynamic nose cone

Mk1-2 command pod

Service Bay (2.5m) containing 2 PB-NUK

X200-16 tank (payload)

Rockomax decoupler

X200-16 tank

Poodle

Rockomax decoupler

X200-32 tank with 4 Advanced Canards at the bottom

Skipper

Rockomax decoupler

Jumbo-64 tank with 4 Advanced Canards at the bottom

Mainsail

MJ says it has 4014 vac dV with useful TWRs at all the stages.

Dead easy to get to a 100km orbit and has a few hundred dV left.

I think the TWR of the middle stage of this is a little low, but this is still workable as a starting point for a standard rocket. So I built it and set all the gimbals to 30%. I then installed MechJeb 2.5.0.0-455 and did some tests in the background while doing other things. Orbital altitude was 75 km, with 0* inclination. Only autostage (0.1 pre, 1 post) and auto-warp were on.

Here are the results:

[table=width: 1000]

[tr]

[td]Start Altitude[/td]

[td]Start Velocity[/td]

[td]End Altitude[/td]

[td]Final Angle[/td]

[td]Shape[/td]

[td]dV remaining[/td]

[/tr]

[tr]

[td]1[/td]

[td]100[/td]

[td]45[/td]

[td]0[/td]

[td]70[/td]

[td]447[/td]

[/tr]

[tr]

[td]-[/td]

[td]100[/td]

[td]35[/td]

[td]4[/td]

[td]65[/td]

[td]444[/td]

[/tr]

[tr]

[td]-[/td]

[td]100[/td]

[td]40[/td]

[td]1[/td]

[td]50[/td]

[td]457[/td]

[/tr]

[tr]

[td]-[/td]

[td]100[/td]

[td]38[/td]

[td]3[/td]

[td]55[/td]

[td]494[/td]

[/tr]

[tr]

[td]-[/td]

[td]100[/td]

[td]35[/td]

[td]3[/td]

[td]55[/td]

[td]422[/td]

[/tr]

[tr]

[td]-[/td]

[td]100[/td]

[td]38[/td]

[td]3[/td]

[td]50[/td]

[td]454[/td]

[/tr]

[tr]

[td]-[/td]

[td]100[/td]

[td]38[/td]

[td]3[/td]

[td]60[/td]

[td]486[/td]

[/tr]

[tr]

[td]-[/td]

[td]100[/td]

[td]38[/td]

[td]3[/td]

[td]57.5[/td]

[td]492[/td]

[/tr]

[tr]

[td]-[/td]

[td]100[/td]

[td]38[/td]

[td]3[/td]

[td]52.5[/td]

[td]484[/td]

[/tr]

[tr]

[td]-[/td]

[td]100[/td]

[td]38[/td]

[td]4[/td]

[td]55[/td]

[td]501[/td]

[/tr]

[tr]

[td]-[/td]

[td]100[/td]

[td]38[/td]

[td]5[/td]

[td]55[/td]

[td]506[/td]

[/tr]

[tr]

[td]-[/td]

[td]100[/td]

[td]38[/td]

[td]6[/td]

[td]55[/td]

[td]509[/td]

[/tr]

[tr]

[td]-[/td]

[td]50[/td]

[td]38[/td]

[td]6[/td]

[td]55[/td]

[td]518[/td]

[/tr]

[/table]

The - for altitude denotes that the 100 m/s kicked in first and overrode that value. I'd also like to give shout out to "tavert" from back in the day that posted a similar table that helped me in the long past.

Cheers,

CyberSoul

Edited May 13, 2015 by cybersol
One last useful datapoint

[Jikahn]

There is no optimal ascent curve and if you pick one you are going to have a bad time. Depending on your upper stages with mass, drag, geometry you may find that a steeper ascent is advantageous. This is also dependent on what engines you are using and where.

The only hard rule I follow is that I wait for my vertical velocity to be >= 50m/s before kicking my gravity turn into gear. Then it's a matter of being more or less aggressive when following prograde to keep your ascent rate ideal.

You are the winner.

[WeirdCulture]

I've managed 3,157m/s vacuum dV to LKO with 1.0.2 pitching to 40° and maintaing vertical velocity at 350m/s until apoapsis is 75km. I think the solution is mostly high TWR, high mass, long elongated rockets and shallow launch profiles.

I am afraid my spaceships would burn up with this ascend

[cybersol]

I've managed 3,157m/s vacuum dV to LKO with 1.0.2 pitching to 40° and maintaing vertical velocity at 350m/s until apoapsis is 75km. I think the solution is mostly high TWR, high mass, long elongated rockets and shallow launch profiles.

This is similar to my best manual piloting in 1.0.2 of 3200 dV with a starting TWR ratio of 2 and a low and hot ascent profile.

I am afraid my spaceships would burn up with this ascend

It's actually much harder to burn yourself up in 1.0.2. Try 1.0, where it was very possible!

[sgt_flyer]

For my chaos star series (especially true with the chaos star 500 on the limits of overheating)

I usually do a pitchover of 5 - 6° off vertical as soon as the rocket goes over 35 or 40m/s of speed - then i more or less follow the prograde marker until i fire the upper stage (at this point, i more or less fastly go to the horizon depending on my veryical velocity) - ends up generally getting to orbit with 150 m/s left of dV in the upper stage - with a total vacuum delta-V of around 3550 m/s (with high thrust lower stages, and high efficiency upper stage)

[WeirdCulture]

It's actually much harder to burn yourself up in 1.0.2. Try 1.0, where it was very possible!

Yes I use a tweak config so I have a DRE feeling like it was pre 1.0.2. in 1.0.2 the heat is more a blush then a hot glowing heat

[NikkyD]

I suggest the following rocket for "very small" testing category:

OKTO2 control, the very small aerodynamic nosecone on top, 3 Oscar B fueltanks and one 48-7S Spark engine.

If you use mechjeb, put it on top of the okto before you place the nosecone, if you attach it to the side it creates drag.

This rocket has a very good budget and doesnt allow crazy maneuvers.

[godefroi]

I suggest the following rocket for "very small" testing category:

OKTO2 control, the very small aerodynamic nosecone on top, 3 Oscar B fueltanks and one 48-7S Spark engine.

If you use mechjeb, put it on top of the okto before you place the nosecone, if you attach it to the side it creates drag.

This rocket has a very good budget and doesnt allow crazy maneuvers.

Your rocket proved unstable in my testing. Also, I don't believe that putting MJ under the nosecone affects anything; only stack-attached parts are shielded from the airstream, radial-mounted parts have normal drag wherever they are (unless it's in a fairing or cargo bay).

Here are my findings (MJ 2.5.0.0-455, with corrective steering, auto-warp enabled; orbit altitude 75km, inclination 0*).

My test rocket consists of:

Aerodynamic Nose Cone

MK1 Lander Can (AR202 MechJeb stuck on the side)

FL-T800

FL-T800 (4x AV-R8 winglets even with the bottom edge)

LV-T45 Swivel

This rocket has, according to KER, 1.76 TWR on the pad, and 3,678 dV. VOID and MJ concur (but I think they're both using the KER code now...).

Note that my end dV numbers are all negative. This means that MJ never achieved orbit with this rocket.

[table]

[tr][td]start alt[/td][td]start vel[/td][td]end alt[/td][td]final angle[/td][td]turn shape[/td][td]end dV[/td][/tr]

[tr][td]5[/td][td]75[/td][td]45[/td][td]0[/td][td]70[/td][td]-244.2[/td][/tr]

[tr][td]5[/td][td]75[/td][td]45[/td][td]0[/td][td]60[/td][td]-222.5[/td][/tr]

[tr][td]5[/td][td]75[/td][td]45[/td][td]0[/td][td]50[/td][td]-217.3[/td][/tr]

[tr][td]5[/td][td]75[/td][td]40[/td][td]0[/td][td]50[/td][td]-lots (and lots of shock heating)[/td][/tr]

[tr][td]5[/td][td]75[/td][td]50[/td][td]0[/td][td]50[/td][td]-187.5[/td][/tr]

[tr][td]5[/td][td]75[/td][td]55[/td][td]0[/td][td]50[/td][td]-177.9[/td][/tr]

[tr][td]5[/td][td]75[/td][td]55[/td][td]0[/td][td]40[/td][td]-224.6 (and some shock heating)[/td][/tr]

[tr][td]5[/td][td]75[/td][td]55[/td][td]0[/td][td]60[/td][td]-206.9[/td][/tr]

[tr][td]5[/td][td]75[/td][td]60[/td][td]0[/td][td]50[/td][td]-178.5[/td][/tr]

[tr][td]5[/td][td]75[/td][td]60[/td][td]0[/td][td]60[/td][td]-218.5[/td][/tr]

[/table]

Note that this is all just golf. There is no globally-optimum ascent profile; otherwise, MJ would just use it and offer no option to edit it. It has a lot to do with TWR, staging, and engines.

Edited May 15, 2015 by godefroi

[Delrey]

I've also been aiming to achieve 350 m/s velocity at about 30 - 40 degree then not exceeding that by leveling out that until I get to 38 km.

I've been building apoapsis at about 56 km with horizontal velocity which has been feeling really good.

[NikkyD]

@godefroi, if you drop the stupid winglets you can get this thing into orbit, at least i can.

I dont know how to express everything here in text so i just note some pointers:

- with realistic drag your rocket should always be prograde, never have the chevron at the border of the yellow circle but always inside centered if possible.

- turn on debug drag info in action menus, right click your nosecone, the drag vector should always be 0 / -1 / 0, if it turns to -0.1/-0.9/0 you are in danger of tipping over

- once you lean your rocket over to the side (start a turn) your rocket will fall by itself depending on its weight, speed and TWR.

-- too fast and you wont fall much, too slow and you fall over (flipping)

- drag increases like crazy when faster than mach 0.8, keep an eye out for the drag number, if it goes up like crazy, you are too fast

- winglets CAUSE DRAG! Using winglets to enforce a trajectory means you do not understand how your rocket should ascend !

Unfortunately you can no longer rely on MJ to do a good ascend for you with some rockets

[meyerweb]

Are there significant differences in drag etc. between 1.0.1 and 1.0.2? I’m still on 1.0.1 and managed to juuuuuuust barely reach LKO and return to Kerbin with a rocket KER said had 2,729m/s vacuum dV. I got my periapsis into LKO with maybe 50m/s dV to spare. (Should’ve taken pictures!) Similarly, I managed to get Valentina into an elliptical high/low orbit with a 2,934m/s vaccum dV rocket.

Edited December 12, 2015 by meyerweb
Cleaned up forum migration errors (Unicode and hyperlinking were both botched—sigh)

[NikkyD]

My personal rule of thumb for average-jeb is: aim for 4000 dV vac and you will get into orbit. Some rockets can achieve low 3k values but you really need a special rocket for it and if you build a different one, everything will be different

[EtherDragon]

I'm working on a new video that compares three ascent profiles. In filming it, just gathering data, I could already see how huge a difference a good ascent profile makes on available fuel after reaching circular orbit.

[godefroi]

@godefroi, if you drop the stupid winglets you can get this thing into orbit, at least i can.

Right, I know; the point wasn't to get to orbit, the point was that different rockets would have significantly different "optimum" ascent profiles. There's no one-size-fits-all gravity turn.

[NikkyD]

not anymore... but that makes it more interesting