[1.0.5] GravityTurn version 1.3.1 - Automated Efficient Launches (1.1 pre-release available)

[legoclone09]

Looks awesome! Maybe for RO you can run a MM patch that makes it run RCS thrusters before second stage ignition? That likely would work. 

[vardicd]

3 hours ago, linuxgurugamer said:

If he is using StageRecovery, the parachutes don't deploy at all, but SR will use them to calculate how fast the stages would be dropping IF they were deployed safely when the stages finally hit the ground or ocean

Please provide a list of the mods you have installed.

Oh dear lord, you want all my mods? Okay: Adjustable landing gear, alternate resource panel, ampyear power management, antenna Range, Asteroid Resource, Asteroid Day, BDynamics Mk22 Cockpits, Better burntime, Chatterer, community tech tree, Contract pack GAP, Contract configurator, contract pack anomaly surveyor, contract pack field research, contract pack tourism, Dmagic Orbital Science, Extraplanetary launchpads, final frontier, fuel wings, gravity turn, hide empty tech nodes on community tech tree, Karibou rover, Kerbal Attachment system, Kerbal aircraft expansion, kerbal alarm clock, kerbal contruction time, kerbla flight indicators, contract pack kerbin space station, Kerbal Inventory System, KSI Placement services, Mark 4 spaceplane system, mechjeb, Mk2 stock alike expansion, near future solar, Near future space craft parts, persistant rotation, portrait stats, Radiator toggle, Realchute parachute systems, stage recovery, stock_jet turbines resurrected, stockalike station parts, stock bug fixes, SXT trucks, Take command, Toolbar {blizzys}, OKS/MKS colonization system, Universal Storage, USI submarine, USI lifesupport, VNG parachute/ plugin.

[vardicd]

@Overengineer1 @5thHorseman @linuxgurugamer 

Here is a link to the video I recorded of me doing my own launch profile, using mechjeb for control, then a second launch with the gravity turn mod, and a short clip at the end showing the appr fuel values in the tanks when it loses control, and the position of the COM and COL. Please, offer any advice, or critiques. I freely admit I'm not so good at building these, I'm still way too used to the old aero system, I didn't upgrade to 1.0 and stayed at .90 for a long time, because of the aero changes. It was one of the reasons I never used FAR, I tried it for 2 months and never managed to get a rocket into orbit.

Edited January 15, 2016 by vardicd

[Overengineer1]

3 minutes ago, vardicd said:

 didn't upgrade to 1.0 and stayed at .90

Are you saying we are looking at version 0.9 in this video?  Well, there's your problem.  I'll add code in the next release to disable GravityTurn for an unsupported game version, since the aero parameters in older versions invalidate the calculations that GT does.  Sorry, either upgrade your game, or keep playing the way you were before.

[vardicd]

1 hour ago, Overengineer1 said:

Are you saying we are looking at version 0.9 in this video?  Well, there's your problem.  I'll add code in the next release to disable GravityTurn for an unsupported game version, since the aero parameters in older versions invalidate the calculations that GT does.  Sorry, either upgrade your game, or keep playing the way you were before.

No, this is the latest version of KSP, I was saying I didn't update for a long time to 1.0, because I didn't want to deal with the aero changes. I would not be so silly as to ask for help with an older version of the game.

Edit: My comment was meant as an explanation of why my skills at building rockets with the aero system sucked so bad. I haven't been using the new version of KSP for very long. Sorry about the confusion @Overengineer1

Edited January 15, 2016 by vardicd

[MK3424]

I didn't saw this one coming!:

https://www.youtube.com/watch?v=WRuUEgKpkzQ

Edited January 15, 2016 by MK3424

[NathanKell]

@mcirish3 I know it's a long thread, but I'm fairly sure that Claw and OhioBob and I discuss that very thing in it somewhere, and OhioBob even posts some helpful graphs.

[Overengineer1]

@vardicd Well then you have 2 problems:  Lots of drag up front, and a lack of positive control.  If you use the R8 fins, I think it would work.  Or a Swivel engine would probably be enough.  And losing those radiator panels.  Any one of these would probably help, all 3 wouldn't hurt.

The version currently in development is slightly more intelligent about how it handles pitch, which may or may not help in a situation like yours.  But it's not ready yet.

[gamerscircle]

I am a lousy pilot/astronaut, okay.. I said it!  This is why I like mods that help me enjoy the game and prolong my addiction, err.. umm passion?  Anyway, I read though all the posts on this and I do have MJ installed, but I also have the mod called, "mechjeb and engineer for all'  , which I think 'integrated' the parts , so you don't  have to attach the parts.

Does GT see that or do I have to attach the MJ part for the integration?

Again, if this was covered, sorry that I missed it.  If both MJ and GT are installed, just set GT to what you want and they [MJ and GT] will work like a well oil machine?

[Overengineer1]

GT requires a MechJebCore module on the ship to do the circularization burn.  It looks like MechJeb For All just adds a MechJebCore module into all command modules.  In that case, GT should see that MechJebCore and instruct it to do the circularization burn.

[gamerscircle]

7 minutes ago, Overengineer1 said:

GT requires a MechJebCore module on the ship to do the circularization burn.  It looks like MechJeb For All just adds a MechJebCore module into all command modules.  In that case, GT should see that MechJebCore and instruct it to do the circularization burn.

Thank you kindly for the quick reply and support.  

[tempsgk]

@vardicd

Craft file?

[Overengineer1]

1.2.1 released for all you inclined tweakers.

[mcirish3]

5 hours ago, NathanKell said:

@mcirish3 I know it's a long thread, but I'm fairly sure that Claw and OhioBob and I discuss that very thing in it somewhere, and OhioBob even posts some helpful graphs.

I thought I read nearly every post at least cursory for main topic idea.   I see lots of really great and detailed info about aerodynamics and thermodynamics.  But what I want to understand before I tackle those topics is the nature of the KSP Aerostatics and Thermostatics.  In other words the nature of the planetary atmosphere when a rocket or space plane is not punching a hole through it.  Of this type of information if did not find any real detailed information.  Certainly none that would be of much use computationally. 

 

2 hours ago, Overengineer1 said:

1.2.1 released for all you inclined tweakers.

I have yet to try your mod, I fully intend to in the near future.( I rather think it will end up being an essential mod)  woot on the new update.

[Superfluous J]

9 hours ago, vardicd said:

Here is a link to the video I recorded of me doing my own launch profile, using mechjeb for control, then a second launch with the gravity turn mod, and a short clip at the end showing the appr fuel values in the tanks when it loses control, and the position of the COM and COL. Please, offer any advice, or critiques. I freely admit I'm not so good at building these, I'm still way too used to the old aero system, I didn't upgrade to 1.0 and stayed at .90 for a long time, because of the aero changes. It was one of the reasons I never used FAR, I tried it for 2 months and never managed to get a rocket into orbit.

Your first launch was WAY too vertical to start, and then you turned too fast at 40km, also, throttling down during launch is never a good idea. If you MUST throttle down, then your ship has too much TWR (note: your ship has too much TWR) and you should remove engine power to save the weight. Why carry what you're not using? During my test launch, my (stockified so may be slightly different) rocket made orbit with fuel still in the center launch stage.

Your 2nd launch in comparison is much better, but the ship does do odd things. I wouldn't call that a "flip" though, more of a veer.

I must apologize, the COM/COT thing is not related to this problem. I agreed at the time with whomever said that, and after I actually built my copy of your rocket I realized that it doesn't matter. That has more to do with SAS getting confused about which way to auto-gimbal the rockets and has nothing to do with aerodynmic stability. Sadly, there is no "center of aerodynamic stress" indicator because that depends on more than just the rocket size and shape, but your COM marker is obviously very low in the rocket, which is what's causing the problem. Really, you should redesign the rocket to raise that COM, and you should start by ditching at least one of those three engines. I know that 1 engine isn't enough and 2 is troublesome to work with but still that's what I'd suggest. 1 engine and 1.1 launch TWR is more controllable AND cheaper (though it'll use more fuel and take more m/s of dV, but see my signature for what I think about that) than 3 engines getting 2+ launch TWR.

One thing I've done in early career is to use a single BACC Kickback SRB on the bottom of my craft for the initial kick off the surface, and not actually stage the LFO engine until the BACC Kickback burns out. It's a cheap way to get some early height and speed, without removing mass from the top of the rocket where you want it during the early rise.

One thing I'd like to know, how does MechJeb handle similar parameters, or how does this mod handle the parameters you plug into MechJeb? I don't know anything about MechJeb (and very little about this mod TBH) but these tests don't seem useful as they're using very different parameters.

Edited January 16, 2016 by 5thHorseman
Changed "BACC" to "Kickback" - sorry about that!

[Overengineer1]

1 minute ago, 5thHorseman said:

throttling down during launch is never a good idea

This needs some clarification, because it gets spread around a lot and it's often applied incorrectly.

The most efficient ascent profile for an average KSP ship will not be riding full throttle the entire time.  GravityTurn will show you this, the primary function of the mod is to reduce throttle to avoid the extremely steep ascents that are characteristic of most KSP players' wildly overpowered ships.

A more efficient ship design will have a lower Thrust to Weight Ratio, and so will be more likely to need full throttle through more of its ascent.  So a more efficient overall mission will use full throttle for its ascent.  The mission is not more efficient because you used full throttle, but rather because the ship itself was more efficient.

Telling a player that throttling down during launch is a bad idea is a mischaracterization of the situation.  And expecting a player to fine tune his mass and staging for maximum efficiency, when he is having trouble keeping his ship from flipping during ascent, that just doesn't seem like a reasonable expectation.

I built GravityTurn because I can't design everyone's ships for them.  But what I can do is show them how much thrust they need by throttling down their ship.

[Superfluous J]

45 minutes ago, Overengineer1 said:

This needs some clarification, because it gets spread around a lot and it's often applied incorrectly.

I did clarify it.

1 hour ago, 5thHorseman said:

If you MUST throttle down, then your ship has too much TWR ... and you should remove engine power to save the weight. Why carry what you're not using?

I remade the ship in 2 ways, one with a Kickback SB and one with 3 BACCs in 2 stages, first 2 of them and then 1. Both of the ships only needed 2 LFO tanks in the 2nd stage and I still dropped fuel back to Kerbin. They needed more control authority as the SRBs don't gimbal, of course, but they were still cheaper than the original, cheap enough for me to not feel the need to make them recoverable.

But anyway, this is ancillary to the mod itself which IS very cool and useful, both as a teaching aid and as a "go make a sandwich during the boring stuff" aid. I'm going to bow out. You DO make very good arguments that I don't (completely) disagree with.

Edited January 16, 2016 by 5thHorseman

[rriegner]

Could you submit this to the CKAN for easy updating? If you don't know how or don't have time, I could do it for you.

[Overengineer1]

@rriegner I put in a pull request 6 days ago.  It's still pending.  

[rriegner]

11 minutes ago, Overengineer1 said:

@rriegner I put in a pull request 6 days ago.  It's still pending.  

Well, one step (or six) ahead of me. Thanks

[NathanKell]

@mcirish3 I'd suggest PMing OhioBob then. He's very friendly. If there's anything he can't answer, loop me in to the convo.

[BlackHat]

KottabosGames has does a great Mod Demo of this mod. I would add the link to the OP.

I am interested in this mod, but will wait until its added to CKAN.

 

 

[danfarnsy]

Overengineer,

This is fantastic. I use FAR, and I've been testing this out with a number of variations in the settings. So far, the "best guess" for a simple 1.25m stack orbiter seems to be pretty good (~96 m/s, 10.7 degrees turn), and increasing the AP hold times from 40 seconds to 45-50 doesn't seem to hurt the bottom line. Changing AP hold times lower (~35 seconds), though, leaves my ascent a little too shallow. Overall delta-V to 80 km for most approaches is about 3350 m/s, which is better than I was getting otherwise, though I haven't specifically measured how much (usually in the ballpark of 3500-3700 m/s). If nothing else, you've reminded me of what a proper ascent profile looks like. Of course, time is a resource too, so I may end up going for steeper ascents just to save time.

If I get better ascents with FAR, I'll be sure to share. Thanks for the mod!

[damerell]

5 hours ago, danfarnsy said:

This is fantastic.

Quite. Second-best entry in the ascent profile challenge, and without a great deal of fiddling.

[mcirish3]

Just curious what happens if you have a slightly underpowered rocket?  I know... I know... that almost never happens.