[old thread] Trajectories : atmospheric predictions

[arkie87]

Staging is not implemented, so when you separate the command pod, it's completely normal that the prediction changes a lot. The prediction for the whole ship probably has similar drag forces, but as the ship is heavier, it will go farther. The command pod alone will fall short of the initial prediction. Ideally, the mod should be improved to allow select the stage that will enter atmospher, but I haven't tried yet, and I don't know if it's possible with FAR.

Is is possible to change the craft that FAR/NEAR/stock lift/drag functions use i.e. do these functions always use the current craft or is the craft an input into the functions. If the former, than i doubt this can be implemented unless NEAR/FAR/Stock lift/drag functions are improved such that the input craft needs to be specified...

The only possible workaround for now is to put some monopropollent and RCS thrusters to get some deltaV on your command pod...

Alternatively, plan to enter atmosphere with the extra weight of your tank and motor and ditch it at the last second before deploying your parachute...

Edited November 19, 2014 by arkie87

[Beetlecat]

Were my eyes deceiving me, or did someone show a version/implementation of this with the target and path showing in flight mode? Now *that* would be cool. [not that this isn't... this is amazing! ]

[Semmel]

Staging is not implemented, so when you separate the command pod, it's completely normal that the prediction changes a lot. The prediction for the whole ship probably has similar drag forces, but as the ship is heavier, it will go farther. The command pod alone will fall short of the initial prediction. Ideally, the mod should be improved to allow select the stage that will enter atmospher, but I haven't tried yet, and I don't know if it's possible with FAR.

Well thx anyway. It certainly is not easy to do that since orientation, shape and weight have a strong impact. Would be wonderful though if it would work!

The only possible workaround for now is to put some monopropollent and RCS thrusters to get some deltaV on your command pod...

I was thinking of that already. That might also give some advantage by using an offset between the geometry center of the descend module and the center of gravity to provide lift during reentry.

Alternatively, plan to enter atmosphere with the extra weight of your tank and motor and ditch it at the last second before deploying your parachute...

I am using deadly reentry. That is not an option. Anyway, the largest effect of the landing location is applied after the hot part of the reentry during the thicker part o the atmosphere. The service module would be burned up at that point anyway.

[Youen]

What do you mean you are more concerned about correctness of aerodynamic force computations-- you think you are using these forces incorrectly (unsure if you implemented it right) or that accuracy of prediction is much better than ability of user/pilot to maintain required AoA anyway (so why bother improving accuracy/speed of prediction)?

I mean that I'm not sure my code is correct (maybe there are still bugs), and also maybe it's inacurate (because of the caching system for example).

There are other sources of drift between prediction and actual trajectory, as you mentionned, which make it hard to know where the problem is exactly when you don't land where it was predicted.

[Youen]

Is is possible to change the craft that FAR/NEAR/stock lift/drag functions use i.e. do these functions always use the current craft or is the craft an input into the functions. If the former, than i doubt this can be implemented unless NEAR/FAR/Stock lift/drag functions are improved such that the input craft needs to be specified...

At this time, with FAR and NEAR, I call functions on each part (actually on the FAR modules installed on those parts) to get the force depending of the velocity and air density. I can skip some parts (because they will be detached by staging), but then I'm not sure I'll get correct results because FAR uses nearby parts for its computations as far as I know. So it may be needed to dive more deeply in FAR code to find out how to simulate a different vessel. Also, maybe it's possible to instantiate a new vessel (that will be a copy of what we'll get after staging) and use it for computations. But I'm not sure if this can work without putting this "dummy" vessel in the physics simulation etc.

[Youen]

Were my eyes deceiving me, or did someone show a version/implementation of this with the target and path showing in flight mode? Now *that* would be cool. [not that this isn't... this is amazing! ]

If it did, that was a bug. Also, implementation changed on how trajectories are rendered in map view, so I doubt they can be displayed in flight view anymore.

The feature could be implemented, but that'd require a bit more work, and that's not something I'd want personnaly (drawing virtual stuff in flight view). If someone wants to do it though, it's easy to merge code on github and contributions are welcome.

[GloriousWater]

So I installed this mod yesterday, and it's really handy, but one thing is bugging me.

With Active Texture Management installed, this mod's toolbar button becomes a black square, which doesn't affect gameplay but just looks ugly. Does anyone have a fix for this?

[Beetlecat]

If it did, that was a bug. Also, implementation changed on how trajectories are rendered in map view, so I doubt they can be displayed in flight view anymore.

The feature could be implemented, but that'd require a bit more work, and that's not something I'd want personnaly (drawing virtual stuff in flight view). If someone wants to do it though, it's easy to merge code on github and contributions are welcome.

Ah cool -- It wasn't here, but I think someone was demoing what it would look like while in flight. It may have been Bahamuto.

[hab136]

I am using deadly reentry. That is not an option.

I re-enter all the time with my engine and tank attached using Deadly Re-Entry (on the default settings, with FAR). If you've increased the deadliness, then that's one thing.. but engines make great heat shields normally, if you're just coming in from a low, stable orbit around Kerbin. You have to stay on surface retrograde marker like your life depends on it though - because it does

[Starwaster]

I am using deadly reentry. That is not an option. Anyway, the largest effect of the landing location is applied after the hot part of the reentry during the thicker part o the atmosphere. The service module would be burned up at that point anyway.

Are you using FAR?

You could try flying the error out. (assuming you're using a lifting body.... if not, disregard)

[INSULINt]

I can't remember if I've thanked you for this mod, so thank you for this mod It's made my latest tour of Jool and its moons way less frustrating when aero braking!

[INSULINt]

I noticed there is a cfg with the following in the folder:

@PART[*]:HAS[@MODULE[ModuleCommand]]:FOR[Trajectories]
{
	MODULE
	{
		name = TrajectoriesVesselSettings
	}
}

Just wondering what this does, because I modified it to only add that module to specific parts yet I can still access the interface without those parts. Just toying with the idea of making some probe cores "flight computers" and such.

[Youen]

This module is just used to save (ie persist when changing vessel and saving/loading) per-vessel settings (descent profile, target, etc.), so most functionalities work without it.

Edited November 30, 2014 by Youen

[Superfluous J]

I just hit Kerbin's Badlands from Low Munar Orbit - with a maneuver node - thanks to this mod.

One word: sweeeeeeeeeeet.

[INSULINt]

This module is just used to save per-vessel settings (descent profile, target, etc.), so most functionalities work without it.

Cool You ever consider doing a "part mode" option like engineer? Just out of curiosity

[Frederf]

I had a thought. What if the length of the lines on the "impact plus" symbol meant something. Propagate a plausible error in each axis through a trajectory and find the ellipse that encompasses these error impact locations and draw the plus legs that match the dimensions of that shape. So a long, flat reentry that is sensitive to initial conditions will have a big footprint (large plus) and the plus narrows as the solution gets more and more certain.

[Youen]

This feature has already been proposed. The difficulty is more about estimating the accuracy than displaying it.

[DivisionByZero]

This mod is fantastic - I just wanted to say that again. I had a tricky mission to Mun in a career mode game where I had to return 850 units of Karbonite to orbit around Kerbin. I'm limited by what I can launch, etc. from Kerbin by FAR+DREC, career science unlocks, KW rocketry fairings, KCT timing and the career is in hard mode so I have to save money. The Karbonite extractor and power source (KSPI fission reactor) was already split from the return mission because of all the extra weight.

So, it comes time to return to Kerbin and get the thing in some kind of orbit, except, I'm realizing after launch that it takes quite a bit more delta-v than I remembered to get off the ground. All of a sudden I'm thinking of economizing and remember I have this fantastic mod that can calculate aerocapture maneuvers! The craft wasn't designed with this in mind, so I was pretty conservative and only dipped down to 34km. One thing I would recommend for the mod is to add an "estimated temperature" calculation in addition to the g-force calculation you already have. Anyway, the aerocaptures were calculated spot on in the game and I was able to pull my apoapsis down from Munar orbit to 180km with 0 delta-v on three passes through the atmosphere and 0.4m/s adjustments in between! It's something like 500+ delta-v savings in the end.

Landings on the Mun were also very helpfully performed with trajectories. Getting close to the extraction base was where I burned all my fuel, but that's not the mod's fault.

Anyway, I just wanted to suggest the "estimated temperature" and give a big thumbs up for the mod again!

[Mystique]

Great mod, but I don't really understand how can I make landing predictions =\ For example, I want to land on KSP with some vessel. I set up landing profile to 180 degrees in all cases and make maneuver to get landing X sign on KSP. But when I'm actually starting to enter atmosphere and aerobrake, trajectory changes and X sign moves closer to my current position, i.e. it seems that I'm braking faster than mod calculated. Could it be because of some modded parts of my vessel, for example, that create extra drag or something?

Update: when I descend to something like 40 km on Kerbin, trajectory prediction stabilizes, but I don't understand why it's changing so much before that (like I was planning to go almost around planet within atmosphere and instead make only a third of that atmosphere trip.

Edited December 3, 2014 by Mystique

[hab136]

I set up landing profile to 180 degrees in all cases

Are you keeping the navball exactly on the surface retrograde marker during the whole descent? (and not orbit retrograde)

Update: when I descend to something like 40 km on Kerbin, trajectory prediction stabilizes, but I don't understand why it's changing so much before that (like I was planning to go almost around planet within atmosphere and instead make only a third of that atmosphere trip.

If you have a really long atmospheric ride, you'll go down, then back up, then down again, and Trajectories isn't too great at predicting that.

What I do is initially aim to overshoot and land just East of KSC, in the ocean. Then once I'm a little lower (like around 40k, just before the re-entry effects start), fire the engines a bit to make the landing more precise.

Oh and we really need a "180" button, or better yet, a configuration option to always default to 180. Having to open up the config window and drag 4 sliders every single time I launch a new ship and then return the reusable booster is tedious, especially since I've been building 3 different space stations and every piece has a booster that needs to be recovered. I never launch space planes (AJE makes them quite difficult) so I always want to come in retrograde (180 degrees).

[LostOblivion]

Is the white line there still until the trajectory reaches the atmosphere, or was it changed to blue?

[Youen]

You have the choice in the mod window to draw the whole trajectory, or only the trajectory that diverge from the stock one (i.e. starting from the first atmospheric entry)

Excepted in body-fixed mode, in this case the whole trajectory is always drawn.

[LostOblivion]

Sweet mod!

[rdwulfe]

Anyone know if this runs with .90 yet? I'll try it out myself, too.

EDIT: It installs with no errors, but has no functionality. The menu shows up, there's no G-force readout, and it does not display anything. Many many errors are reported in the settings dialogues, and the error:

[Exception]: NullReferenceException: Object reference not set to an instance of an object

shows in the debug ALT-12 menu. Whee!

Edited December 16, 2014 by rdwulfe

[theersink]

Any plans to update this to 0.90? Sadly it breaks with the update. One of my favorite mods.