[0.24] If a Booster Has Enough Parachutes On It, Count it as "Recovered"

[Monger]

So why not use chutes to control the landing?

This is not my point. My point is: you should be in control while landing in order to recover costs. There should be no magic formula that guesses whether you could land safely, and do the recovery for you.

[Technical Ben]

Why does the player need to be in control? None of Space X or Nasa/other countries use manned flight for landing... so why can the game not have automatic landing with some restrictions?

Such as the restriction for aircraft autolanding requiring a Kerbal pilot? For rocket, a probe/parachute?

Granted, not all designs can land. But we can design proven landers. So some way to confirm they can land, or some way to go back and land them by hand later?

[rkman]

I'm not a big fan of that. Dropping stages on parachutes is something that basically never happens in real life, yet alone: is applied to every single design. And what you are proposing will do just that: make people attach chutes to every single stage they're dropping.

besides - do an in-game test of dropping some larger booster on a chutes - it explodes most of the time on a contact with ground.

I have done successful landings of booster stages by using landing legs (and parachutes).

But it does feel like it is to easy. More realistic would be a SpaceX Falcon 9 style recovery: retrograde burn and powered landing. There probably is a good reason why they don't use parachutes.

[Kegereneku]

About spaceX the video we know is more advertising than truth.

They would necessarily use parachute for most of the descent, otherwise the deltaV requirement would be absurd. This is not KSP.

As someone said the video of "SpaceX full recovery" is indeed a fantasy in the sense that it won't be so dramatically easy since SpaceX is unlikely to make engines as incredible as they make it look.

For example I have yet to see SpaceX intend to recover the first stage without a second landing pad to allow a cheaper parabolic trajectory. The second stage however seem intended to reach orbit so it could in theory land anywhere.

Lastly the capsule is never going to do an all powered landing, a paraglide for 99% of the descent is a more reasonable and reliable solution.

Why does the player need to be in control? None of Space X or Nasa/other countries use manned flight for landing... so why can the game not have automatic landing with some restrictions?

I think he mean it in the sense that the booster's descent, landing and recovery is done in a controlled fashion (manned or not), rather than a abstract script pretending you did all this and counting the stage as recovered.

That's my point of view in any case.

[R0cketC0der]

About spaceX the video we know is more advertising than truth.

They would necessarily use parachute for most of the descent, otherwise the deltaV requirement would be absurd. This is not KSP.

The Falcon 9 first stage already has the required deltaV for a full landing. It's not as much as you think. My estimate would be 2.5km/s for the turnaround burn and another 300m/s for the landing. Considering that the first stage is really light without a second stage+payload mounted on top of it this is totally doable.

As someone said the video of "SpaceX full recovery" is indeed a fantasy in the sense that it won't be so dramatically easy since SpaceX is unlikely to make engines as incredible as they make it look.

All of the engines in that video have already been fully developed and certified for flight. And with the exception of the SuperDraco used for the softlanding of the capsule all of the engines have already flown multiple times.

For example I have yet to see SpaceX intend to recover the first stage without a second landing pad to allow a cheaper parabolic trajectory. The second stage however seem intended to reach orbit so it could in theory land anywhere.

Lastly the capsule is never going to do an all powered landing, a paraglide for 99% of the descent is a more reasonable and reliable solution.

Why wouldn't they be able to do a full powered landing? It allows for much more precise landing and doesn't require as much delta-V as you seems to think. A conical pod has a pretty low terminal velocity and therefore it takes only very little deltaV to land. I wouldn't be surprised if it used more dV for orbital maneuvers than to land.

[Sky_walker]

I have done successful landings of booster stages by using landing legs (and parachutes).

Yea, me too... on a perfectly flat with huge legs, and retrograde burn while descending on parachutes.

Sadly - perfectly flat ground isn't really a common thing in on Kerbin - and otherwise the risk of tipping over is ridiculously high, especially for taller boosters.

But it does feel like it is to easy. More realistic would be a SpaceX Falcon 9 style recovery: retrograde burn and powered landing.

Hehe, even that is far from easy.

ground plays such a huge role, that even on regular grasslands with excellent descend curve - you never can be sure about a proper landing and recovery.

IMHO this whole notion of a free boosters recovery is rather silly and never should make it to the game.

All of the engines in that video have already been fully developed and certified for flight. And with the exception of the SuperDraco used for the softlanding of the capsule all of the engines have already flown multiple times.

There have been exactly 0 successful landing of a boosters after commercial launch.

At a very best - they descended as planned and dropped into water where they wanted them to do so.

We can't even speculate about possible successful recovery rate at this point.

Edited July 25, 2014 by Sky_walker

[Javster]

Isn't there a mod which works out whther your debris will hit gently or not, and pays you accordingly?

[mythic_fci]

All of this is very true. Differences between real life and KSP:

In real life, boosters can't always land in the ocean and be recovered due to the required waterproofing for the salt water. In KSP, that's not an issue and water is in fact recommended for landing large stages.

In KSP, you can stick 16 parachutes on your first stage and land it perfectly. In real life, that's impossible.

In real life, you need heat shields (for stages that go suborbital), extra guidance equipment, power systems, and more to land a stage. In KSP, just strap a probe core, 2 batteries and a few landing legs and you can call it a day.

Here's my opinion: Have something of a "failure mode". This means in programming terms: Out of say 10 variables, one of them is the failure variable, two are partial failure variables and three are slight damage variables. The rest are success variables. The game would randomly pick from one of them. The failure mode would be a complete loss of the stage, without any recovered funds, the partial failure mode would be about a 50% fund recovery alongside the existing distance from KSC variable, and the slight damage variable would be simply about an 85% fund recovery alongside the existing distance from KSC variable. The success variable would be full fund recovery minus the existing distance from KSC variable. This system could also be implemented for a random failure system - parts with a higher failure rate would have less variables in total and/or more failure/partial failure variables. Failures could be displayed via the in-game toolbar.

Whew! I should post this in the "Addon Requests" part of the forum.

[Sky_walker]

Isn't there a mod which works out whther your debris will hit gently or not, and pays you accordingly?

We're talking here about stock game, not mods.

[M3tal_Warrior]

Personally I'd like to see recoverable parts, since I don't cheat debris out of orbit but deorbit it manually (and in case of engines recover it with parachutes) - and then sit there and watch for 6 minutes the 2859th landing of a out-of-fuel/electricity stage that now costs about 30-40k bucks. Of course I have more than enough funds, but I like my environment clean (and realistic, for that purpose). Even without funds I tried to land as much as possible.

As for ocean landing: Thats what Ariane V boosters do when equipped with parachutes, and what has been done with Space Shuttle boosters all the time. That's because you don't launch a rocket over land (or at least avoid civilized areas). It works in reality (although I don't see the necessity for small boosters in KSP, for aforementioned financial reasons).

Physics bubbles seem a proper way to do it (I'd recommend a bubble testing for chutes, not probe cores and such), and shouldn't be too stressful for a halfway modern computer. Hell, I'm working here on ancient (8yo) hardware, launching massive rockets and all on one (!) 2.33 GHz core (Core2Duo E6650) without having problems at all (about 800 parts create some lag, but only during this trigger-happy physics initiation - that's why I recovered my six experimental planes display in front of the astronaut training center). If you're having problems with newer machines, use x64, learn how to maintain/tweak a computer and stop whining!

Biomes: It seems quite legit to base the recovered value on the biome it's landed in - I do most landings in water (saving poor kerbals huts and homes), almost all work out fine. Also a failure possibility would be quite realistic.

I vote for a quite low recover value, so it doesn't make sense to do it for cheap units, but main stages should work out well. One nice aspect would be a per-part scientific value (just small), which would simulate the knowledge gained about in flight damage of different parts.

Edited July 25, 2014 by M3tal_Warrior

[R0cketC0der]

There have been exactly 0 successful landing of a boosters after commercial launch.

At a very best - they descended as planned and dropped into water where they wanted them to do so.

We can't even speculate about possible successful recovery rate at this point.

I'm fully aware of that. There are still problems to be solved and questions to be answered about their booster recovery program, but it's none of the ones Kegereneku mentioned. I don't think SpaceX will be able to reduce launch costs to the claimed 5-6m$, but I do think that there is a fairly high chance that they will be able to land boosters with a decent recovery rate, though it's still too early to say for sure.

[R0cketC0der]

As for ocean landing: Thats what Ariane V boosters do when equipped with parachutes, and what has been done with Space Shuttle boosters all the time. That's because you don't launch a rocket over land (or at least avoid civilized areas). It works in reality (although I don't see the necessity for small boosters in KSP, for aforementioned financial reasons).

Analysis has shown that the recovery of the Space Shuttle boosters didn't save any money.

[Red Iron Crown]

Analysis has shown that the recovery of the Space Shuttle boosters didn't save any money.

In fairness, those were splashdowns in salt water of solid boosters (which are relatively cheap to begin with). A dry land soft landing of an array of expensive liquid engines stands to save much more.

[M3tal_Warrior]

Analysis has shown that the recovery of the Space Shuttle boosters didn't save any money.

Please do reread my post. You might find your objection there, if being able to do context specific reading.

[SolarLiner]

Almost all the STS boosters have been recovered successfully, they're using parachutes even. SpaceX has its Grasshopper proof of concept which isn't even using parachutes.

"That unbalances the game" - Well, if for you, "balanced" means having no struggle whatsoever in building rockets even with the Funds restriction, then this does not even "unbalance" it.

Why don't we make it so that we get 50% back or something, the same way that recovering ships far from KSC decreases the recovery percentage?

Also, not every single stage needs to be recovered. Massive 1st stages that contains a lot of parts, yes, but single/two part side boosters may not need recovery at all.

[Technical Ben]

Well, I'm now making a SpaceX Style recovery craft for my first KSP Career mode play for a launch of a Duna orbital probe. I've made the first stage 4000k+ to hope at getting the second stage into orbit. As doing a real SpaceX would end up with one of the stages being despawned/atmospherically locked/killed by the game engine.

The second stage should boost the probe to a higher orbit, then return. With the probe (3rd stage/payload stage) taking it's self to Duna.

PS, I'm doing this with half the tech tree too.

PPS, nope, unable to copy Scott, as swapping between craft just does not give me enough time to fly both. STTO seems the only way

Edited July 25, 2014 by Technical Ben

[rkman]

perfectly flat ground isn't really a common thing in on Kerbin

IMHO this whole notion of a free boosters recovery is rather silly and never should make it to the game.

I agree there is good reason not to have stage recovery be automagic. Recovered stages should make a controlled landing on a predefined suitable landing spot. So no "drop and forget".

In my tests i dropped two 1st stage asparagus stacks (not particularly huge) from about 9km, straight down onto KSC. Needed no retroburn and the landings were stable and secure.

I suppose a 2nd or 3rd stage could temporarily be left in high suborbital and landed manually once the rest of the ship is in LKO. Although that gets complicated in combination with Deadly Reentry.

[Monger]

Why does the player need to be in control? None of Space X or Nasa/other countries use manned flight for landing... so why can the game not have automatic landing with some restrictions?

Because autopilots are no fun. This is a game about engineering- and piloting skills. Implementing an autopilot for debris basically enables an autopilot for all other ships as well.

Again: why would you want to add a game feature that actually makes the game less challenging and fun?

Here's my opinion: Have something of a "failure mode". This means in programming terms: Out of say 10 variables, one of them is the failure variable, two are partial failure variables and three are slight damage variables. The rest are success variables. The game would randomly pick from one of them.

Although many games are based on luck (i.e. throwing dices), random events are usually not very well received in video games. If you throw a dice and you loose, then it was bad luck. If the computer throws a dice and you loose, then it feels like cheating.

[tntristan12]

What you didn't get paid back for deployed parachutes? Then slapping on a bunch of parachutes for a relatively cheap but heavy part is no longer economical because you'd lose money. Therefore the player would have to weigh the cost of the extra parts with that of the payload, and would also need to take mass into consideration. Problem solved!

[Torquemadus]

Currently, solid rockets are very attractive for use in a disposable first stage. They are cheap "trash bins full of boom" and have very little recovery value once empty, especially when compared to the cost of parachutes. By contrast, liquid fuelled stages are worth a great deal more than the fuel they contain and are well worth trying to recover.

If players want automatic recovery of their empty "trash bins" I'd say that the game already does it pretty well. The Space Shuttle SRBs were recovered after splash down, but as it turned out, the cost of recovering them wiped out any savings gained from re-using them.

Automatic recovery of liquid fuelled stages is a different matter.

I did a lot of experimenting prior to the 0.24 release to see whether recovery of spent stages was a viable approach. I tried attaching parachutes that activated on separation and focusing on the stage as it re-entered as uncontrolled debris. The trouble with this approach is that most engines and fuel tanks have very low crash tolerances compared to other parts. Even if the stage survived landing (on account of adding a huge number of parachutes), it usually fell over and smashed to pieces. Often, the only parts that survived were the parachutes! I did a lot of experimenting and eventually decided that this approach was unreliable and more trouble than it was worth.

Plan B was to use a kerbal pilot or a probe core to make controlled landings. I found that landing by parachute was still unreliable. Adding the ability to control the stage meant that I could choose where I would land with reasonable accuracy, thus avoiding unsuitable terrain such as mountains, but most designs were still prone to falling over after landing. I eventually took to adding wings and landing gear to each stage so I could return it to the runway.

The approach I ultimately adopted was to use partially re-useable designs that throw away loads of empty SRBs as a first stage and re-use all subsequent stages by returning them to the runway or refuelling them.

I suspect that some players want automatic recovery so that they can carry on using asparagus pancakes without worrying about cost. With automatic recovery, an asparagus pancake would become an "I win button". Simply launch an "efficient" pancake, pay for the fuel, recover most of the cost of the launch vehicle. Profit.

Automatic recovery of liquid fuelled rockets is a ticket to playing the game on easy mode. With launch costs effectively eliminated, payloads of virtually unlimited size can be lifted to LKO. Once they're in LKO, they're half way to anywhere.

[Kerbart]

I'm coming late to this party. But it's hard to run out of funds the way things are right now, and it should be made even harder?

[Kegereneku]

Ah yes, the "Asparagus Pancake", it didn't come to me that not making it easy to recover stages also helped keeping the cost per payload kilogram challenging.

This way we could even end-up with a good reason to use orbital transfer ship and cheap refueling launch.

[Guest]

It'd only work for solid boosters that would splash down. Returning a liquid stage is hard, you need to consider the engine, the tanks and all that. Those things are not good at surviving splashdowns, as SpaceX found out. SSRBs jettisoned their nozzles before splashdown to ensure that the main body survives, since it was essentially a tube, it was much less challenging to recover than a liquid stage.

[Monger]

The approach I ultimately adopted was to use partially re-useable designs that throw away loads of empty SRBs as a first stage and re-use all subsequent stages by returning them to the runway or refuelling them.

This is also what I concluded from 0.24 so far.

Recovering the last stage is always the easiest one, and from there on you work down. Every new technology needs you to revisit this stage, to see if it might be easier to implement as recoverable or non-recoverable stage. The first stage has the lowest benefit from being recoverable, and (again) finally SRBs have a useful purpose in KSP .