Why do recoverable parts need a pod/probe?

[Torquemadus]

Long before 0.24 came along, I did experiments with recovery of launch vehicles. At the time, I was using two stage to orbit expendable rockets, consisting of a liquid fuelled core stage and solid fuelled strap-on boosters. The idea was to recover the core stage down-range by adding a large number of parachutes that would activate on separation. I would then park the payload vehicle in orbit and switch to the core stage so I could watch it re-enter the atmosphere.

The first thing I found was that I needed more parachutes that I had expected. Liquid engines and tanks have very low crash tolerances, so I added what seemed like a lot of parachutes, but as it turned out, the parachutes were frequently the only parts that survived the landing! The second problem was that a launch vehicle core stage is usually tall and thin. The first thing it wants to do after making contact with the ground or water is to tip over. Again, this does not sit well with low impact tolerances! After a lot of trial and error, I concluded that this method of recovery was unreliable and more trouble than it was worth. I switched to doing controlled landings instead, which worked much better. I eventually settled on using winged fly-back boosters that return to the runway at the KSC.

There have been a number of threads on the forums asking for a feature to be added to the stock game whereby spent stages are counted as recovered simply because the player stuck an arbitrary number of parachutes on it somewhere. As I discovered, recovering liquid fuelled stages by parachute is not a trivial matter.

At current prices, solid rockets are worth next to nothing when empty of fuel. Here are some "before" and "after" pictures of an SRB, showing how much it costs with and without fuel. As it turns out, most of the cost of the SRB is the fuel. The empty booster is worth very little.

By contrast, here's a Rockomax "Oil Drum" tank shown both full and empty of fuel. As it turns out, most of the cost of liquid fuel tanks is the tank, not the fuel. In addition, the engines required to burn the fuel add additional cost, which I will want to recover if I can.

Therefore, adding a feature that automatically recovers my "trash bins full of boom" solid rockets is an irrelevance, since their recovery value is negligible. Adding a feature that automatically recovers large liquid fuelled rockets is a way of playing the game in easy mode.

[saburai]

Excellent contribution! Also, very cool note on the relative prices of fuel for liquid vs. solid rockets.

I agree that a feature that automatically recovers large liquid fueled rockets would be unfairly easy. But a feature that allowed you to ATTEMPT to recover a large liquid fueled rocket, but requiring careful design and (perhaps) great skill, would be pretty cool.

[saburai]

SpaceX is certainly trying to get their rocket back in one piece, but I have doubts about them reusing the entire structure. The chances of metal fatigue or any other flaw in the tank walls is too great. There are also lots of fuel lines to purge and clean. I'm betting that they will cut the engines off and reuse them separately rather than relaunch the entire stage. At a minimum they will have to disassemble lots of stuff to repeat the standard inspections that all parts go through during assembly. SpaceX has made it to orbit despite single-engine cutouts, as did Apollo 13's launch. So the insurance people will probably insist that the first "used" engines fly alongside new ones to guard against total failure.

I have doubts about how successful they'll be, but make no mistake, they intend to fully reuse the stage:

“At this point, we are highly confident of being able to land successfully on a floating launch pad or back at the launch site and refly the rocket with no required refurbishment,†SpaceX said in its statement. “However, our next couple launches are for very high velocity geostationary satellite missions, which don’t allow enough residual propellant for landing. In the longer term, missions like that will fly on Falcon Heavy, but until then Falcon 9 will need to fly in expendable mode.â€Â

Source: http://www.spacenews.com/article/launch-report/41350spacex-releases-footage-of-falcon-9-first-stage-splashdown

[NecroBones]

While this is true, there is one important thing to note. Debris that you jettison, i.e. stages, disappear after 2.5km, because they are still in motion. If you were sitting still/crashed/landed and THEN dropped your stages, they would stay as debris. This rule changes when you reach orbit.... anything you jettison in orbit technically is still moving when you go 2.5km away from it, however there are no physics to track other than position at that point, since it's not interacting with an atmosphere. Much easier and CPU friendly to track, therefore it is. I think the devs also knew that people would want to see their debris cluttering orbit, because it's realistic. Sources: Tons of experimentation. And, I slept at a Holiday Inn last night.

The 2.5km range is really the most important factor here. This thread went off in some wonderful tangents, but I wanted to just post one more reply to the OP question just for clarity's sake.

In actuality, the probe core makes no difference at all, except in whether the object is labeled as debris or not. But that's just a label. When recovering debris on the surface, you will get the full value of the recovery, but without the probe core, you simply don't get the UI popup that tells you so. If you try recovering debris without a probe core, and actually take note of your funds both before and after, you'll see that the refund does get processed.

But getting back to the destroyed object side of it:

As soon as ANY object (probe core or not, debris or not) meets all of these conditions, it is deleted as "crashed" or "burned up in reentry":

1. Not the currently controlled vessel

2. Not landed (in motion on a trajectory)

3. On orbital rails (more than 2.5km away from currently controlled object -or- in non-physical timewarp)

4.a. Under a certain altitude in the atmosphere (around 20 or 25km on Kerbin)

-or-

4.b. Collides with surface

As noted in the quote above, objects outside the physics range are on rails, for very cheap calculation. Since the physics aren't being calculated on these objects, the game has to make a "best guess" about what it should do if something is inside an atmosphere, and above the surface on some sort of trajectory. And that guess (currently) is always that it will be destroyed.

Right now there are some mods that will take parachutes into account and assume that it lands OK if there's enough drag (StageRecovery, DebRefund, for instance). In a way I think these mods refund a little too much money, but I would love to see this feature become stock. It creates an incentive to use parachutes and avoid purely disposable designs.

Edited August 6, 2014 by NecroBones