Tail-sitting Spacecraft: How?

[SSgt Baloo]

In the '50s, almost every square-jawed spaceman hero flew a rocket that took off from and landed on its tail. Making one that is supported by its tailfins on the launchpad is easy, but bringing one back to Kerbin tail-first is something of a challenge. In my case, the challenge seems impossible, but I'm hoping that's because I don't know something that someone else might share.

So do tell. Have any of you been able to successfully design ships that land on their tail, which can do that trick in an atmosphere?

Visual aids:

Edited May 23, 2016 by SSgt Baloo

[bewing]

It's easy if you have canards on the top of the rocket, and a big enough engine. I assume that you aren't going to allow any parachutes?

The basic deal is that those "fins" on the tail have to be wings. And then you need enough canard on the front to give you aerodynamic control at several km altitude at a few hundred m/s. You come zooming down out of the sky, then raise the nose slowly into a stall over your target. Once you are vertical and falling retrograde, it's just a standard tail-first landing that you would do on the Mun or whatever -- but you need a lot more thrust, since this is in atmosphere and high gravity. But if the engine is good enough to get you off the pad in the first place, it can get you back down again, of course.

[Norcalplanner]

I would probably depart from the classic look and put a few drogue chutes near the nose. The other trick is to enter the atmosphere backwards with the wings perpendicular to the ground so they generate no lift. You need good control authority to hold a retrograde attitude, though.

[Venusgate]

I do this with ssto rockets. you pretty much have to lock retrograde SAS, enter atmo shallow enough that the engine and fins can take the heat, and not have TOO many fins. I also turn the control surfaces off, as they'll overcorrect and turn you around.

[bewing]

OK, I slapped together a rocket that lands on its tail on Kerbin without parachutes. It's an 11.5 ton MK2-based thing.

Of course, landing on your tail in high gravity takes a lot of fuel. Especially with my sort of design, because you can't exceed 15 m/s during the landing phase, or the thing will flip ends on you. This one takes about 300 fuel to land.

If you want a craft file to try out, just say so.

[Foxster]

One way that works is to not have the tail fins actually take the weight. Instead build an I-beam frame and then mostly cover it with the tail fins. 

Here's some inspiration...

 

Edited May 24, 2016 by Foxster

[Sereneti]

13 hours ago, SSgt Baloo said:

In the '50s, almost every square-jawed spaceman hero flew a rocket that took off from and landed on its tail. Making one that is supported by its tailfins on the launchpad is easy, but bringing one back to Kerbin tail-first is something of a challenge. In my case, the challenge seems impossible, but I'm hoping that's because I don't know something that someone else might share.

So do tell. Have any of you been able to successfully design ships that land on their tail, which can do that trick in an atmosphere?

Visual aids:

Challenge accepted

[cantab]

I'm working on this very concept recently! Here's my first successful "tailsitter" spaceplane, the Shuttle Gallifrey.

 

Some of the things I've learned:

The design is actually nice aerodynamically, at least in FAR. The weight does need to be far back, I get the CoM in the aftmost of the three cargo bay sections, and then the swept wings put the CoL in the right place and the elevators are far back to have good control authority.

With the aid of the cross-struts at the I-beam "legs" the wings take the weight no problem, on 0.5 strength in FAR at least. Heck, I'm now working on a full Shuttle version with big Kerbodyne tanks and it still sits fine on its wingtips before takeoff.

Takeoff and ascent is mostly a doddle, loads of power. The Panther with its afterburner is your friend as a VTOL engine. Do watch for having enough air intake though. My biggest difficulty has been not overheating the nose, because the pointy nosecone has a lower heat tolerance. I put an antenna on the nose and a little fuel in the 1.25-2.5m tank and that helped draw the heat away but I still need to watch the thermals and be ready to throttle back.

Re-entry isn't too hard. Maintain a good angle of attack, control your vertical speed, use S-turns if required, and don't forget the option of pumping fuel fore and aft for balance.

The tricky bit is, of course, landing. My approach is to glide in low and slow but not too slow, then pitch up hard to vertical. I gradually increase the power to make sure the nose doesn't drop away, and Vernors and plenty of SAS help yank it up. Once it's pointing at the sky I hit hold radial out on the SAS. If I do it well I'll gain about a kilometre in this manoeuvre, if I do it badly I'll gain several.

Once in the hover I can't descend too quickly, the plane is after all unstable flying backwards. 30-40 m/s is about the maximum. For this reason I feel jet engines are basically mandatory for the landing, rockets will just guzzle too much fuel. Controlling the descent speed requires practice, a delicate touch on the throttle, and careful attention to the TWR reported in KER.

You need to ensure there's enough air for the jets. I put scoop intakes facing backwards for this, and ramp intakes face forwards. With such an arrangement be aware that least airflow tends to be in the -10 to 10 m/s speed range. You really don't want one of your lift engines flaming out on you.

Beware of having too little TWR as well. The design above has two RAPIERs and two Panthers, and that's enough with no payload, but trying to land with a 5 ton payload put the TWR barely above 1. And thrust is less the higher you are. That meant cases where I would accelerate in my descent and either flip over or hit the ground hard. The successor design has switched to one RAPIER and four Panthers, to give ample thrust for landing without much of an increase in weight actually.

The final touchdown needs to be gentle, a few m/s. The wings give some "spring" and mean it likes to bounce, and it's very easy to tip over. The Vernors up front are meant to help keep it steady of course. If you're playing without quickloads and reverts, don't be afraid to just hit Z and abort the landing if it looks about to tip.

Edited May 24, 2016 by cantab

[Kerbin vonKerbal]

I would say your best bet is to keep the CoM leaning towards your desired tail-side, add some control/landing surfaces on your tail, and just burn retro.

 

Biggest thing would probably be keeping your CoM at the bottom so your tail stays rear-oriented. I could be full of garbage though, but that would be my approach at the situation.

[Brainlord Mesomorph]

in atmosphere, just put a bunch of parachutes on the nose. (I sometimes use rocket-assist w/ chutes)  As long as you come down on level terrain you're good.

non-atm: I don't use tall tail sitters. No reason for it. They are either very short and squat. or I land them on their side, just easier.  (you can't fall over if your're already lying down)

Edited May 24, 2016 by Brainlord Mesomorph

[cantab]

45 minutes ago, Kerbin vonKerbal said:

I would say your best bet is to keep the CoM leaning towards your desired tail-side, add some control/landing surfaces on your tail, and just burn retro.

 

Biggest thing would probably be keeping your CoM at the bottom so your tail stays rear-oriented. I could be full of garbage though, but that would be my approach at the situation.

If you're flying it like a rocket then I think that would work. You'd be relying on the significant fuel drain to shift the CoM from up top during launch to down low for re-entry, and you wouldn't be using aerodynamic lift much for approach and landing. It'd be a lot like landing a Falcon 9 style first stage, which I've seen several KSP streamers do.

The approach I discussed above was rather to fly it like a spaceplane. My own preliminary studies indicated it would be rather hard to make something that can fly aerodynamically forwards and backwards, hence the need to enter the reverse hover for landing.

On chutes, I tried them to yank the nose around and get in the hover, and it works great! I just decided I didn't like using them.

Edited May 24, 2016 by cantab

[Kerbin vonKerbal]

1 minute ago, cantab said:

The approach I discussed above was rather to fly it like a spaceplane. My own preliminary studies indicated it would be rather hard to make something that can fly aerodynamically forwards and backwards, hence the need to enter the reverse hover for landing.

Ah, my apologies. I admit I just skimmed through and assumed you were asking for guidance in landing a rocket. 

[Tiber9]

Now I want to try this, and those example pics are awesome. My initial thought would be that fuel pumping for mass adjustment could allow for two flight modes, especially if those tail fins aren't too huge and there are canards or vernier thrusters. The space-plane approach with last minute maneuver is awesome too. I wonder how you imagine the crew is oriented? Like the mk1-2 or mk3 pods or like the hitchhiker? 

[KerikBalm]

Well, a single stage design in KSP typically has its CoM shift back towards the heavy engine as the fuel is burned off... so that works well.

It should be possible to havethe CoM ahead of the CoL when its full (ie, for ascent when you want the point end facing forward), and the CoM behind the CoL when its nearly empty and you just want to do a propulsive landing.

For my only tailsitting spaceplane... I only use it on Duna... and it refuels in orbit, not on the ground (as, at least in my career game, the lowlands of duna have 0 ore, and I instead supply an orbital fuel depot from ore&fuel from Ike). As a result, it lands, in an atmosphere, with full tanks.

 

I have numerous SSTO vertical rocket designs that land as tailsitters on kerbin, all using parachutes though... its just easier... they are stable pointing retrograde for reentry... and stable flying "backwards"... I just use the chutes for simplicity

[Laie]

 

Can land and SSTO from anywhere but Eve and Kerbin. Can still land on and lauch from Kerbin if you refuel on the ground. Worked in 1.0.5. When accompanied by a ISRU clipper, it can visit any body, even those where the ISRU vessel can't go. With generous strap-ons, it can even handle Eve (that's part of what the docking ports are there for).

Aerodynamically she's barely stable, but that's a boon: Interesting games can be played as soon as there's some room in the tanks. By moving around the fuel you can make it so that she assumes a stable attitude, entering belly down and the nose 20-70 degrees above prograde, as you desire. You can also make her come down tail-first, but that's not good for reentry.

In oder to land on Laythe with enough fuel for the return (= lots of extra mass), you need to do a juicy de-orbit burn of 800m/s or so, otherwise she will go up in flames no matter the attitude.

Without gimbals and overall little control, I opted for a parachute landing. On bodies without atmosphere (or Duna), all these considerations don't matter and she just works.

The engines are surface-attached to the wingtips, the tanks stacked on top of the engines. Will drain the outriggers top-down, then the core. No fuel lines needed. Becomes aerodynamicall unstable when the outer top tanks are drained: stand by with the fuel pumps (or lock/unlock tanks) when launching from Kerbin. Anywhere else you won't take off with full tanks and can trim her ahead of time, while still sitting on the ground.

 

Edited May 27, 2016 by Laie