Payload Fraction Challenge - Rebooted

[camacju]

The payload fraction challenges were a nice way to squeeze as much as possible out of the stock parts in KSP, but they've all fallen into inactivity. Recently I've been playing around with efficient single stage craft, and I've been having a lot of fun trying to maximize the payload capacity. Also, since new optimization techniques have been discovered, I'm curious to see what the players can come up with.

Categories

Categories are arranged in two axes:
-Type of craft (single stage vs multi stage, rocket vs plane). All else equal I expect multi stage designs to have an advantage over single stage designs.
-Techniques used to design the craft (aero optimization)

In my mind I divide this axis into three parts: Purist, Technical, and Anarchy.
Purist: The craft must be physically possible. In other words, no clipping, unless that clipping is incidental and doesn't affect the craft's aero properties. Fairings are allowed if they don't clip through anything.
Technical: Clipping is allowed, as well as node occlusion and root fairings. These crafts are still bound by the physical properties of KSP's wings.
Anarchy: Any aero glitches are allowed. Stacked cargo bays, magic wings, et cetera. However, you must still use an actual engine for thrust - no Kraken drives.

Score is Payload Mass / Launch Mass. The payload cannot have wings or engines, and all resources must be full. (Clarification - the resources don't necessarily need to be full as long as you can prove that no resources were drained from the payload). The payload must be separated from the launch vehicle in a stable orbit (Pe > 70).

Leaderboard:

	Single Stage	Multi Stage
Purist	camacju - 0.614 (Plane - reusable)    Sival - 0.416 (Plane - reusable)    swjr-swis - 0.22 (Rocket - reusable)    pedter - 0.191 (Rocket)	Sival - 0.340 (Rocket)
Technical	camacju - 0.676 (Plane)    camacju - 0.664 (Plane)    zacspace - 0.405 (Plane)
Anarchy

If there are any suggestions for categories or rules, please let me know!

Edited September 30, 2021 by camacju

[camacju]

And here's my submission for the Single Stage Technical category.

This is a previous version of my craft showing how I'm able to take off from the runway. I use brakes to let the Rapier spool up before starting motion, giving me the extra few m/s needed to not hit the water.

Craft in VAB - 40,050 kg. One Rapier, one Nerv.

I'm at 99 m/s when I run out of runway and the craft can't easily maintain level flight until 110 m/s so I pitch up to 9 degrees and barely pull out of the dive before crashing.

After that, it's a pretty standard ascent to orbit. Accelerate until 1690 m/s on Rapier, turn on the Nerv, and point prograde until orbit.

In orbit

Detaching payload

Proof of no drained resources

26.58 / 40.05 = 0.664

[swjr-swis]

57 minutes ago, camacju said:

any suggestions for categories

There should be separate categories for spaceplanes and rockets. There's much too big a gap between those to just lump them all together.

No need for additional categories, but I'd be very interested in seeing the leaderboard mention how recoverable each launcher is (the % of recovery cost shown by the game; screenshot of launcher sans payload in SPH/VAB and screenshots of recovery of each stage should be provided if doing this).

I'll be watching this one, and will likely participate - with a bit of delay, this week is going to be a busy one.

[Pds314]

Is there a requirement that the spaceplane be able to land afterwards?

[camacju]

9 hours ago, Pds314 said:

Is there a requirement that the spaceplane be able to land afterwards?

No, but like swjr-swis said, recovery will be a bonus note next to your name on the leaderboard.

[camacju]

Okay - this isn't an official entry since the "payload" cannot be separated but at the very least it should serve as a good proof of concept for future entries. (Or I can just send up a kerbal to pull all the parts off using EVA Construction)

Category would be Single Stage Technical.

The main improvement comes from my recent attempts to make a maximally efficient orange tank lifter, and discovering that 1 nuke for a just under 55 ton craft was more efficient than 2. Additionally, I found that for the same craft, 2 rapiers is a bit too much for efficiency, and 1 rapier is a bit too little for comfort.

So I doubled the mass of the craft to just over 109 tons and included two nukes and three rapiers.

Craft in VAB - 109.323 tons

In orbit.

13494 liquid fuel remaining is 269.88 Mk0 fuel tanks which is 74.217 tons. Add the 0.1 tons of probe core, reaction wheel, and batteries for 74.317 tons, for 0.680 payload fraction.

[pedter]

Long time lurker, first time poster here with a Purist SSTO (rocket); to no one's surprising it is Mammoth-powered. Initial mass of 319,105kg and payload mass of 61,105kg for a payload fraction of 19.1489%. I suspect someone could improve aerodynamics and/or flight profile to improve the payload mass slightly but short of switching to a plane there's not much more that stock rocket motors can do.

Mechjeb couldn't figure out the dv for a whole two staging events (and shows initial mass as including the two 100kg hold-downs) but KER gets the liftoff mass and dv correct.

The 258t left behind is nothing but the Mammoth and three S3-14400 tanks (15t+81t+81t+81t); simple is good! Barely limped to 72x77km orbit so additional payload will be tight.

[swjr-swis]

10 hours ago, pedter said:

I suspect someone could improve aerodynamics and/or flight profile to improve the payload mass slightly but short of switching to a plane there's not much more that stock rocket motors can do.

In excess of 25% payload fraction has been done in past versions of this challenge, and since then some tanks have been improved, more efficient rocket engines have been added, and decouplers have become lighter. I'd be very disappointed if we don't beat past records this time around.

Good starting entry though, and welcome to the forums!

[camacju]

1 hour ago, swjr-swis said:

In excess of 25% payload fraction has been done in past versions of this challenge, and since then some tanks have been improved, more efficient rocket engines have been added, and decouplers have become lighter. I'd be very disappointed if we don't beat past records this time around.

On the other hand, the aerodynamics have changed significantly, so it's a lot harder now to get to an orbit efficiently.

My best payload fraction for a rocket SSTO (granted, it wasn't really very optimized) is about 20% with a Vector and two Nervs. I suspect 25% will be quite difficult to achieve in the purist category, but I could see a 40% design using some creative aerodynamics.

[pedter]

3 hours ago, swjr-swis said:

In excess of 25% payload fraction has been done in past versions of this challenge, and since then some tanks have been improved, more efficient rocket engines have been added, and decouplers have become lighter. I'd be very disappointed if we don't beat past records this time around.

I've had a look through the three old threads; there are lots of multi-stage submissions in the 26-28% range but the only SSTO submission that made the leaderboards is your own 21.19% D2/D1B SSTO from 1.2.2. The 2% up from 19.15% can very likely be wrestled back out of the vessel but I don't foresee any way for >25% SSTO.

[Zacspace]

I decided to see how one of my existing designs would fare. This SSTO is for the technical category since it has non-trivial clipping and a little fairing nonsense. it's launch mass is just under 126 tons and the three full ore tanks it lifts to orbit are 51 tons according to map view, for about %40 of it's takeoff weight.

I designed this plane to take less ridiculous payloads interplanetary, I'm surprised it was able to even fly this mission. It even had a little gas left over at the end so it could probably squeeze a few more percent out, but I don't think there's anything I could do to it to compete with the more optimized submissions already here.

 

Edited September 13, 2021 by Zacspace
Clarified that it's an SSTO since it kinda looks like a space shuttle

[Sival]

Hi all.

Here's my submission for the purist multistage rocket category. 

It's a rapier based design with an aerospike booster stage.

Launch mass : 31.245 t

Payload mass : 10.633 t

Payload fraction : 34.03 %

Imgur album

[camacju]

1 hour ago, Sival said:

Payload fraction : 34.03 %

Impressive for such a small rocket! Also, welcome to the forums.

[swjr-swis]

On 9/12/2021 at 8:59 PM, camacju said:

On the other hand, the aerodynamics have changed significantly, so it's a lot harder now to get to an orbit efficiently.

Aero was noticeably easier in 1.2.x, but in my experience it's only become easier after 1.3.1, if only marginally so.

 

On 9/12/2021 at 11:20 PM, pedter said:

I've had a look through the three old threads; there are lots of multi-stage submissions in the 26-28% range but the only SSTO submission that made the leaderboards is your own 21.19% D2/D1B SSTO from 1.2.2. The 2% up from 19.15% can very likely be wrestled back out of the vessel but I don't foresee any way for >25% SSTO.

I interpreted your previous words to refer to all rockets, not just single-stage. Still, it's possible to improve on the 1.2.2 numbers, if only because that challenge required a 100 km orbit, and now we can suffice with 70 km.

 

I was curious how my old D1b/D2 lifters would perform in 1.12. So here's an entry for the purist category, a single stage, recoverable rocket. 106.3 t of payload to a 74 km orbit, from a total launch mass of 483.19 t (not counting the 7 x 0.1 t of the launch clamps), results in a clean 22.0 % mass fraction. In the recorded attempt, I landed it all parts intact in the western desert for a 67.6% recovery value.

Spoiler

Note that it was initially tested with a Kerbal in the payload seat, hence slightly different numbers in my craft description, but the resulting mass fraction without the pilot still rounds to the same score.

Craft file: https://kerbalx.com/swjr-swis/SWiS-PMFC-SS-FRR1

Full album: https://imgur.com/a/5VLKBQw

 

 

[Sival]

22 hours ago, camacju said:

Impressive for such a small rocket! Also, welcome to the forums.

Thanks. I'm not sure the payload fraction would increase with scaling. To be tested...

Here's my entry for the purist SSTO reusable spaceplane category. Nothing special about this design, apart from the payload release mechanism 

Launch mass : 31.756 t

Payload mass : 13.640 t

Payload fraction : 41.64 %

Imgur album

[camacju]

Pretty clever deployment method

[camacju]

After realizing that it's possible to fit two Rapiers and two Nervs into a 1.25m fairing without clipping anything, I decided to give the Purist Single Stage category a shot.

60.684 tons launch mass, 37.240 tons payload mass. Payload fraction is 0.614

Spoiler

Spooling up engines with brakes on. The craft can't maintain level flight when it leaves the end of the runway so I have to fall down to sea level before it gets enough speed to fly.

And down it goes, with a very low altitude before it can pull out of the dive.

At 480 m/s I begin climbing to 18 km

At 1710 m/s I turn on the nuclear engines and let the Rapiers flame out

The nuclear part of the ascent uses about as much fuel as the airbreathing part, just because I lose a bunch of dv to drag. The high nuclear specific impulse still makes Rapier+Nerv the best way to go for a single stage craft.

Circularizing

In orbit!

Payload deployed

Payload mass

Because the fuel tanks aren't quite balanced this craft has to enter backward. (This screenshot is from an earlier attempt just to show reusability)

However the craft can still splash down just fine in the ocean. (Again, old screenshot to show reusability)

 

[swjr-swis]

17 minutes ago, camacju said:

60.684 tons launch mass, 37.240 tons payload mass. Payload fraction is 0.614

Question: shouldn't this also be in the technical category? I get that you're focusing on aero and clipping, but if the Purist category is defined by what is 'physically possible', perhaps free-floating gear and payloads phasing through the hull are out of bounds for that.

[camacju]

1 hour ago, swjr-swis said:

I get that you're focusing on aero and clipping, but if the Purist category is defined by what is 'physically possible', perhaps free-floating gear and payloads phasing through the hull are out of bounds for that.

For the landing gear, they're autostrutted by default so I just imagine some struts connecting them to the hull. I could just use bigger landing gear with a negligible penalty to fuel usage - I don't really think that matters. Or I could move the landing gear a bit upward and inward, to the same effect. Essentially this is just an aesthetic thing.

The payload deployment could easily be done by deploying the fairing and detaching the payload, and the craft would still be just as reusable, but I thought it looked better to do it this way. In my opinion it's no different from a cargo bay.

Edited September 15, 2021 by camacju

[18Watt]

1 hour ago, camacju said:

Because the fuel tanks aren't quite balanced this craft has to enter backward.

I often have more trouble accounting for engine mass than fuel burn with spaceplanes.  Enough so that I spend a lot of time trying to mount engines closer to the COM if possible.

With your design, I think you are stuck with having the engines mounted at the very aft of the plane.  However, since you are using elevators both in front of and behind the COM (canards and traditional elevators), I wonder if that plane could be flown backwards successfully?

It looks like you use some wing incidence, so flying backwards would be less efficient.  But after burning all your fuel and deploying the payload, I suspect it would still fly fine with the negative wing incidence.  Once deploying the payload, it doesn’t need to be efficient at all, just flyable enough to get on the ground.

[camacju]

2 minutes ago, 18Watt said:

I often have more trouble accounting for engine mass than fuel burn with spaceplanes.  Enough so that I spend a lot of time trying to mount engines closer to the COM if possible.

Well, I'm usually able to balance both even with engines at the aft. I place the fairing and engines first, then balance fuel tanks around that, put wings at COM, then put the payload at COM. The result is a stable plane at all points in the flight.

The problem here was that my initial plane had too much fuel in orbit, so I took off some fuel tanks and added them to the payload, which shifted dry COM too far backward.

4 minutes ago, 18Watt said:

I wonder if that plane could be flown backwards successfully?

It was surprisingly easy to fly it backward. The one problem was that it had no yaw stability because the vertical tailplane was now in the "front" of the craft.

4 minutes ago, 18Watt said:

It looks like you use some wing incidence, so flying backwards would be less efficient.  But after burning all your fuel and deploying the payload, I suspect it would still fly fine with the negative wing incidence.  Once deploying the payload, it doesn’t need to be efficient at all, just flyable enough to get on the ground.

I sidestepped this problem entirely by flying upside down and backward, and splashing down in the ocean. That way I can still benefit from wing incidence.

[18Watt]

2 minutes ago, camacju said:

I sidestepped this problem entirely by flying upside down and backward, and splashing down in the ocean.

Heh, I considered that also.  My brain is too focused on landing on the runway, and I felt adding a second set of landing gear (to land upside down) would detract from the mission goal- that would be extra weight which wouldn’t count towards your payload.

Hmmm.  If splash down, you don’t need landing gear.  The landing gear could be part of the payload.

I don’t have a suggestion for the yaw stability issue though.

[Sival]

4 hours ago, swjr-swis said:

Question: shouldn't this also be in the technical category? I get that you're focusing on aero and clipping, but if the Purist category is defined by what is 'physically possible', perhaps free-floating gear and payloads phasing through the hull are out of bounds for that.

Is "phasing through the hull" something like : detach payload -> time warp -> payload drifts through the fairing? Or something else? I'm fairly new to KSP and still wondering how the payload was deployed.

3 hours ago, camacju said:

For the landing gear, they're autostrutted by default so I just imagine some struts connecting them to the hull. I could just use bigger landing gear with a negligible penalty to fuel usage - I don't really think that matters. Or I could move the landing gear a bit upward and inward, to the same effect. Essentially this is just an aesthetic thing.

The payload deployment could easily be done by deploying the fairing and detaching the payload, and the craft would still be just as reusable, but I thought it looked better to do it this way. In my opinion it's no different from a cargo bay.

Struts add weight and drag. Same with bigger landing gears. Moving the tiny landing gears inward will create stability issues. It's not just an aesthetic thing.

And wouldn't jettisoning the fairing  mean your plane isn't fully reusable? You'd have to rebuild your hull before next flight (and the reentry should be interesting).

Anyways, more precise rules for categories would be helpful. I can see now that my purist design choices were way too conservatives.

[camacju]

1 hour ago, Sival said:

Is "phasing through the hull" something like : detach payload -> time warp -> payload drifts through the fairing? Or something else? I'm fairly new to KSP and still wondering how the payload was deployed.

Yes, that's correct

1 hour ago, Sival said:

Moving the tiny landing gears inward will create stability issues.

Landing gear stability doesn't matter since I take off completely flat on the runway and land on the water

[camacju]

Submission for Single Stage Technical. 109.6 ton craft brings 74.11 ton payload to orbit, for payload fraction of 0.676

In VAB, 109.6 tons

On runway

Accelerating

Climb

Nukes turn on

Engines cut out

Circularizing

In orbit

Payload deployed

Notably, this payload fraction is over 2/3. In other words, ignoring packaging, this launcher could deliver two fully fueled copies of itself to orbit.

Edited September 30, 2021 by camacju