Heavy SSTO Efficiency Challenge

[herbal space program]

What I want to know Nefrums is what have you got on the front of your plane to keep it from exploding off? I couldn't get anywhere near the speeds you did without my small nosecone exploding. Towards the end last night, I figured out that the small prodecedural fairing could take a fair bit more heat that the nosecones could, but still 1600m/s anywhere down in the soup below ~25km seems lethal to me, and I can't accelerate and climb at the same time breathing air that high up. Is it just that your TWR is so high you hardly even need wings? Judging from the scant wing area you are sporting, I guess that must be it. What is your takeoff speed?

8 hours ago, mk1980 said:

i've had some success with using the Large or Extra Large landing gear (with the suspension slider set to something higher than the default "1") as main gear fairlyclose to the CoM and using one (or 2) medium gears on the front. medium gears a lot shorter, so it's a bit of a hassle to place all the gears so the plane is still somewhat level, but the medium gears in the front are steerable, so that's an advantage.

i've experienced some very annoying "swerving" on the runway when the main gear is not perfectly parallel. that can easily happen when you attach it to something curved like BigS wings. that's especially true if you angle the wings upwards to generate more lift in level flight.

it's often easier to attach it to one of the the mk3 parts of the main fuselage and use offset/rotate to move it a bit outwards.

regarding the cargo bays: i think i read somewhere that there may actually be a bug with the aero overlay/gui. ie. the parts in the cargo bay actually don't produce drag, but the aero overlay still shows the alarming super long drag vectors. not sure what's actually going on there. TBH i never found much use for cargo bays in career, so i rarely use them.

I'm currently using 2 medium gears close together in the front and two fairly far part in the back for a plane that weighs in at a bloated 190t, with the dampers set all the way to the highest setting. Those gave me way less drag than the big ones and actually rolled down the tarmac pretty smoothly. I did not start getting swervy at all until >80m/s, and all I needed to do was up a little to stay on course.

[cerberusti]

The shielded docking port makes a superior spaceplane nosecone when you plan to hit high speeds in the atmosphere, although you will need a bit more engine to push it supersonic (I usually dive a bit).

I have been using no spring and maximum dampening on my gear recently, it seems to help with the insane wobbly bounce.  If putting gear on wings it helps to not only use the rotation as described above to make sure they are straight, but to angle the entire craft upwards a little bit so the back gear makes contact first and you avoid dropping the heavy rear of the plane hard enough to tear the wings off.

How far can we take the detachable part?  Can I move the payload from low orbit out to 300k with a detachable tug and dock again before landing?

Edited June 16, 2016 by cerberusti

[herbal space program]

13 minutes ago, cerberusti said:

The shielded docking port makes a superior spaceplane nosecone when you plan to hit high speeds in the atmosphere, although you will need a bit more engine to push it supersonic (I usually dive a bit).

I think being at all competitive in this particular challenge requires a level of TWR that makes going supersonic something that happens in the first 2 minutes of flight, so the extra drag shouldn't me much of a concern. I'll test it against the procedural fairing tonight and see what works better. I'm starting to think that what I mostly need to do to be competitive with the leaders here is to add one more engine and drop some wings. Trying to get to space using lift more than TWR seems to be a losing strategy for these massive ships.

[mk1980]

i always put shielded docking ports on the nose. i suspect a heatshield would have a similar effect, but it's overkill actually. i honestly never managed to blow up a shielded docking port. the other big adavantage of the docking port is of course that the nose is by far the best location for a docking port - with that configuration you can dock a >100 ton plane without even adding rcs thrusters, just some careful trhusting with the main engines to keep prograde & target vector overlapping.

i did try various thrust levels to get to orbit and i'm now convinced that minimizing engine count is not worth it. my first entry had only 8 rapiers for ~240 tons. the runway TWR was 0.32. it made its way to orbit, somehow, so it does actually work. the follow up plane had a bit less fuel mass and used 12 rapiers and in the end it was more efficient. the final submission (without cargo bay) also had 12 rapiers and even less fuel (only ~180 tons fully fuelled) with a runway TWR of about 0.7 - and it used by far the least amount of fuel of the 3 designs. so my conclusion is simply that you don't make the plane more efficient by using less engines. it just ends up taking far longer to get it superspnic and climbing. by the time the 8 rapier thing finally got to ~400 m/s, it had already burnt like 4000 units of LF.

[Nefrums]

Try to hit mach 1 in under 1 min and mach 4 after 2 min. 

[herbal space program]

6 minutes ago, Nefrums said:

Try to hit mach 1 in under 1 min and mach 4 after 2 min. 

1360m/s after TWO minutes? Yowza! I thought 450m/s after two minutes was fast

[Nefrums]

I improved the stingray design to comply with all the rules.  

Adding another two engines made it a bit better.  It does 0 to Mach 5 in 2min 6s. 

5770 LF, 1980 OX and 8 mp -> 4982p

 

 

[herbal space program]

15 hours ago, Nefrums said:

I improved the stingray design to comply with all the rules.  

Adding another two engines made it a bit better.  It does 0 to Mach 5 in 2min 6s. 

5770 LF, 1980 OX and 8 mp -> 4982p

 

"Abandon all hope, Ye who enter this challenge!" '

D'ohh!  After adding Moar Boosters and pulling a bunch of wings off based on your speed profile, I was finally getting close to beating your prior scores with the rig below, but I don't see how I can beat your current one without starting over, trying to copy your design, and using MechJeb. Flying planes that handle so poorly at such speeds manually is just too harrowing. I'm finding now that if I pull up just a little too much during my early ascent, my plane instantly disintegrates from the aerodynamic forces. I also had to stop using the FAT -455 Main Wing, which made for a nice looking plane,  because it started exploding from the heat!

Anyway, I'm not submitting this for scoring yet, as I came up just a tad short making orbit on my last run and had to go to bed, but I'm  estimating I could get a score just under 6,000 if I do everything right. We'll see about that, but if I can manage it 'll be hanging up my T-square and hoping for second or third place.

Edited June 17, 2016 by herbal space program

[zolotiyeruki]

10 hours ago, herbal space program said:

"Abandon all hope, Ye who enter this challenge!" '

D'ohh!  After adding Moar Boosters and pulling a bunch of wings off based on your speed profile, I was finally getting close to beating your prior scores with the rig below, but I don't see how I can beat your current one without starting over, trying to copy your design, and using MechJeb. Flying planes that handle so poorly at such speeds manually is just too harrowing. I'm finding now that if I pull up just a little too much during my early ascent, my plane instantly disintegrates from the aerodynamic forces. I also had to stop using the FAT -455 Main Wing, which made for a nice looking plane,  because it started exploding from the heat!

Anyway, I'm not submitting this for scoring yet, as I came up just a tad short making orbit on my last run and had to go to bed, but I'm  estimating I could get a score just under 6,000 if I do everything right. We'll see about that, but if I can manage it 'll be hanging up my T-square and hoping for second or third place.

Yeah, I feel the same way now...  I got to a 300x300km orbit using 7200 LF and 3960 oxidizer.  That's in the ballpark of Nefrum's earlier run, but now he's taken it to a whole new level.

This is not an entry, since I couldn't land it--the NERVs were blocked by the rear wing after docking, so I couldn't de-orbit.  Since I had 1k+ extra LF, I'm going back, taking off a couple fuel tanks, and reducing my oxidizer load for attempt #2.

 

[Mikki]

@icedown

Would be glad if you could separate your challenge into "Cargobay-planes" and "other"...

[herbal space program]

7 hours ago, Mikki said:

@icedown

Would be glad if you could separate your challenge into "Cargobay-planes" and "other"...

I think the more important distinction might be between planes flown manually and those flown using MechJeb. Between the extreme speed of the ascent phase and the totally messed-up wheels, actually flying planes that are competitive manually is very difficult, and using MechJeb pretty much eliminates that challenge. There's really no way a fly-by-stick pilot can compete. Also, I have just given up in disgust after spending 4 hours trying to land my mission in vain, because the wheels are so bugged even a perfect touchdown isn't good enough. The one time I was able to roll to a more-or less stop on the KSC grounds in almost one piece (DQ'd because I knocked an engine off as I veered off the runway), my plane just kept doing this silly little dance as the wheels pumped phantom energy into the system, moving this way and that and never coming to a halt. Engaging the brakes only made it worse, and  when I went into 2X time warp, the plane went nuts and literally shook itself apart. It is a real drag that none of this is getting fixed until 1.2, because this is worse than all the other bugs put together IMO.

[icedown]

Multiple categories for the delivery method I am considering.  Separation by chosen control method will not be done.   This is about efficiency, not pilot proficiency. 

@zolotiyeruki Keep it up!  That looks like a good Ship

Edited June 19, 2016 by icedown

[mk1980]

@Nefrums latest design was a bit... disheartenging to be honest. my ugly "open end" bird was probably the best i could manage, and it's stil about 50% worse then the new design.

i suppose i could maybe squeeze out some more efficiency by stealing some of his ideas, but that would be pointless i guess   also i'm not going to install mechjeb. i guess it would help (somewhat), but i don't really like the concept of an autopilot. so i guess i'll throw in the towel. maybe i'll make another entry at some point if i come up with some clever idea, but i doubt it will be much better than what i already posted. i guess ~100 funds per ton to a 300km orbit is not too shabby.

[Nefrums]

Smart A S S is not really an auto pilot. It is just a better sas where you can specify the exact angle it should try to hold.

[herbal space program]

20 hours ago, icedown said:

7 minutes ago, Nefrums said:

Smart A S S is not really an auto pilot. It is just a better sas where you can specify the exact angle it should try to hold.

That makes a HUGE difference when trying to fly a plane that fast and poorly maneuverable to orbit. Why don't you try a few times without it and see where you end up?

 

Edited June 20, 2016 by herbal space program

[Nefrums]

It sure does, but I enjoy the building part of this game more than I do the flying part. So I rather spend the time improving my designs than trying to get a good ascent. So I like the predictability and reproducibility of using  smart A S S. 

And it is true as you say that it is hard to compete with out using it. In the same way as it is hard to compete with smart A S S vs someone using kos scripts.

Edited June 20, 2016 by Nefrums

[herbal space program]

2 hours ago, Nefrums said:

It sure does, but I enjoy the building part of this game more than I do the flying part. So I rather spend the time improving my designs than trying to get a good ascent. So I like the predictability and reproducibility of using  smart A S S. 

And it is true as you say that it is hard to compete with out using it. In the same way as it is hard to compete with smart A S S vs someone using kos scripts.

And let me stress that I'm not claiming you did anything wrong. OP is fine with MechJeb. I just felt it would be fairer if there were a separate category for those who didn't use it, but OP said no. I suppose those of us who don't can just mention that fact informally, and maybe without creating a separate category OP might be kind enough to make a notation to that effect on the leaderboard.

Edited June 20, 2016 by herbal space program

[icedown]

This is my personal logic around the decision about autopilots.  Efficiency is about using all of the tools available to you in the best way that you can.  In aviation we use autopilots for a couple reasons. First is we are lazy.  Second though is that these devices are able to react much faster, more precisely, and more consistently than humans.  Every attitude change takes energy from your craft in the form of drag as the aircraft changes its flight path.  Another reason is passenger comfort or force control on materials.  Pilot induced oscillations are problematic and a very well documented phenomenon. This can cause people to become sick and loads to strain back and forth against hold downs.  The final reason is they are not as prone to operational error as pilots are. 

One other way to look at this challenge.  It is a design challenge, to design and build a heavy SSTO spaceplane that doesn't rely on exceptionally gifted crews to fly.  Think of this challenge as a bid for a contract rather than a personal style achievement.  Both sides have presented designs that are very good and have some diversity in them.  There will be winners, as this is a challenge with that kind of goal, but I am considering a completion badge also for people who have and fly their own unique designs.

NOTE: I do not like "participation trophies." The completion badges are from my thought while looking at these ships of: "That is an awesome design idea, it may not win but it's creativity and ingenuity definitely deserves something"  They will not be earned with close copies of other designs.  While not originally part of this challenge, several designs have made me change my mind about it.

Edited June 21, 2016 by icedown

[Hodari]

2 hours ago, icedown said:

NOTE: I do not like "participation trophies."

I think there is a bit of a difference here though.  The problem with participation trophies is when they are completely meaningless because you get them no matter how badly you do.  You don't have to actually accomplish anything in order to EARN them.  Whereas in this case, just completing the challenge at all is reasonably difficult, so having a badge for it would be fine since it DOES actually mean you achieved something. 

[icedown]

2 minutes ago, Hodari said:

I think there is a bit of a difference here though.  The problem with participation trophies is when they are completely meaningless because you get them no matter how badly you do.  You don't have to actually accomplish anything in order to EARN them.  Whereas in this case, just completing the challenge at all is reasonably difficult, so having a badge for it would be fine since it DOES actually mean you achieved something. 

It is a participation trophy if you minimally modify another design.  The K-Prize covers the completing of this. 

[mk1980]

just to be clear - i didn't mean that "autopilot" comment as a criticism. I have no problem when people use such mods.

[icedown]

3 hours ago, mk1980 said:

just to be clear - i didn't mean that "autopilot" comment as a criticism. I have no problem when people use such mods.

Not a problem.   I knew that this issue would come up when I created this challenge.  Just wanting to make the reasons for my choices available.  I know that most challenges discourage them if not ban them, but what I wanted to see out of this challenge was different from those and you guys are delivering splendidly.  I'm working on my entry, though it won't be part of the competition.

[TaxiService]

Here's my submission to this challenge. Thanks to Nefrums's design, I revised my plane design to successfully fly, dump payload and land on KSC runway. It gots RCS and docking port.

 

 

Fuels before the flight rollout (excluded the payload):
LF: 13835
OX: 8277
MP: 250

Fuels at 300x300km orbit (excluded the payload):
LF: 725
OX: 0
MP: 161

Fuels at KSC runway after the trip:
LF: 444
OX: 0
MP: 88.15

Ascent profile:
It is little complex profile (almost copied from Val's Chibi SSTO 1.0.5 plane)
1) Throttle at 20% until >10 m/s on runway to prevent air starving of two intakes. Throttle to 100% after.
2) Set AoA to 0 at sealevel to accelerate until 7km altitude is reached with over 500m/s
3) Set AoA to 0 again at 7km altitude to accelerate to 1000m/s
4) Set AoA to 10 upon 1000m/s
5) Set AoA to 15 upon 1200m/s before 19km altitude
6) At 22km altitude, go rocket mode and pitch up until no more ox
7) Outside 70km atomsphere, turn on the nuke and climb to 300x300km

Mods used: KER, Alarm clock and MM.
Plane name: Fire Spirit

Points = (13835-444)*0.8 + (8277*0.18) + (250-88.15)*1.2 = 12397

P.S. It is possible to fly with 8 rapiers and 2 nukes with the ascent profile above. Even better, I could drop by 200 points by replacing the MP tank and RCS with LF+OX RCS. But I don't feel like I would try any soon.

(deleted the "you tried" badge after reading the previous pages and discovering I actually got fewer points than other initial designs of other players)

 

Post-challenge: Downsize this plane for my own SSTO launch to 75x75 km orbit while maintaining the same payload.

 

 

Edited June 21, 2016 by TaxiService
More details

[herbal space program]

 

13 hours ago, icedown said:

This is my personal logic around the decision about autopilots.  Efficiency is about using all of the tools available to you in the best way that you can.  In aviation we use autopilots for a couple reasons. First is we are lazy.  Second though is that these devices are able to react much faster, more precisely, and more consistently than humans.  Every attitude change takes energy from your craft in the form of drag as the aircraft changes its flight path.  Another reason is passenger comfort or force control on materials.  Pilot induced oscillations are problematic and a very well documented phenomenon. This can cause people to become sick and loads to strain back and forth against hold downs.  The final reason is they are not as prone to operational error as pilots are. 

One other way to look at this challenge.  It is a design challenge, to design and build a heavy SSTO spaceplane that doesn't rely on exceptionally gifted crews to fly.  Think of this challenge as a bid for a contract rather than a personal style achievement.  Both sides have presented designs that are very good and have some diversity in them.  There will be winners, as this is a challenge with that kind of goal, but I am considering a completion badge also for people who have and fly their own unique designs.

NOTE: I do not like "participation trophies." The completion badges are from my thought while looking at these ships of: "That is an awesome design idea, it may not win but it's creativity and ingenuity definitely deserves something"  They will not be earned with close copies of other designs.  While not originally part of this challenge, several designs have made me change my mind about it.

Truly, if that is all you are after with this, you already have your winning design. It seems like you know a lot about actual aviation, but not really about this game.  I mean, you do understand this is a coarse-grained simulation with only a very limited number of design options, don't you? Nefrums' plane, given your rules and within the very limited design space of available stock parts, is not going to get beaten by more than a couple of percent. It is also not flyable by anybody not using at least smart A.S.S., because to get to that kind of speed performance, you need to reduce the wing and control surface area to the point where the stock SAS won't even exert enough control authority to hold your attitude. I believe the devs actually made it that way to force you to create more plausible designs, and using Smart A.S.S  basically papers over that deliberately created design constraint. Trying to compare it to using an actual autopilot, with the associate real life rationales, is actually a bit laughable. Moreover, there are simply not enough options within the stock part set for anything that doesn't look a whole lot like Nefrums' plane to have a competitive score, and now you have apparently decided post hoc to award the promised badge only to those who came up with different solutions, of which there are none that are competitive. Every single part on that plane is where it is for a reason that is dictated by the design constraint of your challenge, as you would well know if you had bothered to create your own entry first, which by the way is part of the official challenge guidelines in case you haven't read them. I now understand all too well why that is the case. Also, the whole point of offering a badge is to motivate people to spend hours of their precious free time working of their own volition on something that you specified. When they all arrive at similar solutions because that is in fact the correct answer, rewarding only the first participant, who had that craft in their back pocket already, is truly, deeply obnoxious, and frankly has me a bit steamed. I do appreciate though that you have taught me never again to participate in a challenge where OP has not submitted their own entry in advance.

Edited June 21, 2016 by herbal space program

[icedown]

@herbal space program  These are the rules of the challenge which I made clear in advance.  You are free to participate or make your own challenge in this forum as you see fit. I will not continue to discuss them or modify them further. I did not submit mine because I am hosting this challenge.

The only change that is going to take place is 2 categories, External Carriage and Cargo Bay carriage.  There will be Top and Runner Up badges made.