Eve SSTO is impossible!

[GoSlash27]

Someone made a lifting surface comparison I find useful a while ago, http://forum.kerbalspaceprogram.com/threads/29788-How-to-calculate-lift/page3

All wings should have no drag when in level flight. I think the only difference between normal wings and control surfaces is in the AOA calculation, but what you mostly notice is that the small control surfaces have a lift to drag ratio twice that of swept wings.

I've reviewed that thread. The math in the OP is dead on, but the ramifications of what it means seem to be off.

Most parts have the stock drag coefficient of .2. This is multiplied by .008 and the mass of the part to give a "true" drag coefficient. Then it's plugged into the classic drag formula 1/2*p*v^2*Cd.

Wings are more complicated. They have 2 parts to their drag, parasitic and induced.

The parasitic drag works normally, except that it is multiplied by .02 instead of .2. This is critical to understanding how forces operate at speed in a KSP aircraft. The wings are already a fairly low-mass assembly, but that .02 multiplier means that parasitic drag from wings pretty much doesn't exist! I mean... it *does*, but it's so miniscule that it can be safely ignored.

The induced drag is a function of angle of attack. sin AoA * Cdi * p * v^2

At high angles of attack, high velocities, and high pressures, this value is huge. It can be much higher than the parasitic drag of whatever you're lifting. This can cause bizarre and counter-intuitive behavior that wouldn't be readily apparent, like shuttles flying backwards on reentry even though the CoG is ahead of the CoL.

Lift works almost exactly the same way as drag, except the AoA formula is a bit convoluted. It's not exactly correct, but I crib the function as Cl*sin(2*AoA). The wing produces no lift at zero incidence to the airstream and no lift at 90 degree incidence, with a maximum lift coefficient at 45 degrees. Again, this isn't exactly how lift works, but it's close enough for planning purposes.

What's critical to understand for our needs is that any aircraft can be analyzed for efficiency by simply dropping the "p*v^2" part and just looking at how it behaves in a global situation in any angle of attack.

Lift is Cl*sin(2A). Induced drag is Cdi*sin(A) and parasitic drag is mass of the aircraft (ignore wings)*.004; a constant.

If you plug these into a spreadsheet, you can see how a combination will behave for a given AoA. You can see which angle of attack yields the most lift for the least drag (most efficient) and what that ratio will be. You can see where drag from the wings overcomes parasitic drag, which tends to cause instability.

Long -> short, the efficiency of a wing isn't determined by it's overall lift or lift- mass, but rather it's lift to induced drag.

Best,

-Slashy

[GoSlash27]

Tsevion,

Do you happen to know what the current record is for an all wing SSTO? That is, no control surfaces used as wings? I'm at 875 @ 29.5 Km and I'm not sure further improvement is possible...

Thanks,

-Slashy

[cantab]

Doesn't matter. The hard part is getting to low Eve orbit from Eve's surface in one stage, once you're there it doesn't matter what you do.

Although a dodge would be to get refuelled on a suborbital trajectory. It's not really in the spirit of the challenge though.

[Red Iron Crown]

Although a dodge would be to get refuelled on a suborbital trajectory. It's not really in the spirit of the challenge though.

Up to the OP if it's not in the spirit, but even if not it would still be an impressive accomplishment in itself.

[Tsevion]

Tsevion,

Do you happen to know what the current record is for an all wing SSTO? That is, no control surfaces used as wings? I'm at 875 @ 29.5 Km and I'm not sure further improvement is possible...

Thanks,

-Slashy

Really not sure, although I'd be very curious... might be worth starting its own challenge for that. My own attempts to figure an absolute theoretical maximum by running unlimited fuel end up pretty similar to that though, running around a 0.5 TWR, and it's really finicky at that speed and altitude, you end up gaining altitude and losing speed, then losing speed and gaining altitude, and your orientation has a lot less effect than you'd hope. When running pure wing you seem to need a TWR of around 0.6-0.7 to be able to make orbit, and building an ion craft with that high a TWR and still having lift is really hard... and having the delta-V to make orbit is, I suspect, out of reach for an SSTO... it might be feasible using a lot of ion engines, a pair of wings, and a lot of drop tanks / booster stages.

When running using locked control surfaces as wings, on the other hand, you can make orbit with a TWR of less than 0.3. Some of my tests seem to be showing that if you use a lot of locked control surfaces as wings, you can reach orbit with pretty much any non-zero amount of thrust.

Edited July 12, 2014 by Tsevion

[Keldaria]

Up to the OP if it's not in the spirit, but even if not it would still be an impressive accomplishment in itself.

I'm not sure, if we're technically basing the challenge on the term SSTO, I believe a refueling before reaching orbit would constitute a second stage. Although, I would definitely be interested in seeing a suborbital docking/refueling in action because of just how hard that sort of thing would be I Imagine.

[impwarhamer]

Just a couple of crazy ideas:

Would it be acceptable to have a sort of ground mounted mass drive? By which I mean a massive ground mounted engine to boost the plane into the air.

Also, regarding the suborbital refuel idea, surely it would be an ssto if the refuel vessel stayed attached to the ship?

[GoSlash27]

I would say that refueling during the ascent would be staging if the refueling vehicle detaches, though I'd like to see it done!

Likewise for the mass driver; it imparts energy and doesn't go up with the ship, so it's a stage.

I'd also consider a reaction wheel 'copter to be cheating the physics engine, but highly entertaining...

[GoSlash27]

Really not sure, although I'd be very curious... might be worth starting its own challenge for that. My own attempts to figure an absolute theoretical maximum by running unlimited fuel end up pretty similar to that though, running around a 0.5 TWR, and it's really finicky at that speed and altitude, you end up gaining altitude and losing speed, then losing speed and gaining altitude, and your orientation has a lot less effect than you'd hope. When running pure wing you seem to need a TWR of around 0.6-0.7 to be able to make orbit, and building an ion craft with that high a TWR and still having lift is really hard... and having the delta-V to make orbit is, I suspect, out of reach for an SSTO... it might be feasible using a lot of ion engines, a pair of wings, and a lot of drop tanks / booster stages.

When running using locked control surfaces as wings, on the other hand, you can make orbit with a TWR of less than 0.3. Some of my tests seem to be showing that if you use a lot of locked control surfaces as wings, you can reach orbit with pretty much any non-zero amount of thrust.

I think it's a fine subject for a new challenge! It should help development for this challenge as well.

-Slashy

[Technical Ben]

Any rocket design would collapse under it's own weight, no?

[Red Iron Crown]

Any rocket design would collapse under it's own weight, no?

Conventional VTO rocket cannot combine enough dV with a high enough TWR. It's a mathematical non-starter.

[BagelRabbit]

I'm positive this doesn't meet the challenge criteria, but here's some good fun that you can have involving wings.

A few notes:

Making a really long chain of Cubic Octagonal struts really boosts this thing's performance. At the lengths I'm using right now, the craft can get to Duna.

While I used launch stability enhancers in the video (with comical results), any flat platform seems to work.

This works with control surfaces on or off, see picture. I'm not entirely sure if this counts as infinigliding, as such. Also, many struts!

Try it out for yourself. Even if it doesn't meet the challenge guidelines, it's still a very fun little bug to play with.

[cerberusti]

My eve ssto attempt made a great infiniglider.

I turned SAS on and noticed a slight vibration, then watched it quickly accelerate to 1km/s before disintegrating...

[GoSlash27]

Really not sure, although I'd be very curious... might be worth starting its own challenge for that. My own attempts to figure an absolute theoretical maximum by running unlimited fuel end up pretty similar to that though, running around a 0.5 TWR, and it's really finicky at that speed and altitude, you end up gaining altitude and losing speed, then losing speed and gaining altitude, and your orientation has a lot less effect than you'd hope. When running pure wing you seem to need a TWR of around 0.6-0.7 to be able to make orbit, and building an ion craft with that high a TWR and still having lift is really hard... and having the delta-V to make orbit is, I suspect, out of reach for an SSTO... it might be feasible using a lot of ion engines, a pair of wings, and a lot of drop tanks / booster stages.

When running using locked control surfaces as wings, on the other hand, you can make orbit with a TWR of less than 0.3. Some of my tests seem to be showing that if you use a lot of locked control surfaces as wings, you can reach orbit with pretty much any non-zero amount of thrust.

Try looking at it as "thrust to drag" instead of "thrust to weight".

The most efficient stock wing yields a theoretical maximum of 3.17 lift to drag (low alpha and high speed) This means it's efficiency is a maximum of 68%. In practice, it will be below this.

The cheapest cost of picking up a kilo using wings is 320 grams of drag. The cost of the wings themselves is negligible in this case, since their parasitic drag is so tiny.

So you need (in theory) a .32 t/w ratio minimum, but in practice it will be higher than that.

Compounding this is the fact that the optimal AoA varies with the mass being lifted. A higher wing load yields a lower peak efficiency at a higher AoA and more drag. A lower wing loading yields the opposite, but you have to go faster to maintain the same rate of climb, which makes for more parasitic drag.

It's sort of thrust to weight... but also not.

Best,

-Slashy

[GoSlash27]

Tsevion,

have you experimented with various ratios of wing to engines? Does there seem to be an optimal ratio there for the different types?

I don't know how to do infinite fuel... If we can show that adding wings makes things worse instead of better for an infinite fuel supply from 1 tank, then it would conclusively prove that ion SSTO is impossible without control surfaces as wings.

Best,

-Slashy

[Tsevion]

Tsevion,

have you experimented with various ratios of wing to engines? Does there seem to be an optimal ratio there for the different types?

I don't know how to do infinite fuel... If we can show that adding wings makes things worse instead of better for an infinite fuel supply from 1 tank, then it would conclusively prove that ion SSTO is impossible without control surfaces as wings.

Best,

-Slashy

Infinite fuel is in the stock debug menu, alt+f12 to bring it up, it's just a checkbox. as for wing to engine, I'm not sure... I've been playing around and it seems a high ratio is better... something around or more than 2 swept wings per ion engine seems ideal... I need to do some further experiments with even more wings, but around that point the actual weight of the wings starts creeping up, lowering TWR too low.

[Tsevion]

Bit of a breakthrough, I've made a craft with no control surfaces, that with infinite fuel can make orbit... the secret is moar wing:

Javascript is disabled. View full album

I ended using two swept wings and a delta wing per engine, two engines a single xenon tank between them and a small probe core. Ground TWR maxes out around 0.4, but she's flies to the heavens. Seeing if I can use this knowledge to get higher legitimately.

Craft File

[GoSlash27]

Just to let everyone know, we're running a parallel challenge to create an "all- wing" SSTO ion glider from Kerbin.

This eliminates any possibility of infiniglide by restricting entries to one locked vertical control surface. If we get one into orbit there, then this just might be possible....

Regards,

-Slashy

[GoSlash27]

After what I've seen in the Kerbin SSTO thread, I think an all- wing Eve SSTO ion glider is unworkable. Not because we can't find the correct balance of parts, but because any craft that could pull it off would be kraken bait. You would need a minimum of 12 panels per engine to make it run full tilt, but a realistic number for an Eve burn is going to be more like 36. Plus you've got more than twice as many wings and probably 3-5 tanks. Throw in all the cubic struts, and you wind up with a hot mess.

[GusTurbo]

Kraken bait? Do you mean it will have problems running at high physics warp?

[GluttonyReaper]

Kraken bait? Do you mean it will have problems running at high physics warp?

No, it's just ... terrible. From the tests I did on Kerbin, where I attempted to use cubic struts and static panels to create power for a very low weight gain, the game just lags horribly, to about 10 seconds per frame, even though I'm only using ~350 parts. I think that it has something to be with the physics-less nature of the parts. Seeing as I was only using seven ions in my tests, and far more would be necessary to achieve Eve SSTO, I doubt KSP could deal with a suitable craft.

[Galane]

Just to toss something out there, here's a shot of my all stock rocket SSTO I built in .21.1. http://pastebin.com/3t3CEuNu IIRC at the time the only mod I had was MechJeb and a set of command part cfgs to integrate MechJeb. Oh, I also challenged myself to make it work using the default ascent profile in whatever official release version of MechJeb was current then.

SSTO FTW by g_alan_e, on Flickr

Got from the KSP pad to LKO with just a whiff of fuel left. It wasn't just lifting a bare bones lander can, there's a two stage lander on top, and none of its fuel was used in the ascent so it could land back on Kerbin. After that bit of fun, I re-arranged the staging and some fuel pipes to asparagus it and went to Mun and back for the second time (after in-orbit refueling of what made it to orbit).

All rocket stock SSTO from Kerbin = doable but bulky (if you don't use the new parts in .23.5). All rocket stock SSTO from Eve = nuh-uh, no way.

Edited July 17, 2014 by Galane

[GoSlash27]

No, it's just ... terrible. From the tests I did on Kerbin, where I attempted to use cubic struts and static panels to create power for a very low weight gain, the game just lags horribly, to about 10 seconds per frame, even though I'm only using ~350 parts. I think that it has something to be with the physics-less nature of the parts. Seeing as I was only using seven ions in my tests, and far more would be necessary to achieve Eve SSTO, I doubt KSP could deal with a suitable craft.

Pretty much this. The high number of physics free parts lags the game and moving all those parts during the lag can summon the kraken.

[kepicness]

In a realistic setting infiniglide wouldn't exist, but I'd expect electric propellers to. I think that at least a single control surface should be allowed, considering that it's necessary for control and that infiniglide is an unavoidable glitch in the game's code.

[GoSlash27]

Update:

I got to wondering whether the new update changes things with the O-10 monopropellant engine. It doesn't...

Maximum DV from the O-10 engine assuming no payload is 4,619 atm and 6,090 vac. Neither is enough for SSTO from Eve.

The vernor thruster fares worse due to it's poor Isp.

After playing around with ion gliders for half of forever, I'm willing to call this one busted. We still have yet to see one reach Kerbin orbit without control surfaces as wings, and Eve is a much taller order.

(edit)

Galane,

I have designed a Kerbin all rocket SSTO that weighs 10 tons when configured to lift a Mk1 pod. I have another 5 ton model that doubles as a rover. It doesn't take much rocket to pull it off. But Eve requires a lot more DV; it can't be done at any size.

Best,

-Slashy

Edited July 31, 2014 by GoSlash27