Sanzuwu SSTO

[Elmetian]

Horizon Heavy Industries presents the Sanzuwu single-seat SSTO. Its namesake is the three-legged sun crow that appears in many myths and legends of East Asia Kerbasia, a name I thought apt considering its tricycle landing gear and the ease at which it cloaks itself in fiery plasma.

This airframe is the product of several design phases, and a lineage traceable to the 0.18 release of KSP when I created a hypersonic test vehicle. Since then it has undergone countless tweaks and flown to orbit dozens of times. During the process of converting the hypersonic test vehicle into an SSTO, some major design constraints became apparent. For example, the Sanzuwu was conceived from the outset to have a low wing to shield the delicate engine structures from re-entry heating, but this led to instability at high alpha. Only after rearranging the fuselage and adjusting the wing shape was this problem minimised. The engine pods themselves also required a great deal of attention in order to ensure the fuel flow worked correctly, and having them so far forward led to more stability issues.

After doggedly refusing to give up, I can now say with great confidence that the Sanzuwu is one of the most flyable SSTOs out there. It may not be big, it may not have a large crew, and it may not be interplanetary, but it is near perfect at what it was designed to do: take one Kerbal to MKO. For shear ease of use, for its reliability at reaching orbit, and for its cross range capabilities I feel that the Sanzuwu is unsurpassed.

Action Groups:

1 starts the turbojets and turns off the extra intakes (only use this when taking off)

2 toggles the extra intakes

3 toggles the turbojets and all intakes

4 toggles the LV-N atomic rocket

5 toggles the spotlight and the in-line docking port

6 toggles the boarding ladder

Flight Notes:

Take off:

1. Turn on ASAS, open the resources tab, and set the throttle to max

2. Press 1 to start the engines and pull up gently at 90m/s, pitching up to 45°. Bring the gear up

3. At 10km altitude pitch down to 15°

Ascent:

4. At 15km altitude pitch down to 5°

5. At 20km press 2 to open extra intakes

6. Proceed to 30km by gradually reducing thrust. Start doing so when the intake air drops below 0.18. Try to increase your surface velocity to over 1900m/s and your apoapsis to 35km. If you experience a flameout, immediately throttle down a few notches and reorient the craft

Orbital Insertion

7. When you're 20 seconds from apoapsis, press 3 to shut down the turbojets, press 4 to start the atomic rocket, and pitch up to at least 45°

8. Watch the map. As your apoapsis approaches 70km, pitch down so that you are pointing towards your prograde marker and throttle down. 'Tickle' the plane into space by using only a small amount of thrust to overcome drag loses

9. Begin burning again 20 seconds from apoapsis to circularise your orbit

Edited October 11, 2013 by Narcosis

[Cruzan]

Great job! That looks really cool, especially with the low wings. I just d/led it so I can see how it flies with wings that low

Does it need to refuel before returning back to Kerbin, or can it make it back by itself after reaching MKO? (not entirely what altitude range this covers )

[Elmetian]

Does it need to refuel before returning back to Kerbin, or can it make it back by itself after reaching MKO? (not entirely what altitude range this covers )

A good question

With a decent ascent to orbit and without refuelling it has enough delta v for a Mun capture and partial circularisation of its orbit. I just tested this and got a periapsis of 35km and an apoapsis of 185km around the Mun. However, without refuelling it can only achieve a circular orbit around Kerbin of about 2300km, which is significantly higher than your average LKO, but also much lower than a geosynchronous orbit (and so not HKO).

EDIT: without refuelling, the highest circular orbit around Kerbin with enough fuel to return is about 1700km

Edited October 9, 2013 by Narcosis

[Cruzan]

The Sanwuzu works great! I only tested it in atmosphere, flying around between bases and stuff, but takeoff, landing, climbing are all smooth. I flipped it a couple of times because I'm used to small, twitchy planes so I'll chalk those up to pilot error

[Elmetian]

The Sanwuzu works great! I only tested it in atmosphere, flying around between bases and stuff, but takeoff, landing, climbing are all smooth. I flipped it a couple of times because I'm used to small, twitchy planes so I'll chalk those up to pilot error

Thanks for the feedback. She's a bit heavy on take off and will flip easily if you pull back too hard. After returning from orbit you'll find that you can pull tighter turns without it flipping, but I think the low monowing profile will never be as stable as the biplanes that make up the majority of SSTOs on the forums. Call it an aesthetic choice, but I wanted it to look a bit more realistic. Maybe I'll add a caveat into the description

[TheWanderer]

Have you tried none nuclear engines for the push up to orbit or does it not have enough fuel for that? I am always loath to have in atmosphere NERVAs

[Elmetian]

Have you tried none nuclear engines for the push up to orbit or does it not have enough fuel for that? I am always loath to have in atmosphere NERVAs

I have, and they're a lot less efficient. The first iterations of the Sanzuwu had an LV-T45. It got me into orbit, but there wasn't much delta-v remaining. I then shoved an ion engine and a single Gigantor solar array on it for orbital manoeuvres but the acceleration was too low to make it worthwhile, not to mention the extra mass made it harder to reach orbit in the first place.

I guess if I spammed intakes to conserve fuel it would make a T45 more practical, but I think the atomic rocket is still a better bet. I really don't like abusing intakes either. The Sanzuwu has 2 per engine, and that's the bare minimum for making it work as an SSTO. Without airhogging you'd need a much more powerful rocket and lots more fuel.

[TheWanderer]

I would much rather use NERVAs than airhog so its all good :-)

[Rune]

I'd bet it would have much more performance if you halved the number of jet engines, and if put behind a decoupler, you would get the same part count and wouldn't loose the aesthetics... but you would loose more than a couple of tons as dead weight to orbit. Other than that, it looks good, and yes, a plane with a low wing tends to be unstable. You could move the wing up (or fly upside down), or you could also give it a dihedral shape. You know, like a V, my english is not that good. Anyhow, that also increases bank stability by moving the center of lift upwards. Just remember taht your center of mass will want to hang from the center of lift as a general rule.

Rune. My two cents, hope you don't mind.

[Elmetian]

Rune. My two cents, hope you don't mind.

Not at all. I've come to understand how the aerodynamics (or at least KSP's rudimentary simulation of them) affect the plane, and it was a challenge to try and overcome them. I could make it more stable by lowering the centre of mass by adding some ballast in the form of liquid fuel or monoprop tanks, but this would lead to unwanted torque as the direction of thrust no longer passes through the CoM.

For a stable aircraft in KSP you either need a high wing, a mid wing that the engines either clip through or sit behind, or a biplane. None of these options appealed to me as they're not very realistic looking. Most real spaceplanes have a low wing to avoid unwanted localised surface heating on re-entry, and they probably avoid instability by concentrating mass at the bottom of the fuselage like an airliner. I tried a dihedral wing but it ended up clipping through the engines or being very wide, and vanilla KSP can't do pitch and roll properly with canted control surfaces

Flying a low wing aircraft upside down for the ascent to orbit is genius though! Might have to try building a craft specifically for that flight profile.

As for the number of engines, what can I say, I like the extra power. Most KSP SSTOs fly for several minutes and hundreds of kilometres before reaching a high altitude and speed. The Sanzuwu has just taken off in the picture above. That's a little over 40km from the space centre

Edited October 9, 2013 by Narcosis

[Cruzan]

Ya I forgot to mention that I liked the way it looked when I flew it upside down as well . That would be a pretty neat idea to have a plane that ascends with a high wing, then re-enters with the low wing! Hmm....ideas....to the SPH!

[Elmetian]

Ya I forgot to mention that I liked the way it looked when I flew it upside down as well

Well, it did originally have a high wing. This is the prototype hypersonic test vehicle that eventually became the Sanzuwu:

Javascript is disabled. View full album

[Rune]

Not at all. I've come to understand how the aerodynamics (or at least KSP's rudimentary simulation of them) affect the plane, and it was a challenge to try and overcome them. I could make it more stable by lowering the centre of mass by adding some ballast in the form of liquid fuel or monoprop tanks, but this would lead to unwanted torque as the direction of thrust no longer passes through the CoM.

For a stable aircraft in KSP you either need a high wing, a mid wing that the engines either clip through or sit behind, or a biplane. None of these options appealed to me as they're not very realistic looking. Most real spaceplanes have a low wing to avoid unwanted localised surface heating on re-entry, and they probably avoid instability by concentrating mass at the bottom of the fuselage like an airliner. I tried a dihedral wing but it ended up clipping through the engines or being very wide, and vanilla KSP can't do pitch and roll properly with canted control surfaces

Flying a low wing aircraft upside down for the ascent to orbit is genius though! Might have to try building a craft specifically for that flight profile.

Well, as my father says, "ni quinientas palabras más", you seem to have your aerodynamics right on the money. And in the end, you went the kerbal way and picked the crazy solution. Good!

As for the number of engines, what can I say, I like the extra power. Most KSP SSTOs fly for several minutes and hundreds of kilometres before reaching a high altitude and speed. The Sanzuwu has just taken off in the picture above. That's a little over 40km from the space centre

Yeah, again if it is a personal preference, I have nothing to argue. But just for the sake of argument, consider my White Dart (I won't invade your thread by linking, you can look it up through my sig): twice the rocket fuel, half the aviation fuel, and half the jet engines (in a stabojet configuration, so I waste (1-cos15º)*thrust of available "umph"). It still has T/W over 1, so you could fly it like a rocket. It does not waste fuel pushing against terminal velocity, though, and you need quite the steep dive to get mach effects, but it is a very efficient SSTO with awesome Delta-v on the nuke-equipped version. It still makes very high orbit with a LVT-30, and then it has T/W over 1 also in rocket mode, which I can guarantee you ensures a faster climb to orbit, if you want to be inefficient about it. Just in case you want something to measure proportions against . This baby still has plenty of room to grow!

Rune. I mean, that is what I use the spacecraft repository for, improving my builds, so where the White Dart came to existence from. Thanks guys!

Edited October 10, 2013 by Rune

[Elmetian]

But just for the sake of argument, consider my White Dart (I won't invade your thread by linking, you can look it up through my sig): twice the rocket fuel, half the aviation fuel, and half the jet engines (in a stabojet configuration, so I waste (1-cos15º)*thrust of available "umph"). It still has T/W over 1, so you could fly it like a rocket. It does not waste fuel pushing against terminal velocity, though, and you need quite the steep dive to get mach effects, but it is a very efficient SSTO with awesome Delta-v on the nuke-equipped version. It still makes very high orbit with a LVT-30, and then it has T/W over 1 also in rocket mode, which I can guarantee you ensures a faster climb to orbit, if you want to be inefficient about it. Just in case you want something to measure proportions against . This baby still has plenty of room to grow!

The White Dart is a very good looking SSTO I spotted it a few weeks ago when you posted pics on the SSTO showcase thread.

It's certainly very different to mine, but I wouldn't really call the Sanzuwu a 'crazy solution'. I chose not to airhog (much) so it became necessary to have more turbojets in order to approach orbital velocity. While the White Dart appears to have 5 intakes per engine (from what I can see), mine only has 2 intakes per engine so it loses power much lower in the atmosphere. I did try just 1 intake for each but I couldn't get past about 1600m/s which meant the atomic rocket didn't have enough thrust to overcome gravity.

Edited October 10, 2013 by Narcosis

[Rune]

The White Dart is a very good looking SSTO I spotted it a few weeks ago when you posted pics on the SSTO showcase thread.

It's certainly very different to mine, but I wouldn't really call the Sanzuwu a 'crazy solution'. I chose not to airhog (much) so it became necessary to have more turbojets in order to approach orbital velocity. While the White Dart appears to have 5 intakes per engine (from what I can see), mine only has 2 intakes per engine so it loses power much lower in the atmosphere. I did try just 1 intake for each but I couldn't get past about 1600m/s which meant the atomic rocket didn't have enough thrust to overcome gravity.

Oh, the crazy part was the comment about flying it upside down for ascent. You will admit, that sounds like a very kerbal solution. And yeah, two different ships, yours and mine. Mine is a clipped-to-hell airhoggerdesigned for absolute performance that just happened to look way better than it should. But considering yours is lighter, if you kept the intakes as they are but placed a bicoupler behind them, you get the altitude to go very fast (you would have to throttle down anyhow with four engines), the same aesthetics, and half the dead engine weight in orbit. The overall vehicle will be lighter and therefore a better glider, too. That was my humble suggestion, the rest were just arguments.

Rune. Another suggestion: in a three-engine vehicle, you can make the middle engine flame out first as an indicator by placing it last whenbuilding.

[Cruzan]

Another suggestion: in a three-engine vehicle, you can make the middle engine flame out first as an indicator by placing it last whenbuilding.

I never knew that!

[KissSh0t]

The last engine place is the first to flame out always?

[Rune]

The last engine place is the first to flame out always?

Yup, and it's an awesome thing to know when doing SSTO's with an odd number of engines. Just remember, if you remove the others and place them back on, you have to do it all over again.

Rune. Last one in on the SPH, first one out when air gets thin.

[Cruzan]

Rune. Last one in on the SPH, first one out when air gets thin.

*Scumbag Jet Engine.*

[pa1983]

Not at all. I've come to understand how the aerodynamics (or at least KSP's rudimentary simulation of them) affect the plane, and it was a challenge to try and overcome them. I could make it more stable by lowering the centre of mass by adding some ballast in the form of liquid fuel or monoprop tanks, but this would lead to unwanted torque as the direction of thrust no longer passes through the CoM.

For a stable aircraft in KSP you either need a high wing, a mid wing that the engines either clip through or sit behind, or a biplane. None of these options appealed to me as they're not very realistic looking. Most real spaceplanes have a low wing to avoid unwanted localised surface heating on re-entry, and they probably avoid instability by concentrating mass at the bottom of the fuselage like an airliner. I tried a dihedral wing but it ended up clipping through the engines or being very wide, and vanilla KSP can't do pitch and roll properly with canted control surfaces

Flying a low wing aircraft upside down for the ascent to orbit is genius though! Might have to try building a craft specifically for that flight profile.

As for the number of engines, what can I say, I like the extra power. Most KSP SSTOs fly for several minutes and hundreds of kilometres before reaching a high altitude and speed. The Sanzuwu has just taken off in the picture above. That's a little over 40km from the space centre

Low profile wings are not a problem nor unstable. Only real problem I had was engine placment and structrual integrity but once I solved that on my crafts it wirks just fine. My lates craft can fly up side down but rolling from level fligt it usually stopps its roll at a 90 degree bank. Its perfectly posible to fly upside down but it seems to prefer rigth side up.

[Elmetian]

Low profile wings are not a problem nor unstable. Only real problem I had was engine placment and structrual integrity but once I solved that on my crafts it wirks just fine. My lates craft can fly up side down but rolling from level fligt it usually stopps its roll at a 90 degree bank. Its perfectly posible to fly upside down but it seems to prefer rigth side up.

Because I wasn't sure, I thought I'd go test this with a purpose-built plane. Behold, the LowHigh:

The only difference between these two craft is that the body was rotated exactly 180°. The wing was placed before placing the landing gear (because as we know, gear is massless and does not induce drag outside of the SPH) so that the direction of lift passes through the CoM ball but slightly infront.

I allowed both aircraft to get airborne before trying any manoeuvres. There was a difference of about 20m/s between taking flight as the gear is placed slightly differently. I waited until they were both doing 130m/s in level flight and then kept my finger on the S key to perform a hard pitch up without fine controls on. The high wing initially pitched up fast but then the rate of turn slowed as the direction of travel lagged behind where the nose was pointed. At no point did it flip.

I tried exactly the same thing in the low wing. After getting past about 40° it flipped over uncontrollably.

This problem is clearly caused by not having a separate stabiliser. With the addition of this T-tail, the low wing test aircraft became controllable and was able to pull much tighter turns than the high wing (with or without the tail). Note also that the engines had to be moved to bring to CoM further back as I didn't want to have to move the wing section.

Lastly, I tried a low stabiliser placed on a length of massless cubic struts. I also removed the front elevators that can be seen embedded in the wing in the first two pictures in order to bring the direction of lift further back so that I could compare it to the previous T-Tail version without moving the main wing. The aircraft performed well, and although it didn't turn as fast as the low wing with high T-tail, it did not flip.

In conclusion, I was correct that low wings are inherently unstable, but clearly this is only true when they don't have a separate stabiliser. Originally I thought that having a pair of elevators on the leading edge of the wing as well as the trailing edge would act the same as having trailing edge elevators and a separate pair of stabilisers, but they obviously don't. This then is what's causing some instability at high rates of turn in the fully fuelled Sanzuwu. I don't think it's much of a problem though as long as you don't pull any fast turns after taking off

Edited October 11, 2013 by Narcosis

[Michael Kim]

Come on...

Those three Chinese characters are pronounced San Zu Niao instead of San Zu Wu

[Elmetian]

Come on...

Those three Chinese characters are pronounced San Zu Niao instead of San Zu Wu

Are they? Oops

That's what you get for trusting Wikipedia I guess, though I did look elsewhere just in case. Should it be like this by any chance (三足乌), or is it just a badly Anglicised pronunciation of the correct characters?

EDIT: Michael, are you sure that you're not getting these two characters mixed up: 烠and 鳥? Don't get me wrong, I know sweet FA about Chinese other than what I've read online, so it wouldn't surprise me if I'm wrong about this.

Edited October 11, 2013 by Narcosis

[Rune]

Because I wasn't sure, I thought I'd go test this with a purpose-built plane. Behold, the LowHigh:

The only difference between these two craft is that the body was rotated exactly 180°. The wing was placed before placing the landing gear (because as we know, gear is massless and does not induce drag outside of the SPH) so that the direction of lift passes through the CoM ball but slightly infront.

I allowed both aircraft to get airborne before trying any manoeuvres. There was a difference of about 20m/s between taking flight as the gear is placed slightly differently. I waited until they were both doing 130m/s in level flight and then kept my finger on the S key to perform a hard pitch up without fine controls on. The high wing initially pitched up fast but then the rate of turn slowed as the direction of travel lagged behind where the nose was pointed. At no point did it flip.

I tried exactly the same thing in the low wing. After getting past about 40° it flipped over uncontrollably.

This problem is clearly caused by not having a separate stabiliser. With the addition of this T-tail, the low wing test aircraft became controllable and was able to pull much tighter turns than the high wing (with or without the tail). Note also that the engines had to be moved to bring to CoM further back as I didn't want to have to move the wing section. I also removed the front elevators that can be seen embedded in the wing in the first two pictures in order to bring the direction of lift further back.

Lastly, I tried a low stabiliser placed on a length of massless cubic struts. The aircraft performed well, and although it didn't turn as fast as the low wing with high T-tail, it did not flip.

In conclusion, I was correct that low wings are inherently unstable, but clearly this is only true when they don't have a separate stabiliser. Originally I thought that having a pair of elevators on the leading edge of the wing as well as the trailing edge would act the same as having trailing edge elevators and a separate pair of stabilisers, but they obviously don't. This then is what's causing some instability at high rates of turn in the fully fuelled Sanzuwu. I don't think it's much of a problem though as long as you don't pull any fast turns after taking off

You just did quite the aerodynamic study there, good for you! Scientific method FTW. I learned all this in university, so I'm glad KSP weird physics get to the same place. Pa's opinion is probably based on his huge SSTO's flexing to become dihedral wings, and all of them having a big stabilizing tail. Not a coincidence the tail contains what are called horizontal and vertical stabilizers in RL.

Oh, and it's interesting that you got more maneuverability by keeping the unstable wing and stabilizing it with a tail: now you also know first hand why jet fighters nowadays have inherently unstable configurations, kept in check with modern computers.

Rune. Not counting control surfaces, KSP's aerodynamic model gets way more hate than it should. Well, there's the drag issues, those make me cringe a little, but the rest works surprisingly realistic.

Edited October 11, 2013 by Rune

[Elmetian]

You just did quite the aerodynamic study there, good for you! Scientific method FTW.

Heh, well, sort of. My methodology was pretty rubbish because I ended up having to move the engines. I did this because I needed to have the tail on a separate section so it could be rotated independently of the main body, and that required it to be further back. I guess the conclusion is generally correct though. Not bad for a 10 minute design study I guess

Not counting control surfaces, KSP's aerodynamic model gets way more hate than it should. Well, there's the drag issues, those make me cringe a little, but the rest works surprisingly realistic.

True. It definitely does get way more criticism than it deserves, especially considering by Squad's own reckoning we're only a 1/5th of the way to the full release of KSP. That being said, I am looking forward to them overhauling the aerodynamic model, whenever it comes

Edited October 11, 2013 by Narcosis