An idea to deal with in-flight shifting CoM in spaceplanes

[Cirocco]

Right so I'm a pretty big spaceplane nut, I think spacecraft look cooler and more futuristic if they have wings

One thing that has been bugging me for the entire time I've been playing KSP though is the shifting center of mass (CoM) that causes such problems in (heavy) spaceplanes: fully fueled means CoM forward and nose-dive tendencies, empty tanks mean CoM far to the back and somersault tendencies. I used to always try to build around this problem by trying the minimize the drift in the CoM due to fuel consumption, but this lead to designs with either large fuel tanks or engines very far to the front and just looked plain ugly in my eyes.

And then I saw a post on the forums by Janos1986 and I had an idea: angling your jet engines upwards a few degrees with the gizmos gives you a lot of torque which can offset the nose-diving and troublesome take-offs that heavy SSTO's with a lot of engines in the back suffer from, but they would also make your plane flip out instantly if it comes back into the atmosphere at the end of a mission with near-empty tanks and a CoM that lies super far to the back.

But what if you angle all of your jet engines upwards except 2? You leave those 2 horizontally aligned or you might even angle them slightly downwards. Then, when entering the atmosphere with near-empty tanks and a tendency to flip nose-up instead of nose-down, you use those non-angled or slightly downwardly angled engines to compensate. You wouldn't need a lot of thrust anyway because the majority of your mass is gone and hey-presto, stable aircraft at all times, even with extreme CoM shift!

Another option is to slide your engines slightly up- or downward, but in my designs this would be somewhat difficult (though with this new info in mind, my designs may change radically ).

So what do you guys think? Would this work? Do you already do it? Do you have other ways of compensating for an in-flight shifting CoM other then designing around it so it doesn't move much?

Discuss!

This came up in a thread under the "gameplay questions and tutorials" section, but I didn't want to hijack the thread so I figured I'd make a new threa about this idea here. Big thanks to Tassyr for creating the original thread, Janos 1986 for giving me the original idea and Wanderfound for kicking my imagination in the butt so it took off

Edited January 13, 2015 by Cirocco

[DocMoriarty]

Space planes are ALWAYS designed with the center of the tank setup being same as CoM so that flying out fuel doesn't move CoM. The still not yet updated documentation about creating space planes in my signature contains lots of examples how to create perfectly balanced planes up to extremely big ones. Doesn't use the mostly superflous Mk2/Mk3 parts.

[Wjolcz]

Or you can just pump the fuel around. What I do is fiddle around with tweakables for fuel tanks while still in the SPH. And watch how the CoM behaves with full/half-empty/empty tanks. If its in front of CoL then its fine. The best way is to have a tank in the front and on the back, so you can balance the craft in-flight.

Im terrible at explaining things, but hopefully you get the point.

Edited January 13, 2015 by Veeltch

[Janos1986]

AAAAAH TOO QUICK! I replied to you in the other thread!

If you balanced your tanks well, there's no need to overthink your engine placement.

http://forum.kerbalspaceprogram.com/threads/106963-Many-questions-about-SSTO-s-Shuttles-and-spaceplanes-in-general/page3

Here.

[InterCity]

As a matter of fact, during reentry, it's usually good to have your plane a little lawn dart-ish (but not so much that you can't use your control surfaces to deal with it. Too much lift on the back and the plane will spin wildly (under FAR). For takeoffs, fuel pumps are your best friend as mentioned before, and of course the said fuel tanks placed with their bulk around the CoL.

If squad added in-wing fuel tanks, that would make things a lot simpler.

[Wanderfound]

I do a fair bit of fine-tuning with engine angle and position, but I'm typically aiming at minimising thrust torque in vacuum so that I don't waste RCS or need to carry excess SAS units to maintain heading during interplanetary burns. Nothing as annoying as taking your spaceplane to orbit and then finding that it likes to do loops whenever you open the throttle.

Due to the ability of aerodynamic control surfaces to compensate, you can get away with quite a lot of thrust asymmetry while flying in atmosphere. For example:

...flies fine on either engine or both.

As suggested above: if your CoM is shifting significantly in flight, you'd probably be better off getting that facet of the design sorted out directly. Spread the fuel load laterally and centre it around CoM, use fuel lines to control the order in which the fuel drains. There are times when this isn't possible due to practical or aesthetic reasons; this is where temporarily locking off tanks with tweakables comes into play. Even a purely linear fuselage plane can maintain balance if you build it right.

If you can, keep the CoM/dCoM offset at less than one metre. It isn't absolutely necessary, but RCS Build Aid does make this task a lot simpler.

Edited January 13, 2015 by Wanderfound

[Janos1986]

Well uhhhh, this last aircraft is quite... unconventional, indeed.

[Cirocco]

Space planes are ALWAYS designed with the center of the tank setup being same as CoM so that flying out fuel doesn't move CoM. The still not yet updated documentation about creating space planes in my signature contains lots of examples how to create perfectly balanced planes up to extremely big ones. Doesn't use the mostly superflous Mk2/Mk3 parts.

I'm not arguing that you *can't* create spaceplanes that are perfectly balanced, my point is that it's very hard to make aesthetically pleasing stock ones that don't use a lot of clipping or some sort of exploit. In order to keep the CoM from shifting in big and heavy spaceplanes (I'm talking like 200 tons and heavier), you need to either put a lot of tanks in front of the plane, smaller tanks faaaaar in the front, or a lot of dry weight in the front. None of those options ar every appealing to me from an aesthetic viewpoint.

Angling the thrust is a simple method that also makes sense from a physics point of view. Also I do like the new Mk III spaceplane parts so I would prefer to use them

Or you can just pump the fuel around. What I do is fiddle around with tweakables for fuel tanks while still in the SPH. And watch how the CoM behaves with full/half-empty/empty tanks. If its in front of CoL then its fine. The best way is to have a tank in the front and on the back, so you can balance the craft in-flight.

Im terrible at explaining things, but hopefully you get the point.

I get what you're saying, but I personally have an irrational dislike for manually pumping fuel or fiddling with stuff in-flight. I want my designs to have a clear

way of dealing with every (foreseen) problem without having to improvise or solve it on-the-spot.

As a matter of fact, during reentry, it's usually good to have your plane a little lawn dart-ish (but not so much that you can't use your control surfaces to deal with it. Too much lift on the back and the plane will spin wildly (under FAR). For takeoffs, fuel pumps are your best friend as mentioned before, and of course the said fuel tanks placed with their bulk around the CoL.

If squad added in-wing fuel tanks, that would make things a lot simpler.

Well the problem is that my long, heavy Mk III spaceplanes with a ton of engines (including the heavy nuke engines) in the back and tanks more in the middle will often have a tendecy to lawn-dart upon take-off and in early flight, but flip nose-up upon re-entry. They have a CoM that lies near the back, but in order to try and compensate for the CoM shift, I place the CoL as far back as I can. That means CoM in front of CoL on take-off and behind it on re-entry.

These problems are much less of a problem when not using the Mk III parts, but I really want to use those. And they're so heavy that adding more fuel tanks to shift the CoM isn't an option because they would make the plane far too heavy and make it require far too many jet engines. There is of course the option of partially draining the MkIII part tanks, but flying an entire mission starting with only partially full fuel tanks seems horribly inefficient to me.

I'm perfectly aware that many of my problems are a result of my self-imposed restrictions on design, but if that means I get cool looking planes and a better understanding of design and physics because of them, I'm keeping those restrictions

Edited January 13, 2015 by Cirocco

[Wanderfound]

Well uhhhh, this last aircraft is quite... unconventional, indeed.

I was bored.

Not an original concept, though; stole it from the RAF...

[Levelord]

I do a fair bit of fine-tuning with engine angle and position, but I'm typically aiming at minimising thrust torque in vacuum so that I don't waste RCS or need to carry excess SAS units to maintain heading during interplanetary burns. Nothing as annoying as taking your spaceplane to orbit and then finding that it likes to do loops whenever you open the throttle.

Due to the ability of aerodynamic control surfaces to compensate, you can get away with quite a lot of thrust asymmetry while flying in atmosphere. For example:

http://i1378.photobucket.com/albums/ah120/craigmotbey/Kerbal/Challenges/Akademy%20Awards/Kerbodyne%20Scapa/screenshot194_zps3c788c8b.jpg

http://i1378.photobucket.com/albums/ah120/craigmotbey/Kerbal/Challenges/Akademy%20Awards/Kerbodyne%20Scapa/screenshot195_zps64444dd1.jpg

...flies fine on either engine or both.

As suggested above: if your CoM is shifting significantly in flight, you'd probably be better off getting that facet of the design sorted out directly. Spread the fuel load laterally and centre it around CoM, use fuel lines to control the order in which the fuel drains. There are times when this isn't possible due to practical or aesthetic reasons; this is where temporarily locking off tanks with tweakables comes into play. Even a purely linear fuselage plane can maintain balance if you build it right.

If you can, keep the CoM/dCoM offset at less than one metre. It isn't absolutely necessary, but RCS Build Aid does make this task a lot simpler.

Have you seen my craft that uses STABOJETâ„¢ technology?

[Cirocco]

Have you seen my craft that uses STABOJETâ„¢ technology?

http://i41.tinypic.com/34zhk6w.gif

http://i.imgur.com/9nUiP4M.jpg

what manner of sorcery is this?!

[GoSlash27]

I agree with all the others here;

A properly designed spaceplane won't have the CoM shift over the course of the flight.

I understand that you are correcting for a self-imposed condition for aesthetic purposes, but you should remember that you are "crutching" a poor design (from an engineering standpoint), so your results may vary depending on fuel state and flight regime.

KSP aerodynamics is weird.

Best,

-Slashy

[marach]

I wouldn't call it "'crutching' a poor design" more having to work with a broken mechanic. One thing KSP really needs for Spaceplanes is the ability to select the fuel pump location. I should be able to select how fuel flows in the tanks be it front/top to back/bottom, vice versa or balanced (though I must admit bottom to top would be unrealistic). Even if these options were only available in the SP hanger it would be a huge QoL improvement.

[cantab]

The vertically stacked engine design seems like a good idea - you can use thrust limiting to create torque when you need it. Unlike with angling you get full thrust any time you don't need torque.

And ultimately maybe the best option is to just use TAC Fuel Balancer or Goodspeed.

[Wanderfound]

The vertically stacked engine design seems like a good idea - you can use thrust limiting to create torque when you need it. Unlike with angling you get full thrust any time you don't need torque.

That's how I do Buran-ish vertical launch shuttle designs, BTW; one engine under the shuttle, one under the disposable tank, tweak thrust limiters in flight to maintain balance.

[InterCity]

Well the problem is that my long, heavy Mk III spaceplanes with a ton of engines (including the heavy nuke engines) in the back and tanks more in the middle will often have a tendecy to lawn-dart upon take-off and in early flight, but flip nose-up upon re-entry. They have a CoM that lies near the back, but in order to try and compensate for the CoM shift, I place the CoL as far back as I can. That means CoM in front of CoL on take-off and behind it on re-entry.

Okay. I think there is a solution for that: Place all cockpits and payloads as much to the tip of your plane as possible and fuel tanks around the center of lift. You can probably use structural fuselages to extend the nose of the plane. It will then gain momentum large enough to shift your CoM as much as desired. In my experience, however, it results in some rather eyesore looks.

Just now, when thinking about it, I also realized that putting the heavy nukes as close to your center of lift as possible (just like Boeing does ) is the solution to your problem.. No momentum for the engines, no problem.

A great help with this are either swept delta wings extending behind the plane or large stabilizers (the "rear wings"). The engines can then remain pretty much where they are just with some minor tweaks.

Another hint of mine may be avoiding the nukes. I know about their vacuum efficiency, but they're heavy as lead, big, ugly and they make Kreenpeace angry. I use Aerospikes, unless I really need to go interplanetary.

[Cirocco]

In my experience, however, it results in some rather eyesore looks.

yup, that's my biggest issue. I've considered extending the nose forward artificially, possibly through the use of cargo bays. Currently my designs don't incorporate cargo bays yet, sacrificing payload capacity for additional fuel storage, but it looks like I might incorporate them anyway. Extending the nose with structural elements is a no go for me as it's too much of an eyesore.

Just now, when thinking about it, I also realized that putting the heavy nukes as close to your center of lift as possible (just like Boeing does ) is the solution to your problem.. No momentum for the engines, no problem.

A great help with this are either swept delta wings extending behind the plane or large stabilizers (the "rear wings").

yup; that's the idea I'm currently thinking of as well: nukes more forward attached to very small fuel tanks and have em leech off the main tank. That should move the CoM forward a bit. I'll have to see whether the fact that it would then (probably) fire over the wing would damage the wing surface or impede the thrust in any way.

Another hint of mine may be avoiding the nukes. I know about their vacuum efficiency, but they're heavy as lead, big, ugly and they make Kreenpeace angry. I use Aerospikes, unless I really need to go interplanetary.

well that's rather the thing: I plan to go interplanetary with my spaceplanes. Building a cargo hauler to LKO is easy. Building a spaceplane that goes to Duna, lands horizontally, does sciency stuff and comes back to the runway all without refueling? THAT's a challenge.

Mind you a single spaceplane that does that would, while cool, be able to do little else other than that specific mission profile (maybe a laythe mission. Maybe.) so I'm considering letting go of that idea.

Edited January 13, 2015 by Cirocco

[AlexMBrennan]

The vertically stacked engine design seems like a good idea - you can use thrust limiting to create torque when you need it. Unlike with angling you get full thrust any time you don't need torque.

Assuming you are referring to the stabojet: no, you really don't. By angling the engines you gain the ability to generate torque by throttling one of them down, but the price you pay is that the component of thrust normal to the vertical axis is wasted (e.g. with two way symmetry, part of the left engine's thrust is used to turn the craft anticlockwise, and part of the right engine's thrust to counter this rotation... so you get less lift whilst burning the same amount of fuel...)

[Johnny Wishbone]

I agree with all the others here;

A properly designed spaceplane won't have the CoM shift over the course of the flight.

I understand that you are correcting for a self-imposed condition for aesthetic purposes, but you should remember that you are "crutching" a poor design (from an engineering standpoint), so your results may vary depending on fuel state and flight regime.

KSP aerodynamics is weird.

Best,

-Slashy

Have you read "Sled Driver"? If not, I recommend it. Very interesting read.

The SR-71 is generally considered a wonder of aircraft design and engineering and seems to be a favorite of KSP re-creators. In "Sled Driver", the author (a former Blackbird pilot) states that he spent most of the time during the flight pumping fuel around the airframe to deal with the constant shifting CoM and also adjusting the position of the intake cones. The aircraft's autopilot did all the flying; the pilot's job was to keep the thing from falling out of the sky.

[cantab]

Assuming you are referring to the stabojet

I was referring to Wanderfound's design inspired by the Lightning.

[Starhawk]

...well that's rather the thing: I plan to go interplanetary with my spaceplanes. Building a cargo hauler to LKO is easy. Building a spaceplane that goes to Duna, lands horizontally, does sciency stuff and comes back to the runway all without refueling? THAT's a challenge.

Mind you a single spaceplane that does that would, while cool, be able to do little else other than that specific mission profile (maybe a laythe mission. Maybe.) so I'm considering letting go of that idea.

This has been my experience. I have a long-range spaceplane that can do Mun, Minmus, and Ike, but there isn't enough atmo to give it enough lift on Duna. I ended up designing quite a different spaceplane to do the Duna-return mission with horizontal landing and takeoff at Duna, definitely a specific-mission craft.

I never tried it at Laythe, but I see no reason it wouldn't work. I'd need to build a pusher stage to get it to Jool, though, or send a refueler.

[InterCity]

Okay, I've been tinkering about it, and came up with this:

We were talking about big spaceplanes, so I made one. However, it turned out that even with TWR of 1.35 the engines were boosting it just up to 90 m/s, while the lowest takeoff speed was about 115 m/s (I blame that on stock aerodynamics). Adding more engines didn't help soooo...

I know that it's no longer a SSTO, but the boosters gave it enough punch to get the jets running. And since everything went as intended...

...several flight reverts, two coffee cups and a buttload of profanity later I was finally at LKO. It was a long journey through low-fps-land, the old linux-powered Dell XPS i was running it on (travel computer, has some nice battery) didn't seem to like it very much. But it was pretty stable. It will need some tweaks to get anywhere else than LKO (Namely that mainsail), but you get the point. The cargo bay is empty to give momentum and most of the jets are just behind the CoL. Most fuel tanks are ahead, so during reentry it makes for a nice lawn dart.

Such a feat would be WAY easier with Mk2 parts. I swear. The plane was a royal pain to design.

Edited January 13, 2015 by InterCity

[Cirocco]

Okay, I've been tinkering about it, and came up with this:

http://i.imgur.com/5zjAhX1.png

We were talking about big spaceplanes, so I made one.

Nice! this is exactly the sort of scale I was talking about. I was messing around with spaceplanes more myself yesterday and came up with a very similar design: use the cargo bay to push to cockpit (and thereby also the dry CoM) forward. I just need to figure out what to put into the cargo bay now. Rovers, science equipment and sattelites are good contenders. As for the TWR: I personally use 12 jet engines on a lighter design and 4 nuke engines near the CoM. Two on each side, one above and one under the wing. The jets provide just enough TWR to lift off and I'm hoping 4 nukes will be just enough thrust to circularize. If it isn't, I'll replace a couple turbojets with RAPIERS and use the rocket mode to give it a bit of a boost.

http://i.imgur.com/2RUs09N.png

I know that it's no longer a SSTO, but the boosters gave it enough punch to get the jets running. And since everything went as intended...

Aaaaaaaah, good old RATO. Love it to bits and a great solution for heavy planes that have trouble taking off. Might just resort to it myself. Quik note: even though it won't be an SSTO anymore, you *can* make it a 100% recoverable by adding chutes to the SRB's and setting them to deploy at minimum altitude (50m). That way they should splash down in to the ocean before they leave the 2,5 km physics bubble.

Such a feat would be WAY easier with Mk2 parts. I swear. The plane was a royal pain to design.

I know, right? MkIII parts which are designed to look like heavy-duty parts are, ironically, the hardest ones to use when designing heavy-duty SSTO spaceplanes.

Edited January 14, 2015 by Cirocco

[Wanderfound]

This has been my experience. I have a long-range spaceplane that can do Mun, Minmus, and Ike, but there isn't enough atmo to give it enough lift on Duna. I ended up designing quite a different spaceplane to do the Duna-return mission with horizontal landing and takeoff at Duna, definitely a specific-mission craft.

I never tried it at Laythe, but I see no reason it wouldn't work. I'd need to build a pusher stage to get it to Jool, though, or send a refueler.

Weird. I found some things challenging with flying Duna (slowing down takes forever due to the low friction, and bumpy landing strips + low-G lack of traction make for tricky landings), but lack of lift absolutely wasn't a factor. Sea-level Duna atmosphere is no thinner than moderate-altitude Kerbin atmosphere.

Might be another stock vs FAR thing, though.

(see http://forum.kerbalspaceprogram.com/threads/101944-Kerboduna-Part-1 for the mission report if curious)

Edited January 14, 2015 by Wanderfound

[Cirocco]

Might be another stock vs FAR thing, though.

Sounds like it. I run stock and when I did my Duna flight lift was a giant pain in the butt, had to more than double my initial wing surface. I think it has to do with the way FAR calculates lift. A wide plane will have more lift in FAR, even with small wing surfaces. In stock, only the wing surfaces count so I had to cover the entire plane in wing surface to essentially create a flying delta wing.