Need help taking off my heavy cargo spaceplane

[Fearless Son]

Salutations Kerbalnauts,

Long time player, first time poster. Since the .90 update, I have been struggling to make an effective heavy cargo spaceplane for delivering large components to orbit in a cost-effective manner. These components are intended to be attached together into more complex structures, such as space stations or larger interplanetary vessels. I have certainly proven I can build effective spaceplanes that can reach low and mid-Kerbin orbits (almost but not quite as far as Mun in one of my models) but only for the lightest of spaceplanes. Building ones that are as heavy as I need this to be has proven much more challenging. Not the least of which is simply getting off the ground.

So far, every design I have attempted (and this is only the latest of many) has trouble even getting off the runway. I try to build them with large amounts of wing surface area to give them a lot of lift, but no amount of lift ever seems enough, the speed caps out around 70 M/s by the time I hit the end of the runway on only its turbojets, and only a little more than that if I burn the rockets early to give it just that little extra thrust. Even after getting off the runway, I tend to sink just slightly faster than the ground falls away beneath me toward the shore.

Let me illustrate with a few pictures:

Above is the side view of my current design. I have included the center of mass, center of lift, and center of thrust indicators for reference.

This view of it from above underscores just how much wing I try to put on this thing, though it seems to never be enough.

View from below this time, with the cargo bay doors open so you can see the "ballast" I put in here to test the center of mass with a load onboard. My goal is to be able to transport an entire orange fuel container into orbit. It is the volume and mass of it that is more important to my design goals than the amount of fuel itself, since this is not necessarily a tanker, just something that can convey a similar size and mass to this.

View from the front, shielded docking port for utility's sake. You can also see that there are sixteen conical air intakes on this thing to ensure plenty of airflow for the engines.

And of course the reason for all those intakes being the six radially-mounted turbojet engines on the back, though I fear that may not be enough. Nestled between them is a quad of aerospike engines for atmospheric escape and orbital maneuvering.

Detail on the intakes. I angled the ones hanging under the wing slightly downward, because I know a spaceplane of this mass flying at high altitude will likely have some serious incline and I want to ensure my air intake remains relatively efficient during that stage of the flight.

And finally, detail on the landing gear. Note the two pairs of landing gear positioned just behind the center of mass. Both are equally level and adjacent to spread the weight of the plane across them so that they do not end up breaking me just by the weight of the plane pushing down. There is also a single landing gear nestled among the engines which actually hangs off the ground while on the airstrip, but the wheel extends just slightly past the radius of the engine cowling to ensure it does not scrape across the ground while taking off or landing.

Any advice on how I can get a beast like this actually in the air? I mean, I know smaller and simpler is better, but in this case the size is part of the point. I have seen pictures from other people who have constructed absurdly massive spaceplanes that against all common sense make it to other worlds (very Kerbal) but I would be content just to get this into equatorial orbit and back down again.

Edited January 14, 2015 by Fearless Son

[TokiTech]

Hello new forumer. I actually use 4 Turbojets and one RAPIER to get my 24t plane to LKO and back easily. Is a sweet plane you got there but sad to say with current Stock game those Turbojets performance are quite limited on big planes. Now i haven't got a TWR calculation, but by looking i would put 10 more Turbojets on the wings, put it like you did with the intakes and switch Aerospikes to RAPIER which is heavier, but will help some thrust on take off then being dead weight. And if it does takes off it will be trials and error for additional tweaks, as i suspect it will need plenty more shock cone intakes.

Hopefully this SSTO plane design nightmare will be over with the new aerodynamic model, which Squad hinted will be in the next KSP update.

[Anquietas314]

i would put 10 more Turbojets on the wings

Uh, yeah that's not going to be enough with that much fuel and a full orange tank as payload. I would go with double that. You're also going to want a lot more intakes. A lot more.

Alternatively, you could just use less fuel. You don't need quite that much to get to orbit - about half the rocket fuel should be sufficient if you can get it up to ~1600m/s using jets.

[Van Disaster]

Caveat - I have been using FAR for so long I cant remember stock areo, but I would say that is probably not really enough wing area - the fuselage looks awfully heavy - and it most definitely isn't enough thrust. As the post above said, try removing some weight before you start adding more again - it's not being smaller that is better, it's being minimalist.

[Anquietas314]

Caveat - I have been using FAR for so long I cant remember stock areo, but I would say that is probably not really enough wing area

Nah that's probably fine, at least in stock. It might be worth making the wings wider though if you're going to drop some of the fuel (= less surface to stick wings on).

[Wooks]

I find your lack of turbojet engines disturbing, you definitely need moar, no ammount of wing surface will lift that weight without some serious thrust.

Have you considered moving your cargo bay to the middle of the ship, placing it between fuel tanks? It may help balancing the ship on the return trip.

[Anquietas314]

Have you considered moving your cargo bay to the middle of the ship, placing it between fuel tanks? It may help balancing the ship on the return trip.

It's worth considering that on the return trip you would probably expect the cargo bay to be empty, along with the fuel tanks, so that might not be a serious issue.

[Fearless Son]

I find your lack of turbojet engines disturbing, you definitely need moar, no ammount of wing surface will lift that weight without some serious thrust.

Okay, so after posting this last night I went back and did some hand calculations (I use exclusively stock so no mods to do it for me.) Comparing the total wing lift with the total mass tells me that I have less than half the wing I need to lift this much.

I think that I shall have to redesign the layout again, if only because adjusting the fuel supply in this Mk3 setup will alter the center of mass too much for my taste. Someone mentioned splitting fuel tanks across the cargo bay forward and back, but my concern is that it will also affect the center of mass too much by draining one side before the other. I think I am going to have to go with fuel tanks on the sides of the cargo bay so that adjusting both cargo payload and fuel weight should not unbalance the craft. Maybe put some Mk2 hulls astride the Mk3 hull.

I have experimented with "flying wing" type aircraft before, both as a way of cramming more functional components (such as air intakes in the front and turbojets in the back, I dislike most kinds of air hogging) into an aesthetically pleasing package and doubling-up my wing surface area to provide more lift without making a drastically bigger plane, but so far I have not noticed any gain in lift (and in fact most of my planes with that kind of design seem to have much less lift than the sum of the parts lift rating would suggest.) Can anyone elucidate me on the mechanics of making "fat" wings by carefully arranging structural wings together?

[HafCoJoe]

The problem I see in your plane is that the tail is too close to the ground. You can't take off because you can't tilt the nose up due to the landing gear in the back. If you watch a plane take off (Which I'm sure you have ), you can see how it only gets into the air after it tilts considerably up.

CORRECT SOLUTION: Put the rear landing gear directly below the center of mass, and put them far enough down on the plane that you can tilt the nose up without scraping the back.

COOLER SOLUTION:

(Means "Jet Assisted Take Off" but I really think it should be called "boostahs"). Really cool, relatively simple, works well, but sorta silly. I've used this for one of my larger planes which I wanted to land/take off in a short stretch of runway. Edited December 26, 2014 by Avera9eJoe

[Fearless Son]

The problem I see in your plane is that the tail is too close to the ground. You can't take off because you can't tilt the nose up due to the landing gear in the back. If you watch a plane take off (Which I'm sure you have ), you can see how it only gets into the air after it tilts considerably up.

In this case, I do not think that would be sufficient. A plane does not strictly need to be able to nose-up off the runway at Kerbal Space Center due to the runway being elevated. A plane can simply accelerate off the end of the runway, and if it has enough lift and thrust, can then perform a very gentle climb until it achieves sufficient clearance from the rapidly falling ground to point the nose up as far as it needs to for its ascent profile.

With my plane though, it fails to achieve enough lift to even maintain altitude once off the runway.

The more I think about it, the more I think I need to redesign the whole thing. I will post more pictures of my next iteration.

[Laie]

The more I think about it, the more I think I need to redesign the whole thing. I will post more pictures of my next iteration.

While you're at it, search this forum for asymmetric flameout. Once your plane reaches high altitudes, that will be the next bug to bite you.

[Fearless Son]

While you're at it, search this forum for asymmetric flameout. Once your plane reaches high altitudes, that will be the next bug to bite you.

Yeah, I am familiar with that one. I have gotten spaceplanes into orbit before, just never one quite this big.

I actually remembered a trick that worked once for me before and I am applying it to my new design. I use some standard jet engines balanced across the center of mass and positioned to thrust downward while in forward flight. The extra thrust vector directly against the gravity vector helps the excessive mass lift off the ground. Once I get to higher altitude and greater airspeed where the turbojets can shine, I can deactivate the increasingly ineffective lifting jets.

Just got a prototype of this into the air carrying appropriate cargo. Pics to follow.

[Fearless Son]

Okay, new design is ready for shakedown. Initial trials are promising, I can get it in the air under its own power and in a stable ascent profile. Full trials at transitioning from flight to low Kerbin orbit will follow. But for the moment, I have pictures of the new design to share:

She is too big to capture in the space plane hanger in one profile shot, so here she is in perspective. I had to find a lot of ways of extending the wings as much as I reasonably could while still looking aerodynamic, ultimately ending with them in this gentle curve in front and another curve in back to both pull back the center of lift and add more to it in total.

Another perspective shot from below. Note that I elected to go with the Mk2 cockpit instead of the Mk3 for weight reasons, and the adapter component I ultimately ended up using gave me lots of room beneath it to cram two pairs of high capacity intakes. Two other pairs of intakes are on each of the Mk2 fuel containers astride the Mk3 hull (though they are ironically mostly empty for total weight and balance reasons) with two more pairs near the ends of each wing. This makes twenty intakes in total with no air hogging. Testing will tell me if I need more than that.

The main engine clusters. Eight turbojet engines on each side of the hull, four more turbojet engines at the very rear of the hull, for a total of twelve turbojets. Yes, I realize this is slightly too many for the amount of intakes I have (I understand the ideal is two high capacity intakes for every one engine.) Radially mounted around the rear turbojets are four aerospike rockets.

The piece that makes this design workable at all is the extra standard jets balanced across the center of mass on the bottom of the plane. Alone, these give nowhere near enough vertical thrust for VTOL, but they do give just enough upward force to lighten the effective weight of the plane, allowing it to ascend under its own power from the runway, though I still prefer to run it straight off the end of the Kerbal Space Center runway before pulling up so as not to risk dragging the rear of the plane on the runway. Once it gets into a good ascent and is gaining speed, these jets can be deactivated to conserve fuel since they will not be needed or effective at higher altitudes.

Actually getting off the runway does still require a little more thrust than the turbojets alone provide, so I trigger a short full burn from the aerospike engines as I get to the end of the runway, giving it just that little extra thrust it needs to get airborne and nose up. Once it is ascending, the rockets can be cut and let the turbojets do the work.

I am thinking of naming this design the "Stellar Swan", mainly because the shape of the wings and position of the cockpit call to mind the form of a swan. The white wings and hull certainly do not hurt the image.

[Anquietas314]

The piece that makes this design workable at all is the extra standard jets balanced across the center of mass on the bottom of the plane. Alone, these give nowhere near enough vertical thrust for VTOL, but they do give just enough upward force to lighten the effective weight of the plane, allowing it to ascend under its own power from the runway, though I still prefer to run it straight off the end of the Kerbal Space Center runway before pulling up so as not to risk dragging the rear of the plane on the runway.

That's a pretty creative use for jets, I have to admit. Really though, the fact they're necessary suggests you'd be better off slapping more turbojets on with more intakes to match. That way you have more thrust and less dead weight at high altitudes

Actually getting off the runway does still require a little more thrust than the turbojets alone provide, so I trigger a short full burn from the aerospike engines as I get to the end of the runway, giving it just that little extra thrust it needs to get airborne and nose up. Once it is ascending, the rockets can be cut and let the turbojets do the work.

That has the lovely name of JATO/RATO (the latter technically makes more sense), but given how much more fuel-efficient jets are, you really want to avoid it if at all possible. If you can't get off the runway under jet power alone, then you probably won't have the thrust (never mind intakes) to get to orbit efficiently

[Van Disaster]

I've used ( droppable! ) SRBs for RATO on large craft before, saves blowing your jet fuel a little ( and looks awesome too... )

I would agree about the lift engines, you'd probably be better off using something disposable for one manoever in your entire flight profile, as it is you're lugging those lift engines all the way to space without using them when that mass could be cargo.

You might try larger fixed canards with bigger control surfaces to give you more pitch authority also.

Edited December 28, 2014 by Van Disaster

[Fearless Son]

Did a few modifications to the design. Removed the vertical-thrusting jets in favor of more axially-thrusting turbojets.

Added an additional four turbojets, two on each side, in addition to two more jet fuel tanks because I was worried I had a little too little for this many engines to chew through. This brings the total number of turbojets to sixteen.

Added a lot more shock cone intakes, bringing the total number to thirty-two intakes. This seems to be the ideal ratio of intake to engine from what I have experienced.

The total weighs in at 126.5 tons and 345 parts, which understandably takes a while to fly. But never the less, fly it does, this big baby gets off the ground under its own forward-thrusting jet power, if only just barely. Once clearing the runway, it still coasts while very slowly losing altitude, with the ground falling away toward the shore just slightly faster than the plane falls. I have to wait until I am out over the water before nosing up, lest the rear of the plane trail on the ground. But once that nose is up, the plane is off and climbing strongly.

The problem I am now running into is that the plane begins to struggle to stay on course when it gets going high and fast. Partially I think this is due to the high part count, resulting in a little "flex" in the structure that have a subtle effect on its aerodynamics (which I hope the upcoming update will allow me to fix by replacing lots of little parts with a few big parts.) This is controllable though, requiring me to babysit the yoke and make a few minor corrections. However, once I get to a few hundred meters per second and around fifteen kilometers off the deck, this becomes more pronounced, with the plane yawing left and right, overcompensating each time I try to correct, eventually losing meaningful control as it spins. I suspect this is due to the new fuel takes and flow. I can probably solve that problem by removing the new fuel takes, adjusting the fuel to fill up some existing empty tanks more toward the middle of the structure to make up the difference. I might also re-position the intakes to be more toward the back of the plane, resting on the trailing rear of the wing to shift the center of drag back and more toward the middle. That should make it a little more controllable.

Edited December 28, 2014 by Fearless Son

[GoSlash27]

However, once I get to a few hundred meters per second and around fifteen kilometers off the deck, this becomes more pronounced, with the plane yawing left and right, overcompensating each time I try to correct, eventually losing meaningful control as it spins.

This is a common instability mode. Usually it's the result of insufficient dihedral and empennage area.

Try bending your outer wing panels up a bit and add more vertical winglets as far back as you can place them. See if that helps you.

*edit* Also, those vertical canards out on the wingtips probably aren't doing you any favors.

Best,

-Slashy

Edited December 28, 2014 by GoSlash27

[TokiTech]

For preference i prefer 2 Turbojets per 10t for a decent ascend TWR. 16 Turbojets for 126t, i could feel the heaviness..

[Laie]

Actually getting off the runway does still require a little more thrust than the turbojets alone provide, so I trigger a short full burn from the aerospike engines as I get to the end of the runway, giving it just that little extra thrust it needs to get airborne and nose up. Once it is ascending, the rockets can be cut and let the turbojets do the work.

That has the lovely name of JATO/RATO (the latter technically makes more sense), but given how much more fuel-efficient jets are, you really want to avoid it if at all possible. If you can't get off the runway under jet power alone, then you probably won't have the thrust (never mind intakes) to get to orbit efficiently

Full ACK to the second part: if you have enough wing and thrust to reach really high speeds at high altitudes, you also have enough thrust and wing to take off easily. The angle of attack to maintain level flight at low altitudes should be no more than 10 degrees, better five. Early *sustained* climb rate should be no less than 30m/s (preferably 45m/s or over). If you're plane can't do that, taking it to high altitudes will not be worth the time and effort.

However, a short squirt from Aerospikes may still save you fuel, overall. Turbojets are extreme gas guzzlers at low altitudes. You want to get off the runway and to some 2km of altitude quickly (strictly speaking, their ISP won't draw even with basic jets until 5km. But 1500m is already about half the way). One benefit of spaceplanes (as opposed to jet-driven rockets) is that they can turn 100m/s of horizontal speed going down the runway into a 90m/s climb in a matter of seconds. That will quickly taper off and you have to pitch down again, of course, but by then you've already gained quite some altitude. Running the aerospikes for a *few* seconds to prolong that zoom climb, just a little, may be beneficial.

However, with a well-designed plane the gains will be miniscule. It may look cool and be convenient, each of which is a valid reason in and of itself; but if running your aerospikes (or even SRBs) makes a serious difference in terms of fuel economy, then it's quite likely that your plane is underpowered to begin with.

And if your sole problem is that you can't get the nose up, I suggest to try sepratrons. You may want to reduce their fuel load, though -- running them for their full duration may already be too much.

[Anquietas314]

And if your sole problem is that you can't get the nose up, I suggest to try sepratrons. You may want to reduce their fuel load, though -- running them for their full duration may already be too much.

Given how heavy the plane is, unless you have like 10 of them at the tip of the nose you'll probably be able to overcome their thrust using the control surfaces. Sepratrons really suck for moving heavy stuff (16 isn't enough to shift an empty orange tank with a skipper by more than 200m/s deltaV in orbit...)

[Fearless Son]

Getting my nose up is not really the problem I have, it is dragging the tail where I run into issues. As usual, I position the rear landing gear just beyond the center of mass, to allow it to lift itself. Unfortunately, the rear of the plane ends up so low to the ground that even a minor raise in the level of the nose on a horizontal surface will cause it to drag and engines to get shorn off.

I have an idea to move the center of mass back and make this less of a concern. Pictures will follow when I have it implemented.

[Anquietas314]

Getting my nose up is not really the problem I have, it is dragging the tail where I run into issues. As usual, I position the rear landing gear just beyond the center of mass, to allow it to lift itself. Unfortunately, the rear of the plane ends up so low to the ground that even a minor raise in the level of the nose on a horizontal surface will cause it to drag and engines to get shorn off.

You could just stick the landing gear on strut pieces, which would give you more room. Hopefully the bigger landing gears that are coming will make that sort of hack unnecessary though

[Fearless Son]

New revision of the Stellar Swan:

Total part count is over 440, unfortunately. I had to add plenty of extra lifting surfaces to balance the weight. Still cannot lift its nose from the runway, having to run off the edge and turn up over the ocean.

Intake is not a problem. The six intakes above this wing here are mirrored below the wing with bi-couplers, each wing is mirrored across the hull, for twenty-four intakes, plus two angled-slightly-down intakes near the front of the craft for twenty six total intakes. Not sure how drag affects a plane based on position relative to the center of mass, but I wanted as many intakes as close to the center of mass and as far behind it as possible to make sure the plane stays stable while air-breathing. As you can see, I also placed a large number of reaction-wheel stabilizers because this thing is so heavy it needs all the help it can get to turn effectively in atmosphere.

And the reason for the twenty-six intakes (which I now realize I may need to increase by a few) are these eighteen turbojet engines to get this massive beast off the ground. Originally I had four to six aerospike engines here, but they proved insufficient thrust and the only places I could add them upset the balance of thrust across the center of mass. So I ended up ripping out that whole assembly and replacing it with a KR-L2 Advanced Engine. However, that is sill not enough thrust to get from the apex of my air-breathing flight to orbit. I might need more areospikes to supplement it.

In terms of flying profile, this still has some problems. Flying it obviously slows framerate, but I accepted that when I built it. It turns quickly enough, but requires some time before velocity vector becomes parallel with the direction of the plane (it has a lot of inertia, obviously.) When it gets to a certain altitude and speed at full thrust, it has a problem where the nose starts to point upward and I have to reduce thrust to keep the nose where I want it to be. Normally I would think the problem was draining fuel from the front of the craft, but I deliberately lock down the front fuel takes to keep the center of mass forward (and I can transfer fuel back once the rear tanks are exhausted.) So far though this is not too much of a problem because this kicks in just before I have to start reducing throttle anyway to keep accelerating without flameout. Landing has been troubling. I have been able to fly up and turn around in a big arc and head back to the Space Center easily enough. By the time I am lined up with the runway am I essentially falling in an unpowered glide, plane perfectly level with the runway, forward speed less than a hundered meters per second, vertical speed feeling reasonable though it is hard to gauge that considering the framerate is slow. It ends up crashing the plane anyway, then crashing the application. That is some crash! I suspect the mass of the plane has a lot to do with it.

Overall, I think I am going to give up on this project, at least until the next update. While I believe the overall design is sound, I am running into the limits of what I believe is practical in stock. I want to wait and see what aerospace improvements were were told is coming are (I hope larger lifting surfaces and engines and intakes are on the way if only to reduce the engineering complexity of something this size.) I am really fond of this design and I hope to refine it with new parts rather than abandon it.

[Anquietas314]

http://i.imgur.com/xWQesTF.jpg

Total part count is over 440, unfortunately. I had to add plenty of extra lifting surfaces to balance the weight. Still cannot lift its nose from the runway, having to run off the edge and turn up over the ocean.

You have way too many wing parts on that thing!. The reason you can't lift off the runway is because your rear wheels are too far back and the plane lacks the leverage to overcome that. Unless you want to risk tail-scraping, there's not a lot you can do about it with such a large plane, besides putting the gears on struts as I recommended in my previous post.

Not sure how drag affects a plane based on position relative to the center of mass, but I wanted as many intakes as close to the center of mass and as far behind it as possible to make sure the plane stays stable while air-breathing.

You want the "center of drag" (not displayed ingame) to be behind the center of mass - it'll make the plane want to point forwards into the air stream, so your approach is correct .

When it gets to a certain altitude and speed at full thrust, it has a problem where the nose starts to point upward and I have to reduce thrust to keep the nose where I want it to be.

The problem here is almost certainly the way fuel drains; check your payload is also locked if it's fuel tanks.

[Laie]

New revision of the Stellar Swan:

Looks good.

And the reason for the twenty-six intakes (which I now realize I may need to increase by a few) are these eighteen turbojet engines to get this massive beast off the ground.

Oh... you may want to increase this by quite a few.

I really, really suggest that you experiment with a much smaller plane. The physics and design principles are the same, but with much fewer parts and possibly a better TWR, the experiments will consume far less of your time. I started out like you did -- every flight took like 20 minutes until things became interesting. Every new idea, every different approach, every slight variation in ascent profile: I first had to sit through a 20-minute ascent before I could see what happens.

Be smart, start small. Designing huge vessels is a whole lot easier if you have some idea of what to expect.