CrasftyDwarf seeks aircraft stability advice.

[CraftyDwarf]

I know this is an old thread, and some things have changed, but I am having this issue of loosing control on reentry. Been thru the check list, but maybe I missed something. CoM is ahead of CoL, acording to designer at the fuel level. Should stay there too, as checked empty.  Got lots of wing, takes off and achieves orbit very heavy loaded. Very big rudders, four of them.  tried thrusting out of it,but my engines tore off.

[Plusck]

1 hour ago, CraftyDwarf said:

I know this is an old thread, and some things have changed, but I am having this issue of loosing control on reentry. Been thru the check list, but maybe I missed something. CoM is ahead of CoL, acording to designer at the fuel level. Should stay there too, as checked empty.  Got lots of wing, takes off and achieves orbit very heavy loaded. Very big rudders, four of them.  tried thrusting out of it,but my engines tore off.

[snip]

Spaceplane flipping generally is because wings and mass are concentrated right at the back of the craft. Basically it's like taking a hammer, putting wings on the head and expecting it to fly handle-first if you throw it just right. Ain't gonna happen.

Edited October 28, 2016 by Vanamonde
Relevance.

[Vanamonde]

The thread you had posted in was extremely old, and much of the information in it no longer was valid. And so, @CraftyDwarf, your question has been moved to its own thread to avoid confusion. 

[WanderingKid]

Pics + Craft file would help tremendously here.

[HvP]

2 hours ago, CraftyDwarf said:

I am having this issue of loosing control on reentry

What exactly do you mean by "losing control" exactly? Is going into a flat spin? Is it flipping over backwards? Or is it just the opposite, where it won't respond to your controls and maybe even nosedives straight down without being able to turn it at all?

2 hours ago, CraftyDwarf said:

CoM is ahead of CoL, acording to designer at the fuel level. Should stay there too, as checked empty.

This is only part of it. Your angle of attack, the placement of your control surfaces, whether your wings or tailplane is high or low on your fuselage all make a big impact on how easy your plane is to control.

2 hours ago, CraftyDwarf said:

Got lots of wing... Very big rudders, four of them

I don't think you mean rudders (which control yaw), but maybe elevators (which control pitch). In fact, based on having four of them I'm willing to bet that you have them placed along the trailing edge of your main wings. If those wings are close to the middle of the plane's center of mass then you probably aren't getting as much control authority out of them as you may think.

Also, make sure that you don't have pitch, roll and yaw assigned to the same control surfaces. Their movements will interfere with each other. Ones on the tips of wings should just be for roll; ones far at the back of the plane should be for pitch; and ones on rear vertical stabilizers should be only for yaw (until you get the feel for more advanced designs.) Be sure to disable the torque on any cockpits or reaction wheels as well, since these can sometimes twist your plane into angles of attack that are not aerodynamically stable.

Maybe it would be helpful if you could post a picture of your craft to imgur.com and link it here.

2 hours ago, Plusck said:

tried thrusting out of it,but my engines tore off.

It may also be possible that your plane is so heavy that there is a lot of flexing between parts. Try the advanced tweakables to create rigid joints and autostruts to reduce flexing. Your wings might even be too big.

[CraftyDwarf]

A few details before I collect some images. The CoM is ahead of the CoL.  You will see that in the first few from the hangar. Second, I am losing control at about 30 to 25k about 1800-1500 m/s decending. It is a large craft. I designed it to fly supplies to low orrbit as part of a career plan, and it is so close to a success. Some overall stats. This monster is successfully carries 120t to low orbit (80k easy). Noting that, it is no surprise that its 474t with the 120t of cargo, 350t fuel full no cargo. Mod notes, but I don't think they are the reason for the control issue. B9 proc wings, mk3 added parts (cleaver engine, shock cones, fuel tanks).Mechjeb for running stats. I don't know if I can trust it to fly it as I have noticed sensitive craft really get crashed from it. My wings too big is a thought, they are large, but I actually take off from the runway. KSP 1.2. My control issue starts with a random side drift, then some variant of a flat spin. and off course on trying to recover it gets a lot worse. I used riggis  attachment on some fuselage points and wing bases, but I could try the engines. I only witth the cargo bay I had would allow solars on the inside of the bay doors. Nothing will attach there. Current mode is sandbox to confirm design.

[CraftyDwarf]

So here come some hangar pics.

Yes, this is a monster, but it doth fly.

Next, CoM

Very clearly Com is AHEAD off CoL. This carries through empty.

Now it is close, but it is ahead of it. I can try to get it more forward I suppose. The cargo  ittem pulled is a USI Kontainer (2.5) full of Uranite.

[CraftyDwarf]

Ok, made further adjustment as suggested, let me describe in specific, because I had some weird results from some of it.

First, found I could pull back the forward canards, may have helped some. Those engines snap off weird, may be something to that mod. See more about that in a bit. It still spun from this, actually higher than  otherwise, but only tried the once, so might have been a random fit. It is also taking more runway and speed tto take off. It spun at 45k alt, and I managed to recover level flight at 10k, was tuning my approach to ksp a bit later underr power, and the engines snap  off. Glided in for the practice, but I would say that was mixed positive at best.

Tried adjusting the proc wings by pulling the offset back. In testing this though, I also did rigid attach on the  engines. And tthey popped off a little past mach 1/6k  altitude, no  manuveuring happening. Was in map expecting a standard climb, when the ball shook and the engine cuttoff sound went off.

Tried to adjust gear for a better takeoff angle of attack with custom move. Applied autostrut to engines. For some reason the plane dropped harder on the runway and shook hard than ever, and as in landed, the engines snap off. So, after I rest a bit, I am going to see if it was just bad luck. overall, not sure if adjustments are helping or not.

[Plusck]

5 hours ago, CraftyDwarf said:

So here come some hangar pics.

Yes, this is a monster, but it doth fly.

Next, CoM

Very clearly Com is AHEAD off CoL. This carries through empty.

Now it is close, but it is ahead of it. I can try to get it more forward I suppose. The cargo  ittem pulled is a USI Kontainer (2.5) full of Uranite.

Ah well...

First off:

On 28/10/2016 at 3:36 AM, Plusck said:

Spaceplane flipping generally is because wings and mass are concentrated right at the back of the craft. Basically it's like taking a hammer, putting wings on the head and expecting it to fly handle-first if you throw it just right. Ain't gonna happen.

So yes, this is the problem here, and "CoM ahead of CoL" isn't going to solve the problem.

As long as its angle of attack is very low, it'll be fine.

However, imagine dropping it belly-first. The mass is fairly centred over the wing, so drag on that part will more-or-less correspond to the centre of gravity and it could feasibly be stable.
But all that long front section with cargo bay and cockpit is going to catch the air, and with its long lever arm is going to turn it until the craft enters a nice, stable engine-first configuration

Put bluntly, it's a hammer with wings on the head.

The only solution is to move mass - and wings - forward. Wings right at the back + long long nose simply can't work.

[CraftyDwarf]

to the comments about "winged hammer" I get what you are saying, just trying to get the craft deliverable to orbit with an accessable bay. SInce the empty weight is a lot engine, I may have to get really creative. I don't care what it looks like in the end, what I need is reliable 100t delivery cycle to orbit, for a true reusable craft system to lower cost. I like this because it gets supersonic pretty good. But if its the extended nose, I will rearang for a different desing method. maybe lenticular.

[Plusck]

7 minutes ago, CraftyDwarf said:

to the comments about "winged hammer" I get what you are saying, just trying to get the craft deliverable to orbit with an accessable bay. SInce the empty weight is a lot engine, I may have to get really creative. I don't care what it looks like in the end, what I need is reliable 100t delivery cycle to orbit, for a true reusable craft system to lower cost. I like this because it gets supersonic pretty good. But if its the extended nose, I will rearang for a different desing method. maybe lenticular.

Well KSP is a bit difficult for this, since the mass of the engines is right by the nozzle. In real life, of course, the mass of the engines is further foward, inside the nacelle (or taking up about two thirds of the body, for fighters).

So yes, that means that you've really got to move your engines further forward. That lets you move the wings forward, which moves the mass further foward... and so on until you've essentially got a standard-looking plane where the main wing surface is either right in the middle of the craft or at least somewhere in the middle third (-ish).

It's maybe a touch boring, but there's a reason why all our planes look... like planes. (as an aside, it always bugged me watching Stargate, since those planes are just too stupidly far from aerodynamically stable - or even usefully unstable)

[CraftyDwarf]

Well, from your statement, you would see why I  headed to this design. the delta dart is the primary re entry shape of landable glide craft.I have seen some lenticular designs on youtube, so I may go with that next. 

[bewing]

In general, you can make this sort of a design fly properly, by adding enough control surface. The plane already has a smallish canard, which is good. If you exchange that pair for 4 tailfins mounted in an X, this thing may still fly nicely.

[Jarin]

The game desperately needs proper visualization for "Center of Drag" =/

[Plusck]

19 hours ago, CraftyDwarf said:

Well, from your statement, you would see why I  headed to this design. the delta dart is the primary re entry shape of landable glide craft.I have seen some lenticular designs on youtube, so I may go with that next. 

Sure, I understand what you mean.

However, I think the Shuttle was really quite extreme for this, given how far back its main wing was. And at the same time, there were mitigating factors:

• it had positively huge pitch control surfaces, compared to the size of the wing,
• those wing strakes were, I suspect, important in providing lift at lower speeds, so the CoL wasn't quite as far back as it appears
• if you look at the shuttle taking a piggy-back on a 747, you can guess where the Shuttle's CoM is: the 747 wouldn't be able to fly if its CoM were too far back, therefore the Shuttle's CoM had to be somewhere around the rear of the middle third of the craft.

Googling other gliding reentry vehicles gives a variety of delta dart shapes, but they all seem a little less extreme (i.e. wings positioned a bit more "plane-like") than the shuttle.

Finally, as far as I know KSP doesn't (yet!) model hypersonic aerodynamics very precisely. Odd things happen with shock fronts (specifically, if I've understood this correctly, the oblique shock wave from the front section of the shuttle creates higher drag forces lower down the craft... but I'm not an expert about that sort of thing...), and I expect that that makes a big difference to what sort of shapes are stable. Since KSP models drag per part (afaik), this and other real-life design issues like canard close-coupling aren't modelled, so there's bound to be a slight difference between KSP-stable and RL-stable.

[CraftyDwarf]

Ok, so I am working on a lenticular arrangement, and one of the other adjustments I am tryng is to move the engines out. I was evidently right on to low for the wing quantity, as I am not getting take off yet. I am guessing that the weight class I am in doesn't have a whole lot of specific plans yet that people have achieved, so I am in kinda pioneer space for this version. So, trying to figure where the imaginary pull to the side on the runway is coming from, and more about how to improove the envelope so control stays from loaded to near empty.

[Plusck]

5 minutes ago, CraftyDwarf said:

Ok, so I am working on a lenticular arrangement, and one of the other adjustments I am tryng is to move the engines out. I was evidently right on to low for the wing quantity, as I am not getting take off yet. I am guessing that the weight class I am in doesn't have a whole lot of specific plans yet that people have achieved, so I am in kinda pioneer space for this version. So, trying to figure where the imaginary pull to the side on the runway is coming from, and more about how to improove the envelope so control stays from loaded to near empty.

Pull to the side is often due to overloading the landing gear - but in your case you seem to have massive gear already...
However, if you're having problems rotating on the runway, you may need to move your rear landing gear forward. They look like they are in two rows, which isn't going to help either rotating or with the overloading (the rearmost will be suffering terribly as you try to lift off).

For stable flight, you really want the engines and wings to be as closely attached as possible, and as close to the core as possible.

I had a craft that was rolling too much on its own, and was told to avoid attaching wings to the cargo bay, which helped. Maybe autostrut would solve that but I haven't tried any planes with that feature yet. In your case, though, moving the cargo bays further back is probably what you need, so I don't know what to suggest

 

[ForScience6686]

My Experience is runway drift is related to lift.   every time I run into it,  the plane had enough lift to become airborne.   my guess is the lift is raising the plane,  which adds movement in the springs,  allowing it to rock to one side.   however you're dealing with higher loads than I've worked with,  and you said it wasn't airborne,  so my only guess is you're almost there,  just a bit more lift needed.

[bewing]

Runway drift is most often caused by the wheels not being perfectly straight (you need to use Absolute Rotation Mode to straighten them). Additionally, too much friction at the front of the craft, and not enough at the rear of the craft. So lowering the friction at the front almost always helps. The problem with Plusck's suggestion about moving the rear wheels forward is that the closer they get to the CoM, the less stability they give you against drift. (You also run a higher risk of tail strikes.)

[CraftyDwarf]

I would like to say that I appreciate the spirit that info was given, even if it was not in a very useful direction. Yes, there was some (minor) balancing.

I have reserved the craft file, as it is functional. I am trying to figure how to attach it to this, but that is not working so far.

Basically, I needed to center the weight of the cargo bay on the balance zone, being a delta wing that is back, yes. This delta wing works because the engines are rear weights. aside from the cargo bay, I needed a bit MORE forward canard, and tho close to loosing control, which I perrsonally attribute to bad piloting, I was all but the final approach, which again bad piloting. tried to hard to land AT ksp center right off, and just about dived into the ground. Aside from the part shift, which is all the same parts but in slightly different spots, ditching the nuclear for an end cap, and makiing the forward liquid tank a last reserve to help with late term balancing. Infact if noses forward a little hard. during re-entry, you have to try real hard to keep the nose up enough to prevent burnup. The radiators help, but control is critical.

It glides so well, that you actually have will go very far past if you usee the youtube method I was given of set peri at 50k over landing target. I overshot KSP by a lot, but had fuel to turn around. turns kinda slow, but what do you want for a cargo craft?

Final details.

* Orbital delivery: 120t to 80km altitude.

Flaps and brakes do help, and with attention you can handle decent and if you are decent pilot, land prretty well. I was diving almost vvertical down on KSC and pulled out at 3k (no power to engines) and aside from some landing ggear deploment issues, was real close to landing anyway. Kinda bounced and broke.

Other import thing I learned. Rigid attachment for engines can cause them to snap off easier for some reason. Autostrut to grandfather works some.

Probably needs more yaw control, maybe. Gonna change a bank of RTGs to a couple Gigantors (out of a service bay). I wish the CLEAVER had an alternator.

btw, to drop the cargo, i decoupled the senior port, and RCS away from cargo. 120t of Uranite (USI resource used pure for weight testing).