Time to learn space planes... What's wrong with mine?

[MitchS]

Oh boy. Here's the album: http://imgur.com/a/uo8pD

And here are the punch lines.

Successful deployment:


UNSUCCESSFUL dCoM measurements...


Saved my Kerbals, though. Thank goodness. I've never killed Jeb. Full play-by-play of how I saved an aerodynamically destabilized spaceplane from a deadly flat spin included in the imgur album. Highly recommend.

I think I need to use an inline cockpit with a fuel tank for the nose. That ought to balance my fuel much better. this was really frustrating, and it was only a 2t payload SSTO. V2 will have a redesigned front end.

Edited December 16, 2016 by MitchS

[zolotiyeruki]

Congrats!

Foamyesque is exactly right--you want your payload to be near your CoM so that its removal doesn't jeopardize your stability.

 

[AeroGav]

@MitchS

Innovative design,  i like it !

I don't know if my longish posts got buried under the welter of advice but i did recommend swapping to an inline cockpit,  mainly for heat reasons.  Also they are more survivable in a crash landing, which hopefully won't ever come into it.  

What angle of incidence are you using BTW? Are you using ALT S/W to apply 5 degrees when the wings are attached to fuselage or are you using fine rotation mode?

KSP models the transonic region,  there is an area of elevated drag between 240 m/s and 440 m/s.     So, during the subsonic climb, you want to stay below 240 m/s.   When your altitude has become such that you're flying more than 7 degrees AoA just to get lift at these speeds, it's time to go supersonic - use prograde or some other nose down input to shallow dive and accelerate past 400 m/s in a hurry.   Then resume your climb.   Ultimately, the jet powered part of the flight doesn't use anything like as much fuel as the nuke bit, so it doesn't matter that much,  we're just talking "ideals".

Once you're on the nukes though,  you want every scrap of forward thrust to count.    This means you got to stay nailed on the AoA that gives best lift drag ratio no matter what.   If you've got significant amount of wing incidence, 3 degrees or more, then best AoA is going to be when the fuselage is pointing at prograde for min drag.   If you only added one or two degrees in fine rotation mode then you'll need a little AoA to get lift, but not much.     Something with generous wing area like yours has should be ok flying at no more than 5 degrees even if the wings have zero incidence.

Don't worry if your craft balloons up to 35km when you light the nukes, then starts descending again.  It is not yet at orbital velocity and it needs a certain amount of lift, it's either going to get by you pitching to 30 degrees like you do in the next picture or by allowing it to descend to 27km at a low AoA.   Guess which results in lower drag !     I suspect the reason you're not just allowing the Phugoid to run it's course is because you're worried about the cockpit overheating.     This is why you need to switch to an inline !    

Heat issues will only get worse when you switch to RAPIER due to higher airbreathing top speeds. ( 1500 at 21.5km is a good benchmark).

Finally - winglets - KSP doesn't model wing tip vortices so i'd probably leave them off.   They're too close to CoM to do much for yaw.    Add elevons to those strakes you're using for vertical stabilizers instead?

I pity for your CoM woes.   I think a passenger spaceplane is much easier to design as a first SSTO, because generally the idea is to bring them back and it gives you a nice permanent counterweight for the engines.

On cargo ships I find myself moving the nukes so far forward it ends up looking goofy,  perhaps forward-swept wings are the way to go - 

Edited December 16, 2016 by AeroGav

[GoSlash27]

MitchS,

 Congratulations on your first successful SSTO design!

 You're already aware of the cargo bay problem and as you can see one precooler is sufficient to feed an engine and can do it without incurring additional drag.

 Your canard control surfaces aren't hinged perpendicular to the centerline. This can induce some irritating yaw coupling during maneuvering.

 Some additional tips for control surfaces: Canards should only respond to pitch inputs. Ailerons should only respond to roll, and (my personal preference) rudders shouldn't respond to *any* input.

 Finally, be aware that canard delta designs often have problems with the pressure center leading the CoM, causing the aircraft to want to flip backwards. Conventional designs with elevators in the back don't look as cool, but they are positively stable during reentry.

Best,
-Slashy

Edited December 16, 2016 by GoSlash27

[GoSlash27]

A very simple Mk.2 RAPIER crew shuttle design:

http://s52.photobucket.com/user/GoSlash27/slideshow/KSP/SimpleSSTO12

Very docile and easy to fly, stable on reentry and landing. Costs $250 per passenger (not counting the flight crew) to operate round trip.

Best,
-Slashy

Edited December 16, 2016 by GoSlash27

[AeroGav]

3 hours ago, GoSlash27 said:

 

 

 Finally, be aware that canard delta designs often have problems with the pressure center leading the CoM, causing the aircraft to want to flip backwards. Conventional designs with elevators in the back don't look as cool, but they are positively stable during reentry.

Best,
-Slashy

I have to disagree with that, Canards can be made just as stable.  There's nothing particularly wrong with the low speed/high alpha handling of this design.  Or it's crashworthiness...

 

[GoSlash27]

40 minutes ago, AeroGav said:

I have to disagree with that, Canards can be made just as stable.  There's nothing particularly wrong with the low speed/high alpha handling of this design.  Or it's crashworthiness...

AeroGav,

 It's not the low speed end you have to worry about, it's the high speed end, particularly reentry. It's not universal and can often be fixed, but it's a potential pitfall that he needs to be aware of.

Best,
-Slashy

[AeroGav]

16 minutes ago, GoSlash27 said:

AeroGav,

 It's not the low speed end you have to worry about, it's the high speed end, particularly reentry. It's not universal and can often be fixed, but it's a potential pitfall that he needs to be aware of.

Best,
-Slashy

What flight regime is this going to manifest in?  High alpha, or at cruise AoA when fast and high?

This has been debated endlessly on other threads, with neither side showing much sign of movement.   A canard has more centre of pressure up front because of the canard, but the wing is further back so this cancels out.  Also the SPH indicators are  surely going to take account of this.    The only differences I can see are drag and lift from the fuselage, which the SPH blue ball doesn't allow for.  A canard design is likely to have different fuselage layout.       

It is far more likely his problems are being caused by dry CoM.  The design i posted had nuke engines mounted further forward, on forward swept wings.    I recommend

• RCS build aid (indicates dry CoM with a nice red ball)
• CorrectCoL (accounts for fuselage lift)

Also , with the wing area his plane has, it will not need to use extreme nose up angles to re-enter safely. 

[MitchS]

@AeroGav
"What angle of incidence are you using BTW? Are you using ALT S/W to apply 5 degrees when the wings are attached to fuselage or are you using fine rotation mode?"
          I'm using fine rotation, one notch. Right around there. I might have turned off angle snap and adjusted that sliiiightly lower than that one fine notch, because it looked a bit aggressive and high-drag... How many degrees are added with every fine notch? Do I actually want more than one?
"I suspect the reason you're not just allowing the Phugoid to run it's course is because you're worried about the cockpit overheating."
          Phugoid? I'm sorry, what does that mean?
"Finally, winglets... Add elevons to those strakes you're using for vertical stabilizers instead"
          Ah. Good observation. They USED to be well behind my CoM, but now they're vestigial organs from an earlier build and I didn't notice. I'll have the SPH guys knock them off when I get home. Haha.
"I recommend... RCS Build Aid..."
          I use CKAN on 1.2, and RCS Build Aid (a must have in all my other career saves...) is presently unavailable, which has frustrated me every time I've opened up the game.

@GoSlash27
"Your canard control surfaces aren't hinged perpendicular to the centerline. This can induce some irritating yaw coupling during maneuvering."
          THANK YOU. I have been noticing this for ages and haven't known why it happens. I'll fix it. Should have known better.
"...and (my personal preference) rudders shouldn't respond to *any* input."
          Will they still be engaged by SAS if they aren't set to control yaw? The way you make it sound, you don't like rudders at all and may as well leave them off, right?
"Finally, be aware that canard delta designs often have problems with the pressure center leading the CoM, causing the aircraft to want to flip backwards."
          Can you expand on this please? Why would delta canards offer problems where (I assume you mean) square ones wouldn't? What kind of other guidance is there besides aesthetics regarding canard design choice?

Looking forward to round two of design this evening. Has anyone seen the 20-minute video by Kergarin Aerospace (Who is he on the forums? Anyone?) where he completes the Ultimate Challenge with an ISRU spaceplane? Ever since I saw that earlier this week, I have to admit that I've been obsessing over SSTOs and interplanetary spaceplanes. ...My new joystick should arrive in the mail by Monday. Hahaha.

 

Edited December 16, 2016 by MitchS

[GoSlash27]

34 minutes ago, MitchS said:

 Will they still be engaged by SAS if they aren't set to control yaw? The way you make it sound, you don't like rudders at all and may as well leave them off, right?

 The way I do it, I have a vertical fin (or perhaps a pair) that will passively correct any deviation in yaw, but does not *actively* induce a change in yaw. The reason I do this is that SAS will attempt to maintain a constant heading when the aircraft is banked. This creates a cross-controlled situation where the aircraft crabs away from the turn, increasing drag and making flight response sloppy.
 There are no crosswinds in KSP, so active rudders aren't necessary. Passive fins are plenty adequate to maintain stability in yaw.

34 minutes ago, MitchS said:

Can you expand on this please? Why would delta canards offer problems where (I assume you mean) square ones wouldn't? What kind of other guidance is there besides aesthetics regarding canard design choice?

What I mean is designs with the pitch control surfaces in front of the center of mass can behave oddly during reentry. The reason for this is the center of drag isn't the same as the center of mass or center of lift, and is not shown in the hangar.

Eyeballing from the plan view, most of the cross sectional area is ahead of the center of mass, *especially* with Mk.2 designs. At high speed during the heaviest part of reentry, the aircraft actually prefers to fly *backwards* and can be difficult or even impossible to recover once in that state.
 There's nothing that shows in the SPH that will warn you of this condition. My workaround is to use a conventional layout for spaceplane designs because they have no tendency to want to fly backwards.

Best,
-Slashy

Edited December 16, 2016 by GoSlash27

[GoSlash27]

25 minutes ago, AeroGav said:

What flight regime is this going to manifest in?  High alpha, or at cruise AoA when fast and high?

This has been debated endlessly on other threads, with neither side showing much sign of movement.   A canard has more centre of pressure up front because of the canard, but the wing is further back so this cancels out.

 It manifests during the high G portion of reentry. It doesn't cancel out because the CoM tends to be in the rear of the aircraft, giving the canards all sorts of leverage to rotate the aircraft backwards.

30 minutes ago, AeroGav said:

Also the SPH indicators are  surely going to take account of this.    The only differences I can see are drag and lift from the fuselage, which the SPH blue ball doesn't allow for.

That's a large part of it, but the main problem is that the center of drag (pressure center) is completely different from the center of mass or center of lift and it's not shown in the SPH. I believe FAR gives you a tool to visualize it, but stock doesn't.

Now... this isn't to suggest that canard delta designs *always* exhibit this instability or that it cannot be cured. Merely that it is a known problem with canard designs that doesn't manifest in conventional designs. Something to be aware of.

Best,
-Slashy

[AeroGav]

15 minutes ago, MitchS said:

@AeroGav
"What angle of incidence are you using BTW? Are you using ALT S/W to apply 5 degrees when the wings are attached to fuselage or are you using fine rotation mode?"
          I'm using fine rotation, one notch. Right around there. I might have turned off angle snap and adjusted that sliiiightly lower than that one fine notch, because it looked a bit aggressive and high-drag... How many degrees are added with every fine notch? Do I actually want more than one?

 

I'm not sure how many degrees with fine notches but 1 degree might be about right.  In coarse rotation mode the game works in 15 degree steps which is way too much.    You can get a 5 degree step however,  if you hold down shift at the time you are attaching the wings to the fuselage, then press S  (or W?) to angle the wing up or down.  Unfortunately it's a bit fiddly because if you hold shift while attaching it's quite easy to introduce unintended dihedral/anhedral  unless your mouse cursor is perfectly centred.

Of course you need to attach every lifting surface at the same angle or you'll introduce large amounts of unintended pitch trim to the design.  Though I often go over it again with the fine rotation widget on the foremost and aftmost surface so that i get the thing to fly at a small positive AoA "hands off" .

One degree incidence isn't enough imho,  you'll still need a fair bit of body AoA to get adequate lift and that means drag. 

Mach 1.7
Wing AoA	1	2	3	4	5	6	7	8
L/D ratio	2.5	4.79	6.58	8.67	9.15	9.45	9.51	9.47

 

^{I did some tests with the debug menu on showing the lift and drag of the wing panel itself (not the whole aircraft).  In the same tests i found there was little difference between mach 1.7 and mach 5.5 (but a huge differences between mach 1.7 and lower speeds).  I also found that L/D ratio was independent of altitude.}

 

Quote

"I suspect the reason you're not just allowing the Phugoid to run it's course is because you're worried about the cockpit overheating."
          Phugoid? I'm sorry, what does that mean?

https://en.wikipedia.org/wiki/Phugoid

A phugoid or fugoid /ˈfjuːɡɔɪd/ is an aircraft motion where the vehicle pitches up and climbs, and then pitches down and descends, accompanied by speeding up and slowing down as it goes "downhill" and "uphill."   This is a classic example of a negative feedback system.    The phugoid has a nearly constant angle of attack but varying pitch, caused by a repeated exchange of airspeed and altitude.

 

 

[AeroGav]

8 minutes ago, GoSlash27 said:

 It manifests during the high G portion of reentry. It doesn't cancel out because the CoM tends to be in the rear of the aircraft, giving the canards all sorts of leverage to rotate the aircraft backwards.

Now... this isn't to suggest that canard delta designs *always* exhibit this instability or that it cannot be cured. Merely that it is a known problem with canard designs that doesn't manifest in conventional designs. Something to be aware of.

Best,
-Slashy

I think you're saying that in these conditions the canard has a lot of drag (post stall?) .  I take it these are conditions where drag is more significant than lift (the sph only indicates lift)  - stalled?.   The same must also be true of the wing however, Yes it's much closer to CG but it is also much larger...  unless you're saying that the canard surface is more affected by these flight conditions than the wing is?

 Bear in mind on a conventional layout, you can also deep stall the tailplane at very high angles.  You increase it's AoA to get more lift to raise the tail and lower the nose, but since the tailplane is already >30 AoA (aoa of airplane plus aoa from your control input) the increased angle results in reduced lift -  the controls work opposite to as expected.      

I'll have to try two otherwise identical aircraft with the exception of pitch control method, adjust them to have same relative CoL / CoM and see if there's any discernible difference in handling.   BTW this is my "9 minute speedbuild" ssto trying to pitch up to wash off speed on re-entry as it was overshooting KSC.   Inertia and reaction wheel torque briefly got the nose past stall angle, but it wouldn't stay there.  2099m/s at start and 35km up..

Yaw control is more of a problem at high alpha,  i find.

 

[GoSlash27]

1 hour ago, AeroGav said:

I think you're saying that in these conditions the canard has a lot of drag (post stall?) .  I take it these are conditions where drag is more significant than lift (the sph only indicates lift)  - stalled?.   The same must also be true of the wing however, Yes it's much closer to CG but it is also much larger...  unless you're saying that the canard surface is more affected by these flight conditions than the wing is?

Yes, it's a drag thing. Has nothing to do with lift.

The problem is that the CoL of the main wing is very close to the CoM, and exhibits no pitch moment with drag. The canard and all that fuselage out front tho... that exhibits a whole lot of pitch moment. If you're not careful, a whole lot more than you can control.

1 hour ago, AeroGav said:

 Bear in mind on a conventional layout, you can also deep stall the tailplane at very high angles.  You increase it's AoA to get more lift to raise the tail and lower the nose, but since the tailplane is already >30 AoA (aoa of airplane plus aoa from your control input) the increased angle results in reduced lift -  the controls work opposite to as expected.      

Aye... but that's real life rather than KSP. Control surfaces operate differently than wings in KSP and can generate plenty of pitch authority even when perpendicular to the airflow. I'll post some reentry pics to demonstrate a little later.

In your case... You may have enough reaction wheel torque to overcome the effect, so it may not affect this plane. Just be aware that it *is* a thing in KSP.

Best,
-Slashy

[GoSlash27]

 

Here you can see the pitch authority of small elevators during reentry. They don't lack the ability to fly at 90° pitch until well- under the speed where overheating is an issue. Note that pitch instability is never a problem and that I have added no reaction wheels.

Best,
-Slashy

 

[MitchS]

@Slashy (I mean @goslash27), your "test pilot" approach to concept demonstration is very much appreciated!

Edited December 16, 2016 by MitchS
Edit: You aren't @Slashy, and I am @Idiot

[GoSlash27]

MitchS,

 I happened to have an example in orbit from this morning

 Since the demonstration of inherent stability during reentry is part of my certification routine for all of my "manned" vehicles, I had already done this previously before putting it into service in my career. It's kind of an unfair demonstration in that sense because I already knew what would happen.

 So now I'll venture into the unknown and convert it into a canard configuration while changing as little as possible to show how differently it behaves.

Best,
-Slashy

[AeroGav]

OK,

I kept working on my dual nuke single whiplash concept, adding angled wings.     The main lifting surfaces are angled at 5 degrees.  Due to fine tuning with the rotate tool,  the passive tailplanes are angled at 3 degrees, the strakes at 6 degrees,  and the canards at 8 degrees.  As a result the front bits stall first, but when flown "hands off" it maintains an AoA of 1.8 degrees.   When prograde hold is used, this drops to just 0.2deg positive, and supersonic lift:drag ratio of 4:1 is possible.

The forward swept wings make it look like something that belongs on a stick, in an Aztec temple 

https://kerbalx.com/AeroGav/Firefox-II

Nearly 4 tons of fuel left...

In view of what we've just been discussing,  I decided to verify re-entry characteristics.

First up,   I retro burned to 30km  then left it set on prograde to see what'd happen.    However, SAS started going crazy, alternating roll reversals that were wasting power and destabilizing the ship, so below 50km i turned it off.   "Hands off" stability in the upper atmosphere

It ended up skipping off the upper atmosphere so many times, we completed another orbit before coming down for good.    Did this heat soak the cockpit?   Not too bad, internal temp peaked at 890K.  Without 4 tons of fuel aboard, it'd probably re-enter cooler

Next up,  i let it re-enter completely uncontrolled with SAS off and no inputs from 70km.   It started to rotate slowly, then as the atmosphere got a hold of it, it aligned itself to prograde.  Unfortunately , inverted :

Lastly, I tried re-entering at a high angle of attack as some folks insist on.    Holding down full nose up input, the nose bobs up and down, reaching 50 degrees of pitch before the aero forces slam it back down again.   Obviously, the higher up you are, the further you can get the nose up before aero forces overcome inertia and the torque wheels.     This washes off speed incredibly quickly.

@GoSlash27 The problem is this argument is very hard to settle, there will never be a true apples to apples comparison that doesn't get dogged by "whataboutery". All this proves is that we both build aircraft that are stable on re-entry.    Does the Firebird even qualify as a canard?  It has canards and tailplanes (and strakes).  However the horizontal stabilizers are passive with no elevon controls, the nose is pitched up and down only from the front.     And how are the Firebird's combination of mid-mounted main wing and passive tail planes different to a large , rear mounted  wing with the same total area and overall CoL position?

Edited December 16, 2016 by AeroGav

[AeroGav]

@GoSlash27    

12 hours ago, MitchS said:

@Slashy, your "test pilot" approach to concept demonstration is very much appreciated!

OK,  guys i think we have something.

I made a very simple spaceplane,  set infinite fuel and electricity cheats and used "set orbit" to put it in space.

Here is the Canard version 

http://pastebin.com/74XUmWv3

Then i made a tailplaned version.  No changes to the fuselage, i just attached the canards to the engine, then moved the wings forward a bit so that the CoL indicator just touches the back of the yellow ball, as it does in the above example. Note - I do have mod CorrectCoL installed, it is possible this is affecting the results of this test, by giving me more accurate CoL indicators in SPH to work with.

http://pastebin.com/73rZcd88

I did give them a both a quick low altitudye flight.    At night, you cannot tell which aircraft you are flying.  Neither allows more than about 8 degree AoA, and can only be stalled by climbing steeply with the power off.  After stalling, both go prograde - ie. vertically down !

Next, I cheated them to orbit and made a retro burn to 30km.   I set Radial Out and made a note of at what altitude the cockpit reaction wheel could no longer maintain 90 degree nose up   against the aerodynamic forces :

In the canard meanwhile :

Next,  I made a note of at what altitude it is no longer possible to exceed 30 degree (stalled) angle of attack.   On the tailplane test, i did this by leaving Radial Out set,  and pretty soon we started yawing from side to side viciously.  It turns out this is a product of SAS trying the rudder in desperation to get the nose up.  When I switched to Prograde hold then held full UP input, the way oscillations died down.  This is the approximate altitude where the wings became un-stalled :

On the Canard -

 

Edited December 17, 2016 by AeroGav

[GoSlash27]

Observe the difference when I take the earlier example and move the elevators forward.

http://s52.photobucket.com/user/GoSlash27/slideshow/KSP/SimpleSSTO12/SimpleCanard12

Static stability is still the same as the conventional design, yet *dynamically* it becomes uncontrollable during reentry. It's only barely recoverable once the speed comes down. No Kerbals were harmed during this demonstration...

 This example was less harrowing than many delta canards because the CoM is still centered longitudinally. Many of these designs have the CoM toward the rear, which exacerbates the problem. Lots of draggy Mk.2 fuselage ahead of the CoM makes it worse. The only thing that transformed this from a docile spaceplane into a beast was moving the elevators forward.

AeroGav,
 This isn't about your designs vs. mine. I have built many successful delta canards over the years in KSP and there are folks out there who (IMO) build better SSTOs than mine (lookin' at you, @Starhawk ). This is simply about pitfalls that new designers may run into. Canard deltas can be perilous and frustrating for new designers, so it is helpful to know the mechanism by which they have been known to bite you in the butt.

 Best,
-Slashy

 

 

[AeroGav]

@GoSlash27    

Hah ! How did we get such different results !  In a PM, you mentioned dynamic stability.  I must admit, this is something I don't understand properly, so excuse me if i don't talk about it 

However, if you look at pictures 5 and 6 in your album,  i think the problem is one of static stability.    Look at the inputs SAS is making to hold the nose where it is - pitch down !    This mean's the aero forces are trying to push the nose further up.     On picture 6, you're running out of nose down authority and after that it flips.        On my test craft,   SAS is having to apply nose up input to stop the nose falling, eventually it runs out of strength.

I think this is a case of "the stock CoL indicator lied".     It doesn't allow for body lift.  CorrectCoL does, also my mk1 has very little body lift.  Alternatively there could be some weirdness resulting from fuel burnoff.   I used infinite fuel cheat to rule it out.   The canard version has wings further aft, and fuel is stored in the wings but not the canards.  Though if anything that should work in the canard's favour once it burned off, if the ship was balanced with full tanks ?

Edit - the last pic in your album seems to indicate something important

That craft is extremely unstable at 42 m/s and 21M altitude.   You have an AoA of under 5 degrees, yet the SAS is having to use almost max nose down input to stop the nose coming up even more?  I think that's a static stability problem, or a problem caused by CG shifting to rear as the fuel burns off.

Edited December 17, 2016 by AeroGav

[GoSlash27]

26 minutes ago, AeroGav said:

@GoSlash27    

Hah ! How did we get such different results !  In a PM, you mentioned dynamic stability.  I must admit, this is something I don't understand properly, so excuse me if i don't talk about it 

However, if you look at pictures 5 and 6 in your album,  i think the problem is one of static stability.    Look at the inputs SAS is making to hold the nose where it is - pitch down !    This mean's the aero forces are trying to push the nose further up.     On picture 6, you're running out of nose down authority and after that it flips.        On my test craft,   SAS is having to apply nose up input to stop the nose falling, eventually it runs out of strength.

I think this is a case of "the stock CoL indicator lied".     It doesn't allow for body lift.  CorrectCoL does, also my mk1 has very little body lift.  Alternatively there could be some weirdness resulting from fuel burnoff.   I used infinite fuel cheat to rule it out.   The canard version has wings further aft, and fuel is stored in the wings but not the canards.  Though if anything that should work in the canard's favour once it burned off, if the ship was balanced with full tanks ?

AeroGav,

 The actual problem is that the stock indicators don't account for CoP. They do a reasonable approximation of CoL and a very precise accounting of CoM, but it's left to the player to account for CoP and many don't know that it's even a thing. FAR recognizes this and represents the CoP as an orange ball, It exists in stock and affects your stability (or instability in this case) in stock, but is not shown.

 The SAS is responding to forces that it experiences in-game, but there's nothing evident in the SPH to warn you of those forces. That's why conventional designs tend to push the nose forward while canard designs tend to make it want to fly backwards. One approach is inherently stable while the other is inherently unstable. It's always better to have a design that's inherently stable. That way if you lose control, it automatically reverts to a flyable state.

Put another way... There are actually 3 centers. Center of lift, center of mass, and center of drag. The stock SPH only shows 2 of these, The relation between drag and mass is dominant during reentry, so it's imperative that it be accounted for. Unfortunately, you have to guesstimate that. A drag center ahead of the mass is catastrophic and not recoverable until flying speed comes down to the point where lift is dominant.

Best,
-Slashy

 

Edited December 16, 2016 by GoSlash27

[MitchS]

Fascinating. So can one of you (or both!) boil down your last few posts into a couple of takeaways? What are your bottom lines, where do you agree with one another, and where do you disagree? Am I to understand that canards are a less stable design than one without them?

[GoSlash27]

MitchS,

 I think that both of us will agree that canard designs can be tweaked so that they are stable. We have both done that ourselves.

 Beyond that, all of this is just my personal opinion and experience.

 You will have a much easier job if you put the elevators behind your wings instead of in front of them. Putting them in front is still within your ability as a new spaceplane designer, but it will make your ship behave in ways that you won't expect at first. If you put them in back, it will make it more difficult to hold the nose up, but it will be easier for you to control as a new pilot.
 Not saying that you should avoid a canard design, just saying that you should be aware of how such designs behave during reentry and why they act that way.

I'll leave it to AeroGav to outline any points where he disagrees.

Best,
-Slashy

 

[foamyesque]

Personally, I build canarded spaceplanes specifically because they're easier to keep pitched to a high AoA. No disrespect to @GoSlash27, but two standard canards on a craft that small are an absolutely huge amount of control authority, not "small" in any sense. I've driven machines many times the size on just those

In the stalled regime CoL and CoD are functionally equivalent for aerodynamic craft.

I have encountered the CoP/CoL differential @GoSlash27 is talking about with a recoverable booster design that, unfortunately, needed an unstreamlined nose. Pitch control was fine, but my yaw control was very much not. Yaw stability in general is actually not well-indicated in the SPH, since the fuselage has a much larger sail area relative to the stabilizer and so its effects are very noticeable in flight but are not shown in assembly as they don't generate lift.

Oh, and a few other rules of thumb concerning wing surfaces and their positioning (assuming stability):

1. The larger the area you have, the more you bend your path to where your nose is pointing;

2. The larger the lever arm, the more your nose points to your path;

3. The higher the aspect ratio, the more quickly you can rotate about the long axis and the slower you rotate about the other;

 

(I also, FWIW, disagree with @GoSlash27 on simply having a yaw stabilizer and no rudder. You can get away with it on small craft, where the cockpit reaction wheels are adequate to provide yaw corrections, but on machines of appreciable size a rudder is vital for correcting adverse yaw, runway control, fine-tuning landing approaches, and coordinated turns if you're doing anything other than straight forward launches and landings)

Edited December 16, 2016 by foamyesque