Plane veers to the left. Stumped.

[jsisidore]

I tried ckan too, it gives me microsoft net framework error.

Edited December 5, 2016 by jsisidore

[jsisidore]

I tried ckan again, somehow it worked this time and I was able to install the mod.

However, the mod is a bit vague, although I can change the "rotation amount", I have no idea what unit of measurement it is.

Edited December 7, 2016 by jsisidore

[AeroGav]

1 hour ago, jsisidore said:

I tried ckan again, somehow it worked this time and I was able to install the mod.

However, the mod is a bit vague, although I can change the "rotation amount", I have no idea what unit of measurement it is.

Good for you 

Unfortunately, mine just broke when the 1.2.2 update installed 

The way i use it, is to go to the settings button, then the Angle Snap tab.     

By default the mod allows angle snap to work in 1 degree, 5 degree, 15 degree (stock) and 30 degree increments, 

You can increase the step with C and decrease it with ALT  C

However with the Settings > Angle Snap box you can remove presets and ADD new ones.

ie. if you want your wings a 7,  ADD a preset for 7 degrees , click SAVE , then select the 7 degree angle snap step size with C/ALT C

Less error prone than eyeballing it.

[jsisidore]

14 hours ago, AeroGav said:

Good for you 

Unfortunately, mine just broke when the 1.2.2 update installed 

The way i use it, is to go to the settings button, then the Angle Snap tab.     

By default the mod allows angle snap to work in 1 degree, 5 degree, 15 degree (stock) and 30 degree increments, 

You can increase the step with C and decrease it with ALT  C

However with the Settings > Angle Snap box you can remove presets and ADD new ones.

ie. if you want your wings a 7,  ADD a preset for 7 degrees , click SAVE , then select the 7 degree angle snap step size with C/ALT C

Less error prone than eyeballing it.

Angle snap works but only crosswise for some reason. I think I will look for a more user friendly mod.

[jsisidore]

Anyway, I have uninstalled the mod, and just used shift to get 5 degree angle. I found out that shift gives you 5 degrees.

I built another ship, more aerodynamic, it gave me a ratio of around 5-4 at 3000m. Problem is when I try to pitch down to gain speed - I lose altitude and go into a nosedive.

See for yourself: http://pastebin.com/1t7tSdyM

 

[jsisidore]

I made no change and was able to go past 20000ish and past 1100m/s, again the controls are extremely springy, very hard to steer, one wrong move and you're spinning out of control.

[AeroGav]

5 hours ago, jsisidore said:

Anyway, I have uninstalled the mod, and just used shift to get 5 degree angle. I found out that shift gives you 5 degrees.

I built another ship, more aerodynamic, it gave me a ratio of around 5-4 at 3000m. Problem is when I try to pitch down to gain speed - I lose altitude and go into a nosedive.

See for yourself: http://pastebin.com/1t7tSdyM

 

This one's not bad.

The reason the controls are "springy" is because you angled the main wing up 5 degrees, but did not do the same to the strakes either side of the cockpit.  Then you balanced centre of lift with centre of mass with the two surfaces at different angles.

As a result, the indicators in the SPH make it look stable, but it is not !  Your ship is actually tail heavy and unstable, but that is disguised because you've got a canard on the front angled down 5 degrees lower than the main wing.   This is a nose down input, but the thing is , if the centre of lift is in front of com, then it still unstable.  The more the nose goes down, the more it wants to go down.  The more it is up, the more up it wants to go.

What i did 

1.  angled the canards at the front up 5 degrees to match the main wing.

2.  this made the CoL go in front of the centre of mass, so i had to move the wings back a bit.

3. now it's stable but very nose down in flight.   So in fine rotation mode, i angle the front strakes up a bit more, and the rear elevons down ever so slightly, until it flies with a very small nose up angle with SAS off and no control input.

The front strakes are now at a bigger angle of attack then the main wing.   They will stall first.   The elevons are at a slightly lower angle than main wing.  They will stall last.  If you pull up too hard the front end looses lift first and makes the nose go back down.

Also, those tail fins are too small and too near the middle.

I made doubled the number of them, and attached them to the rearmost wings, not the engines, as that lets me get them further back.  It still snakes from side to side a bit if you turn SAS off but it doesn't flip.

Now, if you'll excuse me i need to try flying it to orbit.  The game is paused right now and i'm at 2.5km, but i thought you might appreciate a quick reply !

http://pastebin.com/vS3X33Tm

Edited December 8, 2016 by AeroGav

[AeroGav]

OK - made orbit with 2432LF out of the 7160 it took off with.  I saw a supersonic lift drag ratio of 3.7 which is good.  That was with SAS set to prograde hold.  The numbers get worse when you hand fly or use stability hold because you're not holding exactly to the prograde.

 The best i ever had was 4.3.

It's a 91 ton ship with only 35 tons of fuel aboard, so it's doing ok to reach orbit with a cargo and still have a 30% fuel left.

It's got  a lot of lift, it does like to zoom above the flameout altitude when you're using prograde hold.

First time was right after going supersonic.  It went into a steep climb , reaching 650m/s and a peak altitude of 33km before starting back down.

On the next zoom, it reached 42km and 1200 m/s or so.     The next one it got to mach 4.6+ and the AP was predicted to go 52km+ on the map screen.   Lots of heat bars appeared at the bottom of the last dive, so i pressed space to start the nukes and we went to space this time.

 With more fuel aboard, it might not be so keen to jump out of the atmosphere prematurely.

You might be interested in the mod Configurable Containers .

You can use this to change your LF/O tanks to store only LF.    Also, if you built your vertical stabilizers out of strakes like i tend to do,  you'll have more fuel at the back of the ship, which means you can fill more of your front tanks up and still be balanced. 

 

 

 

[AeroGav]

 

56 minutes ago, AeroGav said:

The reason the controls are "springy" is because you angled the main wing up 5 degrees, but did not do the same to the strakes either side of the cockpit.  Then you balanced centre of lift with centre of mass with the two surfaces at different angles.

As a result, the indicators in the SPH make it look stable, but it is not !  Your ship is actually tail heavy and unstable, but that is disguised because you've got a canard on the front angled down 5 degrees lower than the main wing.   This is a nose down input, but the thing is , if the centre of lift is in front of com, then it still unstable.  The more the nose goes down, the more it wants to go down.  The more it is up, the more up it wants to go.

What i did 

1.  angled the canards at the front up 5 degrees to match the main wing.

2.  this made the CoL go in front of the centre of mass, so i had to move the wings back a bit.

3. now it's stable but very nose down in flight.   So in fine rotation mode, i angle the front strakes up a bit more, and the rear elevons down ever so slightly, until it flies with a very small nose up angle with SAS off and no control input

BTW I used the mod CorrectCoL to work the above out.   You can see in that graph, the line needs to slope downhill from left to right.  If it does not , you need to move your CoL back some more.

The point where the line crosses the x axis (horizontal axis) is the AoA your airplane will try to find if it's got no control input from you or SAS.   I adjusted the canards/elevons until the plane was trimmed to fly at a small positive AoA when left with no input.

The horizontal axis is basically AoA and the vertical axis is the pitching moment the aerodynamics generate.  Positive values on the vertical axis mean the nose is trying to go up, negative values mean it's pushing down.   If that line slopes "downhill" then that means it tries to pitch itself up when the nose is down  , so the plane is stable.

It's hard to see in that picture, but it also generates a thin vertical blue line.  That is the AoA your plane needs for level flight at the conditions you've specified.   So if you want to fly at 21.5km and 1350 m/s, you can put those numbers in and see how much AoA it will need in those conditions. Helps you work out if you need more or less wing.

Edited December 8, 2016 by AeroGav

[AeroGav]

One last thing - 

In the above picture, you will see that the mk1 FL-T800 tank, mk2 rocket fuel fuselage long, and Rockomax X200-8 all hold the same amount of fuel - 800 units.

However, the mk1 has 1.5kn drag,  the mk2  2.95kn (twice as much as the mk1)  and the Rockomax only 0.44 (one third of the mk1, and one sixth the drag of mk2).

So mk2 parts aren't good for drag  - though the mk2 cockpit is less draggy than the mk3 (as well as lighter).  They don't make good fuel tanks though.  I've noticed you are using mk2 fuselages as engine nacelles.   Now, the mk2 bicoupler is in fact the least draggy 2 way engine adapter in the game, but the rest of the mk2 parts are bad, so you want to go straight from that bicoupler to a mk2 to mk1 adapter short and get back to mk1 fuselage size for the front part of that nacelle ASAP.

As well as causing drag, you have to realise that vertical surface ahead of CG is fighting against the tail fin, acting to make the plane less stable in yaw.

[foamyesque]

It isn't that the Mk2 parts are bad, it's that they're also wings, so if you have AoI on them they have lift-induced drag, I believe. Treat 'em as wet wings, not as fuel tanks, and things will go better. I have also noticed that the same object positioned differently longitudinally on the craft will have differing amounts of drag; you can see this, for example, if you build a stack of a nose cone, two identical fuel tanks, and an identical nose cone acting as a tailcone. Hold it on launch stabilizers and check the Cds; the two fuel tanks will have different values.

 

Also there's a rumour out there about the X-couplers being draggy. At least in the case of the quadcoupler, it's wrong; the quad's drag is actually somewhat less than any other means of attaching four engines to something I have found.

[AeroGav]

36 minutes ago, foamyesque said:

It isn't that the Mk2 parts are bad, it's that they're also wings, so if you have AoI on them they have lift-induced drag, I believe. Treat 'em as wet wings, not as fuel tanks, and things will go better. I have also noticed that the same object positioned differently longitudinally on the craft will have differing amounts of drag; you can see this, for example, if you build a stack of a nose cone, two identical fuel tanks, and an identical nose cone acting as a tailcone. Hold it on launch stabilizers and check the Cds; the two fuel tanks will have different values.

 

They still make drag even at 0 AoA - that screenshot i posted has the rocket on prograde at 0 aoa.    Also the lift drag ratio of wings is much better than off mk2 fuselage. I did some tests and got supersonic l/d ratio of 9.5 to 1 on wing/control surface sections, mk2 ain't near that.  Also note that i've never gotten a craft over 4.3 l/d ratio supersonic so even with the fuselage at 0 aoa and with quite large wings at 5-7 degrees aoa, most of the drag is coming off the fuselage to bring our overall l/d down by more than half.

on mk2 spaceplanes i'm lucky to see much more than 3 to 1

[foamyesque]

The correct test comparison would be between a Mk2, no wings layout and a Mk1 with equivalent lift area and mass, with 0 AoI on its wings. I haven't run it yet, but it'd be interesting to see what the terminal velocities on the pad are.

 

'course, you also seem to get much higher L/D ratios than I bother with. Anything above 1.5's good enough for me.

EDIT:
 

@AeroGav, I've run the tests now and you are correct; the Mk2s are draggier than a wing & Mk1 tank combination. I was getting terminal velocities of 430m/s v. 470m/s, and the Mk1s actually had slightly more wing area.

Edited December 9, 2016 by foamyesque

[AeroGav]

@jsisidore pictures of my flight in your "Dragon 001" craft

Ended up in a very wonky 110 x 72km orbit because i wasn't prepared for the zoom climb tendency this has.  Just four Rapiers and 6 nukes with no oxidizer, able to lift a really big ship !

Edited December 9, 2016 by AeroGav

[jsisidore]

Oh snap! Forgot the streaks! Awesome! I think I can rename it into Dragoff now. Can't wait to get it into orbit. Btw were you too short on fuel to head over to Dune? I can't remember my last trip, but 2/3 is what you need right?

If I will go back to mk1 for engine mount I will have to use drag-ugly adapters wouldn't I?

[AeroGav]

No i was suggesting you build the nacelles like this

1. nose cone

2.  mk1 liquid fuel fuselage 

3. a few more mk1 liquid fuel fuselages

4. mk1 to mk2 short adapter

5. mk2 bicoupler

6. pair of engines

 

at the moment you got

1. nose cone

2. mk1 to mk2 short adapter

3.  mk2 liquid fuel fuselage long

4. mk2 liquid fuel fuselage short

5. bicoupler

6. pair of engines

 

RE: trip to Duna

Zero Fuel Weight of your ship with payload  56.04 Tons

Fuel Mass remaining in low orbit                     12.16 Tons

ISP  of NERV engines                                      800

plugging these numbers into a Delta V calculator gives me   1542 dV

You need 950dV to break free from Kerbin and 130 to get a Duna Intercept.

Set your PE for 16km and you can aerobrake into Duna and make a gliding landing, so no more fuel needed at that stage.

[AeroGav]

I didn't have a save from orbit, so I had to fly up again.     By the way, the Old School Anthems Youtube playlist had "O.T. Quartet - Hold this sucker down"  as I was passing 30km.  Got rather an epic screenshot of the sunrise with all engines burning too, but sadly the screenshots aren't saving atm and i'll have to restart ksp.

More importantly, my flying has gotten better with practice.   This second flight brings us a LOT more residual fuel.

Hope these guys remembered their toothbrushes and returned their library books...

[AeroGav]

ok the good news is i still had 1451 fuel remaining when i reached Duna's surface. 

Unfortunately i cannot land this in one piece.   

The landing gear are attached to the outer engine nacelle, which are attached to a wing , which are attached to the mid nacelle, then the inner nacelle, then the mk3 cargo bay.   One of those joints fails and the ship starts coming apart.

Try attaching at least one of the gears directly to the cargo bay then offsetting outward, if you must attach to the nacelle then use struts (real and auto) to strengthen.

Also i think you need vernier thrusters in the floor of the cargo bay, so you can translate downward to act like lift thrusters.

Stalling speed on Duna is about 75 m/s and the terrain is hilly.   Landing above 100m/s will mean hitting the hilly ground too hard and breaking up.

Between 75 and 100, it has too little lift to quickly arrest a descent rate or if a hillock suddenly comes up.  That is where vernier thrusters come in.

On my Duna landings,  i use the wings to maintain a steady descent rate of about 5m/s, and thrust down with the verniers to fine tune it and flare for touchdown.

I my experience canard designs always seem to land a bit slower than ones with elevons because the canards are lifting instead of pushing down at very low speed, but it's probably not the main factor, so don't change your design over it.

 

Edited December 9, 2016 by AeroGav

[jsisidore]

That is a close call... What happened to the docking port shell btw?

2 hours ago, AeroGav said:

No i was suggesting you build the nacelles like this

1. nose cone

2.  mk1 liquid fuel fuselage 

3. a few more mk1 liquid fuel fuselages

4. mk1 to mk2 short adapter

5. mk2 bicoupler

6. pair of engines

 

Makes sense I will try it out.

[AeroGav]

7 minutes ago, jsisidore said:

That is a close call... What happened to the docking port shell btw?

 

It staged off - 1st stage is your rapiers, 2nd nukes, 3rd is the aeroshell right?

[jsisidore]

1 hour ago, AeroGav said:

It staged off - 1st stage is your rapiers, 2nd nukes, 3rd is the aeroshell right?

Right! Note there are no cones on nukes.

[jsisidore]

1 hour ago, AeroGav said:

ok the good news is i still had 1451 fuel remaining when i reached Duna's surface. 

Unfortunately i cannot land this in one piece.   

The landing gear are attached to the outer engine nacelle, which are attached to a wing , which are attached to the mid nacelle, then the inner nacelle, then the mk3 cargo bay.   One of those joints fails and the ship starts coming apart.

Try attaching at least one of the gears directly to the cargo bay then offsetting outward, if you must attach to the nacelle then use struts (real and auto) to strengthen.

Also i think you need vernier thrusters in the floor of the cargo bay, so you can translate downward to act like lift thrusters.

Stalling speed on Duna is about 75 m/s and the terrain is hilly.   Landing above 100m/s will mean hitting the hilly ground too hard and breaking up.

Between 75 and 100, it has too little lift to quickly arrest a descent rate or if a hillock suddenly comes up.  That is where vernier thrusters come in.

On my Duna landings,  i use the wings to maintain a steady descent rate of about 5m/s, and thrust down with the verniers to fine tune it and flare for touchdown.

I my experience canard designs always seem to land a bit slower than ones with elevons because the canards are lifting instead of pushing down at very low speed, but it's probably not the main factor, so don't change your design over it.

That is a lot of fuel... That means I can use some of it for oxidizer to fuel vernor engines. How much thrust do I need btw? Some parachutes maybe? Will cause a bit of drag though.

[AeroGav]

Just now, jsisidore said:

That is a lot of fuel... That means I can use some of it for oxidizer to fuel vernor engines. How much thrust do I need btw? Some parachutes maybe? Will cause a bit of drag though.

I'd say you need about 8 Vernors on a ship of this weight, 200 units of oxidizer maybe?  You can mount the vernors in the floor of the cargo bay and offset them upwards so they're slightly clipping into the floor. If you've done it right, they won't fire when the doors are shut, but will when open.

  This means the game thinks they are shielded by cargo door and are will cause no drag when bay is shut, but when you get to duna can open the bay (even if only by a small amount) and be able to use the vernors.   If you can fit any more reaction wheels in there, that would also be good.

Flying on Duna feels wierd, everything responds in slow motion.    The low gravity partially compensates for the thin atmosphere, but this can catch you out on landing.   You think there is more lift available then there really is because of the low grav.   You're in a steady 10 m/s descent  near the ground, and you pitch up to flare for landing, but it takes ages for the descent rate to reduce, because there is less lift and gravity but just as much inertia as on kerbin.   Same with control forces - pitching and rolling are slower because the controls make less force than on kerbin.

You should definitely practice some landings on duna with the cheat menu before flying a real mission.   Unload most of your fuel in SPH, then ALT F12, cheat, set orbit.   See what you're up against.

 

[jsisidore]

I scrapped the strakes.

On 12/8/2016 at 8:09 PM, AeroGav said:

and the rear elevons down ever so slightly, until it flies with a very small nose up angle with SAS off and no control input.

That solved the problem. Should've guessed.

Redesign: http://pastebin.com/3MmxbM5L

Fixed: http://pastebin.com/keayxA12

The D/L ratio went down all of a sudden.

edit: added strakes back, ratio went up lol. 

Edited December 10, 2016 by jsisidore

[jsisidore]

Hmm vernor engines do not work, yet there is oxidizer in the fuel tanks. Could be a bug.

Oh I had to press space to activate them, confusing.