Shuttle instability

[SlabGizor117]

So I'm working on building a classic space shuttle but I have ridiculous stability problems.  It drifts off of exact east for no reason and any attempt to correct it induces roll and it's all downhill from there because when the SAS system take priority away from keeping the shuttle from immediately nosing to the horizon(thrust vector isn't quite worked out yet) to deal with my command input then ever roll and yaw just drops the nose further towards the ground.  

I have a video uploading to YouTube now that I'll edit in to show more, but in the meantime I have two questions:

1. I think part of it is that my yaw controls are off center.  The vector engines are obviously way out of whack for yaw, and the tailfin is super high above the CoM.  Should I put hidden control surfaces in for yaw?  What else could I do for that?

2. How can I fix my thrust vector so that I'm not constantly wanting nose down, but then keep stable after I lose my SRBs?  How can I have one without the other??

I'll edit when I finish uploading the video, it's taking a while.

 

EDIT:  Video!

 

I'm not sure if you can see my frustration at the end there, but Suffice it to say Val preferred suicide to landing the shuttle safely.  Watch the navball and how everything snowballs as I try to correct stuff.  This is without payload by the way.  Lord help me if I ever tried that.

Edited January 21, 2017 by SlabGizor117

[GoSlash27]

22 minutes ago, SlabGizor117 said:

So I'm working on building a classic space shuttle but I have ridiculous stability problems.  It drifts off of exact east for no reason and any attempt to correct it induces roll and it's all downhill from there because when the SAS system take priority away from keeping the shuttle from immediately nosing to the horizon(thrust vector isn't quite worked out yet) to deal with my command input then ever roll and yaw just drops the nose further towards the ground.  

I have a video uploading to YouTube now that I'll edit in to show more, but in the meantime I have two questions:

1. I think part of it is that my yaw controls are off center.  The vector engines are obviously way out of whack for yaw, and the tailfin is super high above the CoM.  Should I put hidden control surfaces in for yaw?  What else could I do for that?

2. How can I fix my thrust vector so that I'm not constantly wanting nose down, but then keep stable after I lose my SRBs?  How can I have one without the other??

I'll edit when I finish uploading the video, it's taking a while.

 

EDIT:  Video!

 

I'm not sure if you can see my frustration at the end there, but Suffice it to say Val preferred suicide to landing the shuttle safely.  Watch the navball and how everything snowballs as I try to correct stuff.  This is without payload by the way.  Lord help me if I ever tried that.

SlabGizor,

 That's actually normal behavior from a shuttle stack. The SSMEs are so far off the centerline that vectoring them to control yaw imparts roll and roll imparts yaw. I recommending turning down their vectoring range and practicing how to fly it that way. You need to learn to use your rudders to impart roll and your stick to correct yaw. It's tricky.

Best,
-Slashy

 

[SlabGizor117]

So what about the tailfin being so far off the CoM?  Does that cause problems when yawing?

Also if I turn down my gimbal range that'll give me big problems when the CoT and CoM shift as the ET runs out of fuel and I drop my SRBs.  What am I gonna do if I have to balance the CoT for liftoff and SRB seperation, and then I add the CoM shifts of a payload??

Edited January 21, 2017 by SlabGizor117

[GoSlash27]

8 minutes ago, SlabGizor117 said:

So what about the tailfin being so far off the CoM?  Does that cause problems when yawing?

Not that I've noticed, but I never use active rudders on my planes. I don't imagine that having your rudder active is doing you any favors, though. It'll give you the same roll coupling as the engines.

Best,
-Slashy

*edit* Nice lookin' shuttle, by the way

Edited January 21, 2017 by GoSlash27

[SlabGizor117]

Huh, interesting point.  I never thought about deactivating it.  What if I need it for maneuvering during reentry and landing?

Also yeah how am I gonna balance my CoT to account for liftoff configuration, booster seperation, and payloads?

Edited January 21, 2017 by SlabGizor117

[steve_v]

30 minutes ago, SlabGizor117 said:

Also if I turn down my gimbal range that'll give me big problems when the CoT and CoM shift as the ET runs out of fuel and I drop my SRBs

Rather than reducing gimbal range, what about just disabling yaw and roll on the vectors? You should have enough yaw/roll authority from your aero surfaces, at least for a while...

[SlabGizor117]

I don't think you can disable individual gimbal directions on engines.

[GoSlash27]

1 hour ago, SlabGizor117 said:

Huh, interesting point.  I never thought about deactivating it.  What if I need it for maneuvering during reentry and landing?

Also yeah how am I gonna balance my CoT to account for liftoff configuration, booster seperation, and payloads?

SlabGizor,

 You won't need it for reentry or landing. Just having it back there is enough to keep your turns coordinated. Plus, having it active with SAS engaged tends to make the plane crab during turns. You dip a wing and it attempts to maintain your previous heading by yawing away from the turn. Makes everything a lot sloppier. Try building a simple plane with the rudder deactivated and you'll see what I mean.

 As for how to keep the CoT centered, that's the real trick to shuttles. Basically, you do this:

I start with the bare orbiter, making sure that the OMS engines are thrusting through the CoG.

Then I add the tank. I set up the tank to drain in such a way that the CoG descends along a line towards the main engines. Then I point the engines along that line. When I add the SRBs, they are offset slightly away from the shuttle so that the total thrust vector is through the CoG.

 It will take a little tweaking and trial runs to get it all running smoothly, but if you do it right you never have a thrust misalignment during the mission.

Shuttles are tricky

-Slashy

[SlabGizor117]

33 minutes ago, GoSlash27 said:

SlabGizor,

 You won't need it for reentry or landing. Just having it back there is enough to keep your turns coordinated. Plus, having it active with SAS engaged tends to make the plane crab during turns. You dip a wing and it attempts to maintain your previous heading by yawing away from the turn. Makes everything a lot sloppier. Try building a simple plane with the rudder deactivated and you'll see what I mean.

 As for how to keep the CoT centered, that's the real trick to shuttles. Basically, you do this:

I start with the bare orbiter, making sure that the OMS engines are thrusting through the CoG.

Then I add the tank. I set up the tank to drain in such a way that the CoG descends along a line towards the main engines. Then I point the engines along that line. When I add the SRBs, they are offset slightly away from the shuttle so that the total thrust vector is through the CoG.

 It will take a little tweaking and trial runs to get it all running smoothly, but if you do it right you never have a thrust misalignment during the mission.

Shuttles are tricky

-Slashy

Wait...  Correct me if I'm wrong but if the SRBs are lined up with the CoM of the External Tank and the Vectors are lined up with the CoM of the shuttle by itself then everything will work itself out, correct?

Also I don't like having to angle the shuttle.  Can I do it just veritcally?  I don't mind if the Vectors push me backwards till I nose back enough.

EDIT:  One more question.  What's causing that yaw drift?  I don't have any symmetry that I know of causing one sided drag.

Edited January 21, 2017 by SlabGizor117

[GoSlash27]

5 minutes ago, SlabGizor117 said:

Wait...  Correct me if I'm wrong but if the SRBs are lined up with the CoM of the External Tank and the Vectors are lined up with the CoM of the shuttle by itself then everything will work itself out, correct?

SlabGizor,

 No, afraid not. The relative t/w is static for the orbiter by itself, but changes wildly for the booster and tank. Plus, you spend most of the time during the launch without the SRBs. That's why controlling the movement of the CoG and aligning to it's path is so critical.

8 minutes ago, SlabGizor117 said:

Also I don't like having to angle the shuttle.  Can I do it just veritcally?  I don't mind if the Vectors push me backwards till I nose back enough.

 Yeah, you can do that. I just find it's easier to launch consistently with the center of combined thrust directly beneath the CoG.

Best,
-Slashy

[AeroGav]

One thing that may be hindering us is that the Vector is actually nothing like the real RS-25 (SSME).  It is much heavier and more powerful than the real shuttle engine ; the real shuttle actually had 80% of takeoff thrust from the SRB.

I did try building a stock shuttle with 3 swivels on the orbiter (if you offset them forward so only the bell protrudes from the mk3 engine mount, they look a lot more like shuttle engines) and 2 kickback SRBs.   This gave us a CG that was further forward and it would have been aerodynamically stable during re-entry and landing.

Unfortunately it would still have taken a lot of work to make it fly properly on the way up and I couldn't be bothered to stay on it.  

On liftoff, CoM is well below the Orbiter, over time this moves upwards as the tank and SRBs get lighter, which increases the pitch up moment from the SRB thrust as their distance from CoM increases.    At the same time,  CoM is moving aft as the tank empties, which reduces the aerodynamic ability to fight the pitch up, as does increasing altitude.

In my sig. is a link to my KerbalX craft - I've a few "nearly shuttles" in there, they all have four SRBs mounted symmetrically around the Orbiter and internal fuel instead of ET, and have much lighter main engines than the Vector - these are much easier to design and fly, if you're willing to stomach the compromise.  (smaller cargo bay as a result of the internal fuel)

 

[SlabGizor117]

Ok.  Well I brought the gimbal range down to 70 and that helped and I turned off the tailfin yaw but I still have that yaw drift.  I don't understand why it's doing that.  And it turns everything into a big mess to try and correct it because yaw on such a weird vehicle as you know is just an absolute nightmare.  I did realize however that when possible it helps alot to set it to align to prograde.  I wish there was an autopilot to control roll and yaw so you would be left to only deal with pitch.

[GoSlash27]

21 minutes ago, SlabGizor117 said:

I still have that yaw drift.  I don't understand why it's doing that.  And it turns everything into a big mess to try and correct it because yaw on such a weird vehicle as you know is just an absolute nightmare.

SlabGizor,

 There will always be a little misalignment in the early going because the airspeed is so low. The yaw drift is a result of the engines trying to correct for roll. You have to train yourself to not respond to the yaw misalignment with yaw, but rather roll. Likewise, if your wings aren't level you need to correct that with yaw.

 The plane wants to respond in such a way that the prograde vector is below the nose and centered. If you kick in a little left yaw, it will immediately roll left to try to get the prograde back below the nose, and vice versa. If you roll left, it will create a yaw left response. It's a weird way to fly, and takes some practice, patience, and finesse to get the hang of.

I find that most of my work during launch is done using the yaw vectoring of the engines, with just a little roll to keep the turns coordinated.

HTHs,
-Slashy

[steve_v]

2 hours ago, SlabGizor117 said:

I don't think you can disable individual gimbal directions on engines.

Sure you can:

[SlabGizor117]

1 hour ago, steve_v said:

Sure you can:

Oh u rite.  Maybe that'll fix my yaw drift problem then, if Slashy is right about what's causing it.  Noice!

[SlabGizor117]

Ok, yet another problem, the shuttle still yaws after disabling yaw and roll gimbal, meaning I basically have no yaw control on the shuttle, other than RCS.

EDIT:  So I chose after messing up my launch to land and to my surprise when I landed it still had that left hand yaw.  Looking at the Aero overlay(and taking a bunch of pictures on steam), the tailfin was producing lots of lift, which is really weird.  Glitch?

 

EDIT 2:  Tried it with that really tiny little basic fin to see what happens and it still has a yaw.  Testing with no vertical stabilizer now.  Also launched with no SAS and I dunno if it's the same effect but by the time it nosed into the runway it only had a slight left roll.  Probably the same thing happening.

EDIT 3:  Still yaws even with no stabilizer. WHYYYYYY MEEE

EDIT 4:  So the best I can tell is that when the SAS pitches down to account for the pitch up of the uneven thrust it induced a slight roll for some reason which brings it off the 90 and as it goes too far the other way and brings it back it rolls back the other way so that it keeps going north.  Then gravity takes over and pulls the nose down.  Launching without SAS does the same thing when you manually pitch down.  Launching and then letting the shuttle nose back to the ground ends with the shuttle at a slight left roll as well.  I don't understand what's causing this, I even disabled roll on every control surface.

Edited January 21, 2017 by SlabGizor117

[seaces]

On 1/21/2017 at 11:01 PM, SlabGizor117 said:

Ok, yet another problem, the shuttle still yaws after disabling yaw and roll gimbal, meaning I basically have no yaw control on the shuttle, other than RCS.

EDIT:  So I chose after messing up my launch to land and to my surprise when I landed it still had that left hand yaw.  Looking at the Aero overlay(and taking a bunch of pictures on steam), the tailfin was producing lots of lift, which is really weird.  Glitch?

 

EDIT 2:  Tried it with that really tiny little basic fin to see what happens and it still has a yaw.  Testing with no vertical stabilizer now.  Also launched with no SAS and I dunno if it's the same effect but by the time it nosed into the runway it only had a slight left roll.  Probably the same thing happening.

EDIT 3:  Still yaws even with no stabilizer. WHYYYYYY MEEE

EDIT 4:  So the best I can tell is that when the SAS pitches down to account for the pitch up of the uneven thrust it induced a slight roll for some reason which brings it off the 90 and as it goes too far the other way and brings it back it rolls back the other way so that it keeps going north.  Then gravity takes over and pulls the nose down.  Launching without SAS does the same thing when you manually pitch down.  Launching and then letting the shuttle nose back to the ground ends with the shuttle at a slight left roll as well.  I don't understand what's causing this, I even disabled roll on every control surface.

This reminds me of my spaceplane that was pulling me to the right on takeoff. Turned out I wasn't using symmetry for one or more parts or something like that. My plane just did not have two sides balanced properly. So try that put your shuttle on its wheels on runway and if it yaws again mean that something is probably asymmetrical. In the end if you cant find the asymmetrical part I would just do the orbiter from the scratch.

[SlabGizor117]

 

1 hour ago, seaces said:

This reminds me of my spaceplane that was pulling me to the right on takeoff. Turned out I wasn't using symmetry for one or more parts or something like that. My plane just did not have two sides balanced properly. So try that put your shuttle on its wheels on runway and if it yaws again mean that something is probably asymmetrical. In the end if you cant find the asymmetrical part I would just do the orbiter from the scratch.

I looked everywhere and found nothing asymmetrical