Shuttle thrust direction problem.

[MedwedianPresident]

I have the following problem: as the thrust is uneven due to the system’s natural asymmetry, I had to tilt the SSME’s away from the tank/booster stack. That’s what they do in real life. The shuttle now flies “properly” without becoming unresponsive halfway through ascent like most of my earlier designs. However, at launch the shuttle does not fly directly upwards but rather sideways because the tilted Vectors generate perpendicular thrust. This means that I have to do an ugly correction maneuver in the first seconds of flight and lose a lot of fuel. I have thought of tilting the whole shuttle on the launchpad, but that would be quite unrealistic. I’ll have to see how it behaves when there is a payload in the cargo bay, maybe it could serve as some sort of a counterweight. But how can I negate this perpendicular thrust? I have already tilted the boosters themselves by 2-3 degrees. Maybe I could move them further “downward” on the tank to move the CoT and change the DoT? Or should I add additional, smaller boosters that fire for 20 seconds and keep my shuttle from moving in the wrong direction until it is tilted down enough for the thrust discrepancy to be unimportant?

Yes, I already tried limiting the thrust of the main engines, but this would give the whole thing a TWR under 1. FYI: The boosters are 1.25m ones from NovaPunch, scaled to 1.4m by TweakScale.

[DrLicor]

The first seconds aren't that important. It's not about the fuel, but about the delta-v. The first seconds the shuttle will for example burn 2 m/s²  while in the last seconds of the stage it reaches 100m/s². 
Try to angle the shuttle a degree backwards so it's horizontal velocity is reduced a bit. But if you ask me, that extra booster is a bit of overkill.

[Helmetman]

It's very hard without a actual picture. Preferably one from the side with the CoM and CoT?

Dragging the SRB's further down certainly helps.
A few other things may help.

Moving the SRB's slightly away from the CoM, By this I mean moving away from the shuttle.

Re-angle the main engines to get balance right again. Just a little bit without making it look bad. May help to counter the effect the main engines provide with the SRB's itself.
But don't angle them.

Move the shuttle as close to the surface of the external tank. Maybe use the 'move' tool and make it attach very closely to the external tank surface. Also try to drag the shuttle itself as far down. Or re design the engine stack so that it sticks further down.

Also pay attention to the relative thrust of the SRB's against the main engines.
The shuttle engines needs to carry it's weight and the external tank once jetissoned but no more. That while you want the SRB's to deliver most of the thrust at launch. While it depends on the design a 'typical' shuttle at launch would want about 0.90 or 1.00 TWR from the SRB's and the rest (0.5+) from the shuttle main engines.

Going overboard on shuttle engine thrust will severely add to the sideways motion. Limit the thrust on the shuttle engines if they produce to much thrust comparatively.

You can perhaps save a degree of rotation from the shuttle engines off the center of Mass if you have sufficient reactions wheels to counter any unbalance on the way up.

Here is a example of a dummy shuttle where I applied all of the above. The thrust on the vectors is limited to 24 for a desirable thrust on the main engines. Remember that the tank representing the shuttle is just to mimic weight. The real shuttle can be longer and I'm sure one could it get stable with the engine stack and the weight further to the top.
It also lacks the vertical stabilizing effect the wings produce and she flew fine after jetissoning.

 

Sorry for Hyperedit window
 

It's not none at all, but very little of the effect you described.

Edited October 13, 2017 by Helmetman

[Spricigo]

Something to keep in mind:

KSP gives you all the pieces you need to make a vessel that looks like the real life shuttle but the vessel that looks like the real thing probably don't fly like the real thing (used to).

PS: for craft specific question, craft specific pictures. 

 

[Palaceviking]

The shuttle always took off ugly, that's the nature of it

 

[Laie]

I'm not sure what you're trying to get at with your bigger question, especially teh course correction upon launch, but this one caught my attention:

13 hours ago, MedwedianPresident said:

how can I negate this perpendicular thrust?

And the answer is, you don't. You just roll with it. The real shuttle didn't fly in the direction it''s cockpit was facing, and you should do likewise. During ascent that's not difficult, merely eyeballing it and pointing your vessel a little above or below prograde will get you good-enough results.

Once you are in space and need to do precision maneuvers, you want your thrust lined up with the CoM as well as possible. You may need a dedicated Orbital Maneuvering System for the purpose. Hovever you do it: if your thrusters are angled relative to the cockpit, add a small probe core or docking port somewhere that's tilted at just the same angle, and use the "control from here" function so that your idea of "prograde" lines up with the actual direction of thrust.

Example. The Nuke is badly angled, but there's a small probe core between the NCS adapter and small nose cone. That's how the vessel can fly straight in space.

[Geschosskopf]

16 hours ago, MedwedianPresident said:

I have the following problem: as the thrust is uneven due to the system’s natural asymmetry, I had to tilt the SSME’s away from the tank/booster stack. That’s what they do in real life. The shuttle now flies “properly” without becoming unresponsive halfway through ascent like most of my earlier designs. However, at launch the shuttle does not fly directly upwards but rather sideways because the tilted Vectors generate perpendicular thrust....

FWIW, the real shuttle also went kinda sideways at launch.  Or rather diagonally, pointing straight up but moving both vertically and horizontally.  You can see that happening in the 1st 2 launches of this video, where it's shown in side view.

I really remember being freaked out by the 1st shuttle launch.  I'd been used to watch Saturn launches and they went straight up very slowly.  Then the shuttle went up sideways very quickly

 

Edited October 14, 2017 by Geschosskopf

[Grand Ship Builder]

Make sure the vectors are pointing at the CoM. Just gimble a bit. It's natural for the shuttle to be upside down.