my space shuttle starts to make barrel rolls

[amateur astronaut]

so i built a NASA style spaceshuttle, and when i dont touch the controls, its flyes pretty well... but when i touch the pitch, its starts to tilt to the left.. and then make barrel rolls uncontrolably. also, when i reach 800 m/s, its tilts backward and starts to descend. I tried to eject down SRBs at 800 m/s but there were plenty of fuel on them.... what a waste...

Edited October 24, 2018 by amateur astronaut

[Xd the great]

1 hour ago, amateur astronaut said:

so i built a NASA style spaceshuttle, and when i dont touch the controls, its flyes pretty well... but when i touch the pitch, its starts to tilt to the left.. and then make barrel rolls uncontrolably. also, when i reach 800 m/s, its tilts backward and starts to descend. I tried to eject down SRBs at 800 m/s but there were plenty of fuel on them.... what a waste...

Rotate the space shuttle so that its back faces east. Aka where you want to go. Auto gravity turn, no keimg required.

Problem solved. 

[amateur astronaut]

It is facing where i wanna go, the problem is that when i use pitch, its also bit tilts to the left, and later it's become uncontrollable (I guess because i dont used symmetry tool to take on the vector engines, because if i would, it would not allow me the thrust vectoring) . but when i dont touch anything, it flies nicely until it reaches 800 m/s, (in the 4th picture you can clearly see that the navball shows that its flies straight, but the vectoring makes it go forward

you can see on the navball that i tried to tilt it to counteract the forward motion of the vectoring, and how it tilted to left (for no reason!)

[bonyetty]

It would appear like the direction of thrust is not pointing through the center of mass. It helps to do so although this is difficult with NASA style shuttles, as their COM changes during assent. I find that having the control point also pointing through the COM prevents the roll and yaw inputs from yourself and the SAS from being misaligned. When trying to yaw with a misaligned control point some roll will also be applied to the craft, and vise versa. An extra docking port or probe core attached pointing through the COM can reduce this. NASA style shuttles are tough, good luck and hope this helps. 

[amateur astronaut]

9 minutes ago, bonyetty said:

It would appear like the direction of thrust is not pointing through the center of mass. It helps to do so although this is difficult with NASA style shuttles, as their COM changes during assent. I find that having the control point also pointing through the COM prevents the roll and yaw inputs from yourself and the SAS from being misaligned. When trying to yaw with a misaligned control point some roll will also be applied to the craft, and vise versa. An extra docking port or probe core attached pointing through the COM can reduce this. NASA style shuttles are tough, good luck and hope this helps. 

do yo mean that the vector of the CoT should point to the direction of the CoM? (i checked it, and its just barely points to the CoM, not exactly to the center)

update: i put in some 2.5m fuel tanks to the cargo bay so now the CoT points towards the CoM. but now its too heavy and tilts to its back

Edited October 24, 2018 by amateur astronaut

[ARS]

An helpful tip: reduce gimbal range of all vector engines by 50%, (I find that the gimbal range of vectors is way too large, which, coupled with it's high thrust, is prone to causing loss of control) that way whenever there's any input, the engines doesn't over-correct your craft's maneuver. I did this in all of my shuttles

About the design, assuming an optimal design, as long as the craft is oriented on the right direction on launchpad, you don't have to input any directional control whatsoever during the ascent phase. Let the mass of the shuttle set the ascent profile, tilting automatically for gravity turn, with your supervision for staging. Only assume direct control when it's relatively safe for reaction wheel/ RCS to reorientate the craft without atmospheric disturbance for orbit circularization

[bonyetty]

As I said shuttles are very difficult. Also note the SRBs have a lot of thrust and so they burn a lot of fuel very quickly and thus lose mass very quickly. With the tools and information given us in KSP a lot of trial and error is the only way to get a NASA like shuttle to fly. It does give one a new appreciation of the aerospace engineering required for the real one however. 

[amateur astronaut]

30 minutes ago, ARS said:

An helpful tip: reduce gimbal range of all vector engines by 50%, (I find that the gimbal range of vectors is way too large, which, coupled with it's high thrust, is prone to causing loss of control) that way whenever there's any input, the engines doesn't over-correct your craft's maneuver. I did this in all of my shuttles

About the design, assuming an optimal design, as long as the craft is oriented on the right direction on launchpad, you don't have to input any directional control whatsoever during the ascent phase. Let the mass of the shuttle set the ascent profile, tilting automatically for gravity turn, with your supervision for staging. Only assume direct control when it's relatively safe for reaction wheel/ RCS to reorientate the craft without atmospheric disturbance for orbit circularization

well... I've done it but now something strange going on...

at 362 m/s its do the same barrel rolling, without touching the controls

[ARS]

Try putting toggling gimbal lock on action group. Stabilize the craft using only reaction wheel during ascent, with vector gimbal capability turned off. Turn it on later once you left the atmosphere. Building NASA style space shuttle is hard, second only to SSTO, it's a quite common topic of discussion in this forum

[Harry Rhodan]

I'd say the usual suspect for unwanted rolling is insufficiently strutted SRBs. Once they bend ever so slightly to one direction they produce a lot of torque.

[Geonovast]

I found my shuttle launches got more stable when would disable the control surfaces on the shuttle itself.

[MaianTrey]

Here's a post I've made in the past when someone was asking about making a stable shuttle. I've revamped my shuttle since the pictures below, but still used the same concept. Another option to solve your SRBs becoming too powerful is to change the fuel levels to be different in each pair, so they burn out differently and you'll get lowered thrust as you climb higher. You can also leave the fuel levels but change the thrust limiter, either way can get it done. It takes some experimenting with it.

If you can get the thrust to be more stable, it will likely help your rolling problem. I don't use the wing control surfaces on launch, and I never activate the Tail-Wing surface for use in control. If you right-click your control surfaces, you can then click 'Show Actuation Toggles' and then turn Yaw off for all of them, and turn roll off for the inner flaps (outer flaps should be roll). With my method of burning out the SRBs at tiered timing, I've found I can just use the gimbal of the SSME's to control my direction.

Quote

The best thing to do, which is how I built my shuttle that gets to orbit very easily, is to angle the 3 engines on the shuttle to move through the center of mass with only the external tank attached. I've posted a couple of screenshot examples below. That purple line at the bottom is the 'Center of Thrust'. As long as that points through your 'Center of Mass', it will fly without flipping (for example, I have to hold a 15° angle above prograde to actually be vectoring through prograde). I don't have it shown with screenshots, but the little sidepod engines (Orbital Maneuvering System - OMS) angle through the shuttle's Center of Mass without the External Tank and Boosters.

You may have to adjust how high or low the external tank and/or boosters sit to get the thrust correct. If possible, I'd recommend a mod used in the editor, called RCS Build Aid. It adds extra markers along with the Center of Thrust, Mass, and Lift markers already there so you can see rotation that your engines will cause. It also allows you to see both CoM with full tanks and with dry tanks, and tells you where the average is so you can base it on that.

 

 

Edited October 27, 2018 by MaianTrey