Another launch; much faster speed.
This one was a much more direct route; all horizontal velocity was reduced to 0, whereas the previous had a small horizontal velocity (50 m/s or so). I would advise against any sort of powered rotation or a slowing of the rotation the craft gains, as there seems to be a problem which causes your rocket to shake itself apart, as shown with the instantaneous disintegration of the entire craft in the log. But at 4.3% the speed of light, it's almost to the point where it would be a viable interstellar speed.
Another interesting note, as it approached the very core of the sun, the velocity was under 1 million; the G force meter zoomed upwards, and the craft almost instantly accelerated to its speed of 14 million m/s. I think the rotation of the craft comes in part from gravity. As it's calculated independently for the parts, the extreme gradient near the center of the sun causes it to rotate the oblong craft as it pulls on the ends of the rocket at different rates.
The shaking of the ship components seems to continue even after the ship shakes itself apart, I think it may have something to do with floating point precision at such high velocities; though I'm not certain.
Edit: Next successful flight went even faster!
61 million m/s, or approximately 20% the speed of light. Overall a pretty decent speed for a sub-luminal interstellar spacecraft.
Again, horizontal velocity slowed to near 0; this time I thrusted horizontally just as I was reaching the center (in the ~5 second window between G force rising drastically and hitting the center), though I'm not sure if that was the cause. Had a slight rotation of the craft going on, and it appears my first craft's massive rotation may have been due to an orbit around the center, as it had some horizontal velocity left. This craft was ripped apart upon its speed up at the center, with the bottom half of the rocket being torn off, with the remainder oscillating very quickly with an offset caused by the oscillation of about 1 meter. Another thing to note, as the craft does have its RCS still, the velocity seems to be unable to be changed through linear thrusting; likewise I noticed it wasn't decelerating as it left the sun. Again, I suspect some sort of floating point or other precision problem is at the root.
My particular design for these endeavors is a circular design like I previously mentioned, with hexagonal symmetry. Stages are concentric rings, with 1, 6, 6, 12, 12, 12 engines respectively. All liquid fuel; and all fuel comes from the outermost ring via fuel pumps. Rings are reinforced with struts, with fuel stacks of 3, 4, 4, 4, 4, 3 in the rings. Thus, the thrust is 49 engines in stage 1 drawing from 36 tanks, 37 in 2 drawing from 48, 25 from 48, 13 from 24, 7 from 24, 1 from 3, respectively. Unfortunately, this build is probably impossible on most machines, as it makes even the 8-core i-7 930, geforce gtx 480 alienware machine in the game dev lab run slowly, particularly in stage 1 with 49 engines firing.