Help me understand why my design is unstable?

[IncompetentSpacer]

I returned to this game after many, many years of hiatus. Last time I played was right after careers were added. So coming back to the game after this long time, it is like playing new game all over again. I have managed to land on Mun and Minmus several times and so on but some elementary stuff eludes me.

I am trying to simplify my design and the R-7 version is 42 tons lighter and can achieve pretty much same as previous iterations at same cost (I am playing career mode). This is also my first design with Nerv motors in upper stage.

The problem is not to get this in orbit... The problem is that due to the flight path I end up using first stage before reaching orbit and have to use upper stage for that final push. Simply put the rocket starts to flip whenever I try to angle it over 2-4 % towards the east before I reach 20000 meters. Which means that I am flying stright up too long and have too low horizontal speed before I make the sharp turn. I want to understand why this happens.

See below for the design:

https://imgur.com/a/aTrZbKF

[Streetwind]

You're violating the golden rule of rocketry: "heavy stuff at the front, draggy stuff at the back".

Specifically, you have a high amount of parts and assorted greebling sticking out your upper stage. The blunt, uncapped docking port is an absolute no-go - put a nosecone on that, you can eject it once you leave the atmosphere. Additionally, there's that gap in the rocket where the stages connect. I cannot quite understand what is going on there, but depending on how it's set up, it's a major candidate for crippling drag.

I see you have the center of lift marker activated. Unfortunately, this marker does not actually show the center of aerodynamic pressure that is relevant for rocket stability. The game does not feature a way to display this to you. You'll have to work with educated guesses based on the CoM location, and test flights to verify your guesses. In your case, those test flights tell you "nope, you guessed wrong".

You can use a fuel flow trick to force the CoM a little further up in flight, by configuring the fuel tanks on your first stage so that they have decreasing priority from the bottom up. That is, if your fuel tank priority is, I dunno, 30 by default, then leave the bottom-most fuel tank at 30, set the one directly above that to 29, the next one to 28, and so on. This will make the bottom-most tanks drain first, leaving the others completely full. As a result, the average Center of Mass ends up higher than if all tanks were draining equally at the same time.

This is however only a band-aid. It can make the difference between slightly unstable and barely stable. It will not fix an unflyable rocket on its own.

Edited June 7, 2020 by Streetwind

[AHHans]

1 hour ago, Streetwind said:

Specifically, you have a high amount of parts and assorted greebling sticking out your upper stage. The blunt, uncapped docking port is an absolute no-go - put a nosecone on that, you can eject it once you leave the atmosphere. Additionally, there's that gap in the rocket where the stages connect. I cannot quite understand what is going on there, but depending on how it's set up, it's a major candidate for crippling drag.

I think that's the main issue.

In addition it also might be that the thin connection there has too little mechanical strength and "bends" under stress. In that case SAS can actually make things worse.

[IncompetentSpacer]

What did the trick was to to prioritize the flow of fuel as Streetwind suggested.  I tried that suggestion first because I am very skeptical that drag is an issue at 20 000 meters and makes rocket flip over (but see how that may be an issue looking at fuel consumption and maneuverability in thicker atmosphere below 14 000 m). Now the rocket can do Split S if needed at any height - figuratively speaking, she is no nimble beast below 7000 but rather maneuverable outside atmosphere.

I did test the drag with shielded docking port and it improved the drag by 30-40 KN at 10 000 meters (worth mentioning, thanks for the tip). Improving the fairing for Nerv engine had minimal effect what so ever (that is what you are seeing on the middle of rocket - default fairing is even worse). The rocket has no shaking what so ever, so I am leaving that one be. Removing antenna removed another drag factor but difference is 20 m/s gained at 10 000 meter. Total drag at 10 000 meters stands now 351 kn*.

 

*Using shell to encase entire upper stage reduces drag to 127 KN but increases weight and fuel consumption leaving me with less fuel at 10 000 km. So drag is not everything in this game...

 

 

[jimmymcgoochie]

That’s a lot of drag, and a lot of it is unnecessary. Get rid of the first stage Vernor/RCS and the monopropellant tank, then use the weight saved by that to put a fairing over the payload which will reduce drag considerably and then pop the fairing at about 25km as by then it is no longer required.

Edited June 8, 2020 by jimmymcgoochie

[IncompetentSpacer]

4 hours ago, jimmymcgoochie said:

That’s a lot of drag, and a lot of it is unnecessary. Get rid of the first stage Vernor/RCS and the monopropellant tank, then use the weight saved by that to put a fairing over the payload which will reduce drag considerably and then pop the fairing at about 25km as by then it is no longer required.

I did wrote the longer response with test results but it all went away.

Basically doing as you suggested but replacing RCS with dialed down vernor gave best result with 80 MS left in fuel tanks even in crap orbit because I messed up in final burn. I think I had 79+ KN of total drag at 10 KM but I am not sure. 

Same weight reductions but without airstream fairings left me 20 MS of dv in fuel tanks in a standard circular orbit at 78 KM.

All test did not use perfect flight path but rather stright up to 10 km and then turn to east at 45 degree climb (so I could recreate same situation every time.

So the take here is - weight distribution through fuel use gives stability, airstream shell is fine but in my case only had an effect when itss effect could be offset by reduction in weight elsewhere.

[jimmymcgoochie]

2 hours ago, IncompetentSpacer said:

I did wrote the longer response with test results but it all went away.

Basically doing as you suggested but replacing RCS with dialed down vernor gave best result with 80 MS left in fuel tanks even in crap orbit because I messed up in final burn. I think I had 79+ KN of total drag at 10 KM but I am not sure. 

Same weight reductions but without airstream fairings left me 20 MS of dv in fuel tanks in a standard circular orbit at 78 KM.

All test did not use perfect flight path but rather stright up to 10 km and then turn to east at 45 degree climb (so I could recreate same situation every time.

So the take here is - weight distribution through fuel use gives stability, airstream shell is fine but in my case only had an effect when itss effect could be offset by reduction in weight elsewhere.

You don’t need ANY RCS on a first stage booster (unless you want to try a powered recovery, that is, and even then reaction wheels are probably sufficient). 
 

And you DO NOT under any circumstances want to do a split-S in a rocket! A split-S is rolling inverted then doing a downward half-loop to fly on the other direction; in a plane that is, doing it in a rocket is generally known as ‘tumbling end over end and exploding’. Perhaps you meant a gravity turn?

[Spricigo]

1 hour ago, jimmymcgoochie said:

And you DO NOT under any circumstances want to do a split-S in a rocket! A split-S is rolling inverted then doing a downward half-loop to fly on the other direction; in a plane that is, doing it in a rocket is generally known as ‘tumbling end over end and exploding’. Perhaps you meant a gravity turn?

Quoting for emphasis. If your Launch vehicle is capable of a nudge in the right direction just after the take off that is all the maneuverability you need. If that is not enough to reach orbit you have a stability problem, a piloting problem or both.

 

[IncompetentSpacer]

5 hours ago, jimmymcgoochie said:

You don’t need ANY RCS on a first stage booster (unless you want to try a powered recovery, that is, and even then reaction wheels are probably sufficient). 
 

And you DO NOT under any circumstances want to do a split-S in a rocket! A split-S is rolling inverted then doing a downward half-loop to fly on the other direction; in a plane that is, doing it in a rocket is generally known as ‘tumbling end over end and exploding’. Perhaps you meant a gravity turn?

While for "some reason" nobody thought about buzzing Christopher C. Kraft Jr. Mission Control Center with a Saturn rocket raging 50 meters over ground I assure you that would be a sight to see.

That said, I was speaking figuratively, I haven't really tried to do that (but now I thinking about making variant B just to see it done...)  - I just turned rocket here and there to see if it would vibrate too much (follow the line on NAV ball) or flip over just because I turned too quickly or too much.

I kept 4 verniers for when outside the atmosphere but wasn't really needed so I deleted those. Also winglets at the bottom are gone now.