Rocket returns to 90 degrees and reverses direction..?

[Arugela]

I fly for a bit and start to get around 10km in a very large rocket. I'm flying around 60 degrees with the indicator at about 45 degrees. After a while the ship uncontrollably starts going the other direction. I can't figure out why.. It eventually returns directly the opposite direction to 90 degrees then goes the other way and I loose all control. I can't seem to isolate it. Has anyone run into this problem. I'm not sure if it's an engine issue or some weird structural issue or what. I'm lost atm. The ship is very symmetrical. So I'm not really understanding where the force is coming from. I think the fuel is draining evenly. So I don't think that is it. And it has happened on several version of the craft. I have big boosters on and it doesn't do this without them. But I don't know if that is the boosters or the boosters bringing the issue out and it isn't seen because of other issues...

If I put SAS on retrograde I can kind of stop it but It's hard to figure out. If I do maintain direction it seems to slowly rotate while doing so... I'm pretty sure the engines are even though. Unless they are misplaced up farther. The only other thing I can think of is that they are so stacked up that they are blocking each other randomly and throwing off the general direction of thrust somehow.

Warning: 1100+ parts ship.

http://www./download/e3a6y81oaey541n/The_Orion_Mining_Spaceship_Basic.craft

Also, does anyone know which way an arrow should be pointing on a decoupler? Away from the main craft of towards it. I"m using TR-38-D.

Edit: I just noticed. As my fuel drops and my velocity increases, even though they are strutted, they seems to be moving wildly around. I had put struts to stop this. But this doesn't necessarily explain why it goes in one direction sometimes. Or does it? Maybe the boosters aren't stable on their own... I can't tell but some of the struts may have broken. It's night time while I'm testing it. But the problems primarily start while the struts are definitely still attached.

Edit2: I'm adding another layer of those struts between the boosters on the upper tanks to see if it helps. Hopefully it works, but I don't think it's the whole problem. Also, I'm doing a tear down and rebuilding the middle and engine sections soon. So I can get rid of lots of little structural issues this thing has and test out a new layout to connect the middle and back sections. Hopefully that make sit more sound structurally.

Edited October 15, 2015 by Arugela

[Slam_Jones]

You may want to try slapping some more Reaction Wheels on it, or place some stabilizing fins at the bottom of each stage. just make sure the bottom set of fins are the largest.

Reaction wheels will help it keep pointed the direction you want, and the fins would help make sure the rear end stays at the rear (kinda like the feathers on an arrow).

As far as I know (and I could be wrong) that issue is usually caused by the payload becoming heavier than the booster (due to shifting CoM from burning fuel), which makes it want to flip around.

[Higgs]

Pictures of it?

[Mastikator]

As the rocket burns fuel the upper tanks empty first.

This pushes the center of mass lower.

As the center of mass moves behind the center of drag it flips over. (you can do an experiment to check this out in your own home, throw a dart end-side first, it will flip over since the fins at the end have a high amount of drag and the tip is heavy)

You can put fins at the bottom, but that would increase both the mass and the overall drag thus reducing your total delta-v. Considering that it's a mining ship that may or may not be a problem.

Another solution is to disable the upper fuel tanks until the lower ones are almost empty.

[Slam_Jones]

1210 parts! Impressive! You must be running a supercomputer I'm surprised my .craft parser even got thru the whole thing

[Arugela]

I'll be adding more reaction wheels on the tear down. I'm actually burning from the center. The top has 1/3 of the total fuel and the boosters 2/3 of the total fuel on launch. They are very condensed. There are actaully 9 total tanks directly in the same position and 2 of those layered in a set of 4 boosters.

I'm testing again with extra struts. Hopefully it helps. But I need to get to faster speeds to see if it acts up again when the tanks are closer to empty. (Edit3: I'm hitting around 700m/s surface velocity and the boosters are trying to go all over the place. The struts linking them together seem to be holding, but I don't know if their internal struts not holding together individually or what. I'll have to figure out how to make them more stable possibly. Basically once the boosters are half way through their bottom tanks they start to flail...Edit4: On the upside 1200+ parts explosions are very entertaining! 8p)

It's a little laggy to fly. It's takes a while to do test flights sadly. I'm not running hardware that is that good. I'm surprised more people can't run this sort of thing. It's that or because I always build craft this bit I'm used to always flying in lots of lag...

Specs:

Fedora 22 x64(Linux)

Phenom II 1100t

8gigs 1600mhz ram (2x4gig)

256gb SSD main drive

Nvidia 560ti 1gb 384 core edition

750gb raid 10 (6x250gb drives. Random old drives and somewhat slow write because of differences.)

It could be the SSD helping...

Edited October 3, 2015 by Arugela

[Slam_Jones]

Usually much easier to work with screenshots taken during the day, but luckily I have GIMP and could change up brightness/etc to make it more visible.

I'd say stick on fins on the end of the two booster "legs," with at least 3x symmetry on each. After that, add at least 3 Reaction wheels to each leg, plus a similar amount to the top of the craft, just behind the cockpit.

I don't have too much experience with the level of clipping you've got going on with those fuel tanks, but that might be an issue as well. Make sure they're properly strutted, as oscillations there would only create a positive feedback loop (meaning, the movement causes more movement, which cause more, and more, etc in a loop til something breaks)

This should boost stability greatly, and keep it pointed mostly the way you want to go. Just remember not to be too aggressive on your turns, or it will cause too much drag and, potentially, some big issues. keep your heading no more than 5 degrees from your prograde marker.

Edited October 3, 2015 by Slam_Jones
Grammar, spelling

[ForScience6686]

I'm wandering about the aoa you're using. You should be facing prograde +- 5 degrees until out of the thick atmosphere. From your description, you are moving off prograde and causing the aero forces to hit the side of the rocket causing it to flip. You need to be above 25k before doing any aggressive maneuvers. Also make sure you are viewing surface velocity.

[Kryxal]

Definitely turning too hard, too soon, then not being able to follow prograde. TWR looks like it's probably a bit on the high side too. Try kicking over not quite 5 degrees at 100 m/s, hold till your velocity vector catches up, then FOLLOW that vector at least to 20-30 km as best you can. See if that's a better launch profile, adjust as needed.

[MarvinKitFox]

Its a simple case of too much drag up front.

That second container ship on the forward bulkhead simply does not counter the wind drag of the portside oil refinery.

.

.

At least not while you are sideslipping at 500 miles per hour.

Edited October 4, 2015 by MarvinKitFox

[Arugela]

Yea, it could be an issue of me flying wrong. I fly full power and angle. I don't do anything more complicated. Although there were a lot of structural stuff I have since solved on the new build. I have basically only flown space planes since I started playing. So I'm used to burning full speed to gain an apoapsis. I haven't really learned any of the subtleties of rocket flight. I'm flying it now trying to maintain a lower surface velocity and a lot of the problems are not there for now. Although I'm still early in a test flight.

[MarvinKitFox]

Ok, lets rephrase my comment.

Your rocket has the intrinsic smooth aerodynamic shape of... a feather duster.

That is being shoved backwards.

in a barrel of treacle.

You should expect to see some slight steering issues.

If you really *must* use that configuration, then at least have it flying directly prograde.

Your problem is that you are presenting a very draggy, long, and likely wobbly shape to severe crosswinds.

Either smooth the shape, OR add so much drag at the back that it overwhelms the turning forces, OR fly so close to prograde (surface speed!) that there is no crosswind.

[ForScience6686]

Try and stay under 300 m/s until above 10km, then full throttle. Begin your turn between 50-100m. At 5-10 degrees, then follow prograde up above 30k before making big maneuvers. Larger payloads I use a more gentle turn until I'm up out of the thick atmosphere. And I almost always use fairing now.

[Arugela]

Ok, lets rephrase my comment.

Your rocket has the intrinsic smooth aerodynamic shape of... a feather duster.

That is being shoved backwards.

in a barrel of treacle.

You should expect to see some slight steering issues.

If you really *must* use that configuration, then at least have it flying directly prograde.

Your problem is that you are presenting a very draggy, long, and likely wobbly shape to severe crosswinds.

Either smooth the shape, OR add so much drag at the back that it overwhelms the turning forces, OR fly so close to prograde (surface speed!) that there is no crosswind.

Yea, I've been adjusting by using the follow prograde button to stear. I tilt it 5 degrees at start then follow prograd and let it fall naturally. It seems to work really well. At 10k or more I change from surface to orbital velocity and it changes the prograde marker and it follows it. Kind of like an auto pilot. It seems to be working so far. I've given up stearing in atmosphere. And yes I must keep the shape. It is a replica ship of "The Orion" that is going to be in Star Citizen.

Here are the latest save files. I fixed some issues with it. Though I added more parts. I might need to thin out some of those struts.

https://www./folder/phyp96txcn9o8af,01tduvwnor4go4u/shared

It still needs coms, science, probes, and other cargo. But I can't get my computer to take more parts.

BTW, would adding nosecones or something on the forward/upward facing side of those containers reduce drag(Or reduce drag in a way meaningful to getting into orbit.). I'm not sure what the current calculations are for drag to tell how much stuff affects it. I don't know if it's good to take struts off or add something to see if it alleviates it and gets me into orbit faster.

I don't currently have any nosecones no this ship. the only thing similar are two CLAWS on the front of the ship to grab asteroids. And I don't think they help with that.... Or do they?

I forgot to add tail fins to the boosters. I'll have to add those next and see what they do.

And I'm aware the angle is steep. I oops. I put the pick up though to demonstrate the drag lines for the question about nosecones.

Edited October 4, 2015 by Arugela

[ForScience6686]

Be cautious switching to orbital so early, you are still in significant atmosphere at that point. Other than that congratulations

[Arugela]

If I put nosecones on some of the forward/upward facing surfaces would they work if I also sunk them in or do they have to be above other surfaces and be exposed?

[Kerbart]

Most people will tell you to start turning as quickly as possible for an efficient ascent. And they are right.

But this is not "most ships." And your "effeciency ratio" gets multiplied with a 1|0 factor indicating if you made it into orbit or not.

As others pointed out, the aerodynamics of this vessel are nearly not existent. So I'd shoot this one straight up in the sky and only dare to make a turn once you hit the 50km altitude.

Or consider using fairings. Or a different design. Or both.

[ForScience6686]

Most people will tell you to start turning as quickly as possible for an efficient ascent. And they are right.

But this is not "most ships." And your "effeciency ratio" gets multiplied with a 1|0 factor indicating if you made it into orbit or not.

As others pointed out, the aerodynamics of this vessel are nearly not existent. So I'd shoot this one straight up in the sky and only dare to make a turn once you hit the 50km altitude.

Or consider using fairings. Or a different design. Or both.

I agree with fairing but 50km is a bit long to wait to turn. Losing a lot of time to build horizontal speed which is more important than vertical speed.

[Arugela]

If I put fairings around all of those ore tanks would it get rid of the drag? I wonder if I could put some around those hub parts to make it look like a fuselage. Or A big one around the engine pod to give it the proper aesthetics. Do large fairings make a lot of lag or are they a good parts count reducer if used heavily?

Edited October 4, 2015 by Arugela

[ForScience6686]

If I recall correctly, they are counted as one part and remains as a part on the ship until the base is discarded. I would put the whole payload in the fairing or just use nose cones.

[Arugela]

Do nose cones reduce drag now? I read a while back they don't do anything. so far I have been flying with some on and I don't think it's doing much.. I'm confused how it works as I have more drag lines than exposed surfaces. It's like it creates drag in the middle of the plane even with no airflow.

I was also reading something about struts causing lots of lag. But that it only occurs on the parent part. I wonder if it would help to reverse my struts and start them on the fuselage instead of the ore containers.

Before Cones!

After Cones:

Does anyone see a difference? Besides the kerbals faces.

I only see a little less fuel in the one with cones. Probably from the extra parts.

Edited October 4, 2015 by Arugela

[ForScience6686]

Nose cones are highly recommended, not sure about the strut question though.

[Arugela]

I have an interesting question. If you use two reaction wheels. One on top of the other. And one is turned using the rotation tool to face the opposite direction and sitting on the exact same space as the other one, so they are completely stacked but facing opposite directions, will they double the reaction power or nullify each other?

My testing indicates they do not nullify each other. But is there anything else to this?

Edited October 5, 2015 by Arugela

[KerikBalm]

"Do nose cones reduce drag now? I read a while back they don't do anything. so far I have been flying with some on and I don't think it's doing much.."

Yes, they do, you probably read something that was written before the big 1.0 aerodynamic overhaul.

"I'm confused how it works as I have more drag lines than exposed surfaces. It's like it creates drag in the middle of the plane even with no airflow."

Every exposed part will have some drag, but if the node above them is occupied (by another fuel tank, or a nose cone) they will have less drag.

More red lines, but the sume of the drag is lower.

Test it yourself, go in sandbox, take a RT-10 SRB, put a mk1 inline on top of it, and some stabilizing fins at the bottom.

Launch it with no nosecone, see how high you get.

Launch it with a standard rounded nose cone

Then try an advanced nose cone

Then try a shock cone.

Then try a 2x2 structural panel placed flat....

You should see big differences

[Arugela]

One of the problems is, as I have finally discovered, is that I have 384 large reaction wheels and they are not properly struted(and I'm not sure if they can be). They wobble a bit and make it impossible to maneuver properly. I'll have to see what happens now that I have strutted them a bit. Are they going to add a new reaction wheel the size of the Kerbodyne S3-14400 tank. The game could use one. It should probably have 120 torque and simple 4x the stats(assuming double the diameter) or whatever is appropriate for the size increase to the diameter or however it works. I have to use too many parts for size of the ship and it really needs the torque. Any idea why one of these is not in game yet?

I have strutted some of it down but the inner stuff still has a lot of wobble more than likely and I keep deviating slightly from my ideal path in flight.. I wonder how much of this is the reason for my shakes after I get in orbit with full thrust.

Here are the updated craft files. I have more parts on the main ship and I put wings on the boosters.

http://www./download/kd21hsksdmki1w5/The_Orion_Basic_wBooster%283%29.craft <- With Boosters

http://www./download/fdwhpek32w1yd55/The_Orion_Basic%282%29.craft <- Without Boosters

Edited October 6, 2015 by Arugela