Loss of control at launch due to slight drift corrections

[Waxing_Kibbous]

It's getting pretty predictable when a launch will go sour for me (and I may be using the wrong terminology, hopefully the idea gets across). If my rocket starts to drift off the 90 degree vector, going above or below it on the trackball, adjusting it back to the 90 generally causes the ship to go upright (it moves back from the prograde), or in extreme cases flip and lose control. How do the pros handle such drift elegantly?

Edited November 2, 2015 by Waxing_Kibbous

[jarmund]

I'm far from a pro, but: control surfaces at the rear.

[Harry Rhodan]

Yeah. If it flips if the drag at the front overcomes the drag at the back. You counter that by having more draggy stuff like fins at the back as Jarmund said and/or by moving your center of mass up by having more heavy stuff at the front to lessen the leverage of the drag at the front.

[r_rolo1]

Well, this means that your rocket is more aerodynamically stable if travelling backwards, something you don't want in a rocket That is caused by too much drag in the front of the rocket compared with the back of it and can be solved by either decreasing the drag in front ( fairings or bays will help in here ) or by forcing the rocket to stay directed to the prograde vector by aerodynamical stabilization ( aka putting fins in the back of the rocket, dart-like )

[Skorpychan]

Thrust vectoring, and gentle control adjustments. Drag eats efficiency, and thrust vectoring works well. That, and more TWR. Driftiness sometimes goes away if you can just accelerate a little.

[Alshain]

Stabilizers (aka 'fins') do not work because of drag. While it's true that drag at the rear of a vehicle can in some cases hold the rear end stable (such as with trains, in which we are referring to friction drag not aerodynamic drag), stabilizers actually work because of air pressure on either side. If the air pressure is equivalent on both sides of a flat symmetrical wing (i.e. not a airfoil) then the wing cannot move and therefore anything attached to it doesn't have anywhere to go. The pressure pushes inward in both directions equally and it prevents slip. Same principle on the vertical stabilizer of an airplane. The stabilizers are at the rear because that is where they are most effective on the nose. (like a lever)

Edited November 2, 2015 by Alshain

[Waxing_Kibbous]

One thing I always use are 4 of the delta deluxe fins at the bottom of my middle rocket, and for really small ones the small fins. Just sometimes I get a lot of drift away from the 90, the only thing I can figure is that some parts up top that are attached asymmetrically are causing the nose to veer off.

[Motokid600]

By saying 90 I'm assuming you mean the prograde vector?

And you never want fins on the middle of your rocket. If you have fins on your second stage then you must have a set on your first as well. Look at the way sounding rockets do it.

So what exactly is going on? You say your rocket drifts off of prograde. Which leads me to believe you either have no pilot and/or no reaction wheel torque with SAS. It can be done, but you need to design your rocket appropriately to easily fly without SAS.

So... Do you have the ability to lock your angle of attack with SAS? You should be able to take your hands away from the controls with little input needed. If not then something is wrong. Could be asymmetric drag as you said. You should never have to fight the controls. A good design and a perfect gravity turn will nearly fly itself to orbit.

Post a picture if you can.

Edited November 2, 2015 by Motokid600

[Waxing_Kibbous]

I mean the 90 on the navball which ideally the prograde marker would move along like it was on rails. By middle rocket I mean the very center, "root" of the first stage, which is generally designed to get into orbit or pretty close, which will have fins as low as possible. Also, if not a pilot I use a probe core, and unless I forget I put a reaction wheel near the top quarter unless it's a really light rocket. I really think I'm underestimating the drag calculations... What I don't get is how people launch huge rockets that look like Christmas trees with all the stuff hanging off them if that’s the case.

[Alshain]

One thing I always use are 4 of the delta deluxe fins at the bottom of my middle rocket, and for really small ones the small fins. Just sometimes I get a lot of drift away from the 90, the only thing I can figure is that some parts up top that are attached asymmetrically are causing the nose to veer off.

Don't use Delta Deluxe, they are active control surfaces. Even my 2.5m rockets have basic fins, that is really all you need. The very big rockets don't usually need fins at all.

- - - Updated - - -

By middle rocket I mean the very center, "root" of the first stage, which is generally designed to get into orbit or pretty close, which will have fins as low as possible.

The terminology for that is 'center stack'. Also, try a 2 stage center stack, with a better ISP for high atmosphere it allows you to make the rocket significantly smaller and easier to fly, not to mention cheaper. So for example, at 1.25m, use a Swivel engine on the first stage and a terrier on the second, typically you only need radial SRB's depending on your payload mass. For example, here is a Tier 4, 4 ton lifter. Try to shoot for somewhere close to 30km on the first stage and the second stage will be high enough it doesn't need stabilization.

Other good 2 stage stack combinations include:

Skipper/Poodle

Mainsail/Poodle

Mainsail/Skipper

Mammoth/Rhino

You can go on, these are just good if staying within the same size, but there is no real reason to do that, it just depends on your abilities and how you want to build. The key is recognizing that some engines perform better in low atmosphere while others perform better in high atmosphere or space.

If you are using fuel lines on the lifter at less than 120 tons payload (at least, possibly more), you are wasting money. I know this because I have lifters built for those sizes, adding more LFO engines other than the 2 stage stack is always significantly more expensive. Asparagus staging is an old practice that is no longer necessary or desirable for most lifters in 1.0.

Edited November 2, 2015 by Alshain

[Motokid600]

Larger rockets are actually easier believe it or not.

@Alshain - a serial staged rocket ( a small one anyway ) is trouble is it not? Unless that first stage carries the second through most of the atmosphere its very difficult to stage without flipping. You have to be perfect.

I find its much better for smaller rockets to not get shorter as they ascend unless the first stage really does some work.

[Alshain]

Larger rockets are actually easier believe it or not.

@Alshain - a serial staged rocket ( a small one anyway ) is trouble is it not? Unless that first stage carries the second through most of the atmosphere its very difficult to stage without flipping. You have to be perfect.

I find its much better for smaller rockets to not get shorter as they ascend unless the first stage really does some work.

Absolutely not. The first stage should get you close to 30km, at that point you probably will not flip. The only trouble it adds is figuring Delta V for optimal performance. If using KER you need to make sure you add the first stage delta V at 0km and the second stage at 30km using that little slider, same for TWR. If you only add it at ground level it's misleading, as you can see in the picture KER shows total of 2835 m/s but it's actually closer to 3600m/s (I shoot for 100km orbits).

Keep in mind, that assumes a proper gravity turn. By 30km, my nose is on a 10 degree incline. If you go up to 10, 20, or even 30km before turning, you need about twice the fuel I use.

One rule of thumb I have see suggested is each stage should burn for ~2min and actually after reading that I went and checked and indeed most my lifters show about 2 mins from ground level. It's not a perfect rule though.

Edited November 2, 2015 by Alshain

[Waxing_Kibbous]

I will give those tips a shot Alshain- I've actually had your lifter pack zipped on my desktop for a wek now, maybe I should actually check them out

[Alshain]

I will give those tips a shot Alshain- I've actually had your lifter pack zipped on my desktop for a wek now, maybe I should actually check them out

Lol, I actually have a few more since then. I haven't updated yet and I've been doing some additional fine tuning. They should work but if you do a really good launch some of them may have even more fuel than you need.

[Motokid600]

30km is a good stage. I find most people who are having trouble with rockets flipping during staging ( not the case here I know. Not yet at least lol ) is because people underbuild there first stage and it doesn't carry the rest of the rocket far enough out of the atmosphere and when they go to stage it flips.

Then again just realized I'm basing all my experience with FAR, so..

Edited November 2, 2015 by Motokid600

[Alshain]

I made a video of my 20ton lifter launch. (My first KSP video) Enjoy.

[magnemoe]

Don't use Delta Deluxe, they are active control surfaces. Even my 2.5m rockets have basic fins, that is really all you need. The very big rockets don't usually need fins at all.

- - - Updated - - -

The terminology for that is 'center stack'. Also, try a 2 stage center stack, with a better ISP for high atmosphere it allows you to make the rocket significantly smaller and easier to fly, not to mention cheaper. So for example, at 1.25m, use a Swivel engine on the first stage and a terrier on the second, typically you only need radial SRB's depending on your payload mass. For example, here is a Tier 4, 4 ton lifter. Try to shoot for somewhere close to 30km on the first stage and the second stage will be high enough it doesn't need stabilization.

Other good 2 stage stack combinations include:

Skipper/Poodle

Mainsail/Poodle

Mainsail/Skipper

Mammoth/Rhino

You can go on, these are just good if staying within the same size, but there is no real reason to do that, it just depends on your abilities and how you want to build. The key is recognizing that some engines perform better in low atmosphere while others perform better in high atmosphere or space.

If you are using fuel lines on the lifter at less than 120 tons payload (at least, possibly more), you are wasting money. I know this because I have lifters built for those sizes, adding more LFO engines other than the 2 stage stack is always significantly more expensive. Asparagus staging is an old practice that is no longer necessary or desirable for most lifters in 1.0.

Personally I found that the AV-R8 works well with the T30 who don't have gimbal. My lightweight Eve accent unit has 8 AV-R8 and need them.

It also depend on the shape of the item you launch, if its less aerodynamic than an oil platform you need lots of fins, an 2.5 meter stack with an streamlined fairing don't need any.

You might also want to use your payload as upper stage and refuel in LKO, this saves lots of weight.

Else I agree that <120 ton has no need for fuel lines with two execptions, first is the stupid payloads, second is early in career before you unlock the skipper and want to do an all minmus biomes or Mun science rover.

[paul23]

I'm always adding active control surfaces: going to far as to lock gimbal on engines untill large height and use only active/aerodynamic surfaces. I found those (when placed as a group of 4) way more responsive & predictable. Mainly because you can actually see the force vectors during flight (alt+f12), while you can't see the gimbal of engines so you can't debug that.