KSP Tutorial: Missing the Ground

[rmaine]

1 hour ago, Meecrob said:

Ok let me try a different way. Say we have a pen stood on end vertically. It is stable, all forces are balanced. If I tip it slightly from the vertical axis, it will fall over because the forces are now unbalanced. Now imagine I do the same experiment, only the surface I am conducting this experiment on can accelerate upwards and also rotate (pitch) so that it is always acting through the longitudinal axis of the pen as it tips. The pen still tips as in the first experiment. The only difference as far as the pen is concerned between the first and second experiments is that in the second, the pen is now moving upwards and/or downrange...there is no force to prevent the pen from tipping. There is no force to restore the pen to a stable condition.

A little hard to follow your analogy, but I think I got it. It is still wrong. I'm not sure how to explain this to you, but I assure you I *DO* know the physics. Spent my career at NASA as an expert in flight dynamics. But pulling the "I'm an expert" line isn't good, so I'll try otherwise. The only reason your pen tips is that the forces are *NOT* acting along the centerline through the center of mass. By adding a moving tilted surface, you are actually making the situation much more complicated rather than simpler, so it doesn't help much as an analogy. No, attaching something to a rotating, accelerating surface is not anything close to the same as attaching it to the ground. It is far easier to look at the actual situation instead of trying to imagine analogies that introduce more complicated dynamics than the original. 

As MechBFP and Ashandalar alluded to, you need to realize that there are 3 "degrees of freedom" (to use a technical term, I'm afraid) here. I real flight, there are 6, but as these diagrams are all in a flat plane, only 3 of them are at issue. Two of them are the motion of the center of mass in the vertical and horizontal directions. Those are the two that gravity turns are about. Force components through the center of mass are what affect those. Those forces include both gravity and the component of thrust along the centerline. In the case at hand, all the thrust is along the centerline. The third is the direction that the vehicle is pointing. It is absolutely essential to understand that this has nothing at all to do with the direction of motion of the vehicle. You can turn around and fly backwards and that doesn't change the velocity of the center of mass. Conversely, changing the velocity of the center of mass doesn't cause the craft to rotate. Repeat after me, these are completely independent. The wiki diagram you looked at is causing confusion because it is only about the velocity of the center of mass and yes, it talks about rotation of the velocity vector, but that has nothing to do with the rotation of the vehicle. The only part of a force that causes vehicle rotation is the component of the force that is not directed through the center of mass. If you gimbal the engine so that it is not  thrusting down the centerline, you'll get such a component. Just like things don't move if nothing pushes them (gravity counts as a source of pushing here),  things don't rotate unless something twists them (that torque word mentioned above).

It's actually easiest if you don't try to separate the thrust into vertical and horizontal components. The cleaner separation here is the component through the center of gravity and the component perpendicular to that. But if you really insist on looking at the component of the thrust that is upwards, ok, we can do it that way. Yes, that component of the thrust gives some twisting force (torque) that would tend to push the nose down. But you don't get to just stop there. If you are looking at the upwards component of the thrust, you have to *ALSO* look at the horizontal component; it has both except when the rocket is vertical. That horizontal component is aimed below the center of gravity and thus gives twisting force tending to push the nose up. Turns out that, assuming the total thrust is aligned along the centerline, that nose up twist exactly counterbalances the nose down twist. This is really the harder way to do the analysis - awkward choice of axis system (and it gets really awkward when you are doing real dynamics instead of sticking to a flat plane and adding in aerodynamic forces  - which is why nobody does it that way). But if you want to stick with vertical and horizontal directions, you do have to include them both. When you earlier talked about the engine thrusting upwards, it hadn't even occurred to me that this was because you were ignoring its horizontal part; can't do that.

Edited February 1, 2023 by rmaine
"it's" -> "its". My mind knows better; fingers don't

[Vl3d]

4 hours ago, rmaine said:

The only part of a force that causes vehicle rotation is the component of the force that is not directed through the center of mass.

We won't understand gravity turns if people keep confusing them with craft rotation.

There is torque like rmaine said caused by a force that doesn't go through the COM. That's not just gravity. How could gravity cause torque by itself on a free falling body? It can't. A vertical pen falls to one side because the COM becomes misaligned with the point where the table pushes on it. Same for rockets.

The confusion has it's origin in the fact that we're accustomed to thinking that the non-gimballed engines thrust always goes through the COM. But when gimballing the force components get separated at the point where the engine is, not at the COM.

What is a gravity turn? It's just a parabola. It's the path any moving mass takes when acted upon by gravity. It's always there when a velocity vector is not aligned with the vector of the gravity. It's why orbits are curved. It has nothing to do with the direction an object without thrust is rotated or the shape of the object if there's no atmosphere. Aerodinamics are a way to passively keep the orientation of the craft in line with the velocity vector.

Conclusion: rotating your craft passively using a gravity turn only works when in an atmosphere. Otherwise you need gimballing or RCS or reaction wheels.

But where do the real DeltaV savings of the gravity turn come from? From not needing to spend fuel to rotate the velocity vector (not the craft!). And from minimizing drag loses as the craft is generally aligned with the velocity vector.

Still, there are also other factors to take into account.

 

Edited February 1, 2023 by Vl3d
i reformulated what I understood better

[Vl3d]

PS: The tutorial mentions "gravity turn" but does not explain it.

And I'm not ashamed to say I've never done a gravity turn. I always manually controlled craft on ascent. I just now understood that you have to pitch and then lock to prograde instead of using WASD. Thanks @rmaine for making me curious!

@Ghostii_Spaceplease pass on that we need a dedicated tutorial on what is and how to do a gravity turn. I've wasted so many hours and so much energy on manually controlling ascent until now.

Edited February 1, 2023 by Vl3d

[The Aziz]

I don't trust the "lock on prograde and don't touch the controls". All my rockets are different, some turn too aggressively, some barely at all. The outcome varies from ending up horizontal at 15km, through wasting tons of fuel on circularization because the rocket didn't turn much, to straight up flipping mid flight (rarely, because I know to observe the drag arrow)

Also, again, people, you're making rocket science where no rocket science is needed, this is a beginner tutorial. Reading long posts on what a gravity turn actually is ain't gonna help anyone flying Kerbal X for the first time. Time for more in-depth tutorials on controlling the craft, inclination changes and whatnot will come later.

Edited February 1, 2023 by The Aziz

[yoarebignoob]

geez really a small animators mistake sparked a whole discussion

[Ghostii_Space]

7 hours ago, Vl3d said:

PS: The tutorial mentions "gravity turn" but does not explain it.

And I'm not ashamed to say I've never done a gravity turn. I always manually controlled craft on ascent. I just now understood that you have to pitch and then lock to prograde instead of using WASD. Thanks @rmaine for making me curious!

@Ghostii_Spaceplease pass on that we need a dedicated tutorial on what is and how to do a gravity turn. I've wasted so many hours and so much energy on manually controlling ascent until now.

Thanks for calling it out! 

[LHACK4142]

How do you think these tutorials are stored and presented? As video files, or something else?

If they are just pre-produced video files, I imagine any inaccuracies found during the early access might be kinda difficult to fix?

Edited February 1, 2023 by LHACK4142

[poopslayer78]

On 1/27/2023 at 12:01 PM, The Aziz said:

Kids stuff? I'm nearing 30 and I enjoyed it

Same.  I plan on watching all the KSP2 tutorials even though I'm a KSP1 vet.  Mostly for the feels, but I'm sure there will be a few humbling "wait you could do that all along?" moments occasionally.

[yoarebignoob]

On 2/1/2023 at 10:30 AM, Vl3d said:

And I'm not ashamed to say I've never done a gravity turn. I always manually controlled craft on ascent. I just now understood that you have to pitch and then lock to prograde instead of using WSAD.

not to hate or anything but locking on prograde wont make a good gravity turn. also literally turning your rocket using WSAD counts as a gravity turn

[MechBFP]

28 minutes ago, yoarebignoob said:

not to hate or anything but locking on prograde wont make a good gravity turn. also literally turning your rocket using WSAD counts as a gravity turn

Exactly. The only launch that couldn’t be counted as a gravity turn is one that is  launched straight up and then has one immediate 90 degree turn to the horizontal. 

As long as a launch has some horizontal component before it reaches 90 degrees, then it is a gravity turn. Obviously the efficiency of the gravity turn will vary wildly here  as a result of how exactly it was performed  

Now a “perfect” gravity turn is a whole different story and that is pretty much impossible to accomplish on a regular basis given the tools in the game. Thankfully a “good enough” gravity turn that is still >=85% efficient when compared to a perfect turn  is fairly easy to accomplish. 

Edited February 2, 2023 by MechBFP

[ArmchairGravy]

On 2/1/2023 at 4:30 AM, Vl3d said:

PS: The tutorial mentions "gravity turn" but does not explain it.

And I'm not ashamed to say I've never done a gravity turn. I always manually controlled craft on ascent. I just now understood that you have to pitch and then lock to prograde instead of using WASD. Thanks @rmaine for making me curious!

@Ghostii_Spaceplease pass on that we need a dedicated tutorial on what is and how to do a gravity turn. I've wasted so many hours and so much energy on manually controlling ascent until now.

 

On 2/1/2023 at 6:03 AM, The Aziz said:

I don't trust the "lock on prograde and don't touch the controls". All my rockets are different, some turn too aggressively, some barely at all. The outcome varies from ending up horizontal at 15km, through wasting tons of fuel on circularization because the rocket didn't turn much, to straight up flipping mid flight (rarely, because I know to observe the drag arrow)

Also, again, people, you're making rocket science where no rocket science is needed, this is a beginner tutorial. Reading long posts on what a gravity turn actually is ain't gonna help anyone flying Kerbal X for the first time. Time for more in-depth tutorials on controlling the craft, inclination changes and whatnot will come later.

I found locking onto prograde tips me over way too soon. That method seems to be all about getting the throttle through max-Q just right. It's easier for me to just fly the rocket. Some I kick up high before turning, some I blaze right on through. It just depends on what is going to space today.