Why are Eve landers listing?

[Laie]

Hello,

it's a common problem, won't stay behind their heat shields.

My latest example would be this:

2nd picture shows the rocket ready for launch. During reentry, the rocket itself is drained. All the fuel is in the two tanks at the bottom, the CoM is just above the Lander Can. The vessel is symmetric except for the science container -- if that was the main problem, I'd be off on the pitch axis. But if you look closely, you see that the worst control input is on yaw.

This isn't just a momentary flutter: for quite some time during descent, SAS has to apply a lot of yaw control in order to keep the vessel lined up with srf_retrograde. I see no reason why this should be necessary.

If I turn off heat damage and do an uncontrolled entry, the vessel will settle in an attitude about 20° off from retrograde. It's by no means flipping, but has a strong desire to stick it's neck out. What is going on here?

[bewing]

Looks to me like your red dot and your green dot are not precisely aligned. I would guess that there is a small amount of horizontal displacement in the attachment of one of the bottom parts -- ie. something got attached in a non-centered way.

I suppose you can build it so that the science container is on the side that it tries to yaw towards, so that it will automatically counteract the yaw.

[Laie]

I think I've got it: the direction of force the  depends on the heatshield's rotation. Rotate it by 90 degrees and it becomes pitch, another 90° and it yaws in the opposite direction. There's a ModuleLiftingSurface on the heat shield -- could that be the reason?

I guess it's supposed to provide some body lift for the pod during entry or something? Can't say that I ever noticed.

ETA: however, the problem doesn't entirely go away if I remove the heat shield. It becomes much better, five degrees rather than twenty, and there's no longer a preferred direction. But it seems as if a stick with a blunt tip just won't fly straight, no matter what.

Edited March 10, 2019 by Laie

[Foxster]

Is it a balancing a pencil on your finger effect? 

There is a strong force on the leading edge of the craft and not so much at the rear end so it is inherently likely to tend to wobble at the back end. 

Adding LOTS more airbrakes to the back end usually fixes this for me. 

Oh and a continuous axial spin might help. 

Edited March 10, 2019 by Foxster

[Laie]

33 minutes ago, Foxster said:

There is a strong force on the leading edge of the craft and not so much at the rear end so it is inherently likely to tend to wobble at the back end. 

Yup. For the tall stick to act as a stabilizer, it has to be in the airflow somewhow. Put differently, AoA has to be non-zero for the stabilizer to work.

34 minutes ago, Foxster said:

Adding LOTS more airbrakes to the back end usually fixes this for me. 

Oh, the vessel above behaved well with just the four airbrakes. I was miffed that I needed any at all, though.... but yeah, a long stick alone just won't suffice, I see that now.

As it seems that a heatshield will induce a force in a predictable direction, it might be possible to build it such that a single airbrake would do. Guess I'll have to try that someday.

[Foxster]

4 minutes ago, Laie said:

... AoA has to be non-zero for the stabilizer to work.

 

Except in KSP, where the atmosphere blows straight through everything including heatshields. So your airbrakes work even though they are behind the heatshield. 

[Laie]

Just now, Foxster said:

So your airbrakes work even though they are behind the heatshield.

Yea yea, I get that. I was referring to my expectation that it should work without airbrakes -- no other stabilizer but the long tail. Nice idea, but if the tail is perfectly lined up, it will create no force.

[Foxster]

I see what you mean. I guess though that the stabilising effect will not be right at the back, so you'll get a pivot effect.  

[Superfluous J]

1 hour ago, Foxster said:

Except in KSP, where the atmosphere blows straight through everything including heatshields. So your airbrakes work even though they are behind the heatshield. 

Source?

I'm not running the game now but I recall testing my Eve lander and noticing the drag arrows going crazy whenever it wobbled into the airflow like the OP's craft. Also, entering a planet sideways as opposed to head on (with a lower surface area in the wind) most assuredly causes you to slow down faster.

[Foxster]

9 minutes ago, 5thHorseman said:

Source?

I'm not running the game now but I recall testing my Eve lander and noticing the drag arrows going crazy whenever it wobbled into the airflow like the OP's craft. Also, entering a planet sideways as opposed to head on (with a lower surface area in the wind) most assuredly causes you to slow down faster.

Is this not widely known? 

Occlusion in KSP only works for heat and for adjacent joined parts. In other words, you can't reduce the drag of trailing parts by have a larger part out front. Similarly, draggy parts like airbrakes are always in the slipstream even with larger parts out front.

"Also, entering a planet sideways as opposed to head on (with a lower surface area in the wind) most assuredly causes you to slow down faster." That's because you are presenting a lot of flat draggy faces of parts into the wind. If you go in head on, all the parts are still draggy, just a lot less so, even if visibly behind a large thing like a heatshield.

This  is the last big physics non-realistic effect in KSP that there seems to be no sign of being addressed. 

Edited March 10, 2019 by Foxster

[Superfluous J]

1 hour ago, Foxster said:

Is this not widely known? 

I don't know if it's widely known or not. I didn't know it.

And while it goes directly against my observations using airbrakes, I am willing to accept that I'm probably just seeing what I expect to see (that they peak out from behind the heat shield, get some drag, and then get pushed back) instead of what's happening (as they get farther from the line of the vessel's trajectory they have more torque, so when they're behind the heat shield they just have less affect because they're in line).

[Foxster]

For places like Eve, airbrakes can be mercilessly exploited. As long as they are inside a cylinder behind the rim of a forward heatshield they are occluded from the effect of heat but are quite unoccluded as far as drag is concerned. This means they can be quite effective in producing lots of drag at the the back end whilst be immune to re-entry heat. 

Edited March 10, 2019 by Foxster

[Laie]

1 hour ago, 5thHorseman said:

And while it goes directly against my observations using airbrakes, I am willing to accept that I'm probably just seeing what I expect to see (that they peak out from behind the heat shield, get some drag, and then get pushed back) instead of what's happening (as they get farther from the line of the vessel's trajectory they have more torque, so when they're behind the heat shield they just have less affect because they're in line).

Maybe I'm merely rephrasing what you meant, but just to be sure: a tall, spindly rocket is pretty aerodynamic and has little drag as long as it's pointed into the wind. If you start turning it sideways, drag increases -- mostly because you are presenting more surface area.

This works regardless of whether it is in the "shadow" of a heat shield, it only depends on the angle of attack (of course, when you stick out you have a higher AoA than when you don't, hence also higher drag). The heat shield shadow only matters with regards to heat, while all the rest of aerodynamics (lift, drag, ...) is blissfully unaware of slipstream or occlusion.

That nosecone will protect the golden tanks from a hasty ascent, but it does nothing to reduce their drag. If there were control surfaces mounted in their place, they'd just work.

Edited March 10, 2019 by Laie

[Foxster]

I'm not sure how entirely unrealistic it is though. In the picture above, in reality the sharp back end of that nose cone would produce a lot of turbulence behind it and a fair amount of air would interact with the tanks. There'd be no clean cone of air protecting the tanks behind the nose cone from the effects of air flow, it would be more complex than that. 

I'd say it's a thing that is difficult to model in KSP. By this point though I think we should be further along than having air just flow straight through everything. 

Edited March 10, 2019 by Foxster

[OHara]

This simplification never bothered me too much, either.

You'll have a shock wave through most of re-entry, in the shape of some kind of blunt cone, and that's where you'd expect most of the compression heating to happen.  Behind the shock cone, the air is moving much more slowly (subsonic versus supersonic) relative to the vessel, but still pretty fast and a lot more dense because it was squeezed at the shock front.  Most of the mass-per-second of air that the frontal area of the craft meets, still passes pretty near the craft behind the shock wave, just slower and more dense.

Still, the FAR whole-body approach does look better for a game like KSP, in hindsight.