Jump to content

Rocket flipping and spinning on powered decent


Recommended Posts

I'm trying to land on Laythe and I'm using the infinite fuel and electricity cheat. Got myself into a nice tight orbit and picked a land mass. Narrowed my trajectory down to land on solid ground near a crater. So I'm coming down now in reverse. The rocket's light because there is no fuel except infinite fuel. I do have a lot of fins on the bottom and the base is that huge engine with four nozzles, but drained of fuel. The game usually lets me descend quite a ways and then it starts with the spinning. i can stop it manually, but  only for a moment and I have other things to do like control my descent speed. I have at times slowed the rocket way down. It still flips. Tends to point right at the land for impact which ends badly for anyone on board. i could deploy my chute, but I wanted to save that for earth or maybe just never use it. I like the idea of powered descent and see no reason why my rocket suddenly decides it has to spin and flip when I'm controlling the descent nicely with the application of thrust. I've tried turning off SAS. That cures the spinning problem but the rocket still tends to flip and become uncontrollable. I need it to point in the direction opposite to it's trajectory so that I can apply thrust and slow down. It works on the moon but not on Laythe, maybe because it has an atmosphere. 

To me it feels like the game thinks the rocket should be hurtling through atmosphere when its actually creeping. The game is thinking, I'm flying through an atmosphere in reverse and I'm unstable and stuff so I've got to go into a death spin because I'm the game and i insist that you use a parachute or a plane if you want to land on a planet with atmosphere.

In other words the game is a jerk. 

Edit: No ones here. fine. I'll talk to myself. I've crash landed on Laythe. Stopped the spinning by turning off SAS, but when I got very low the rocket still wanted to flip and it ended up crashing on it's side. My Kerbal survived, but am now stranded. I think the aerodynamics being all wrong are part of the problem, but Space X has similar aerodynamics and they've been landing. It should work. 

Edit 2  I reloaded and lowered myself very fast. Then I gave it high thrust for a brief period and managed to lower my incoming speed greatly. The rocket swung back and forth. I managed to apply thrust at critical points and land on my landing gear at a pretty high speed. Probably around 50m per second. My landing gear consist of long structural beams that radiate out and down and are nearly as wide as the rocket is tall so the rocket bounced and settled down. I destroyed a few short segments of my landing gear but its still usable. Due to high gravity I was unable to exit the craft as I'd never be able to climb back up and the jet pack isn't powerful enough to get me to the capsule. I found that out on a previous crash. I took off after awhile and I don't think there was any damage to the engine nozzles or the fins. However, the rocket flew bizarrely. I think the atmosphere is unusually thick because the rocket heated up really fast. I had to throttle down and I had a lot of trouble going in the direction that I wanted to go. The rocket kept swinging around. Is this caused by gravity from Jool? Jool is very close, quite huge and moving. I managed to get out of the atmosphere and right away the rocket started to behave itself so i think the atmosphere and the game engine are to blame for all the bizarre behavior. 

 

Edited by Chik Sneadlov
science
Link to comment
Share on other sites

2 hours ago, Chik Sneadlov said:

I'm trying to land on Laythe and I'm using the infinite fuel and electricity cheat. Got myself into a nice tight orbit and picked a land mass. Narrowed my trajectory down to land on solid ground near a crater. So I'm coming down now in reverse. The rocket's light because there is no fuel except infinite fuel. I do have a lot of fins on the bottom and the base is that huge engine with four nozzles, but drained of fuel. The game usually lets me descend quite a ways and then it starts with the spinning. i can stop it manually, but  only for a moment and I have other things to do like control my descent speed. I have at times slowed the rocket way down. It still flips. Tends to point right at the land for impact which ends badly for anyone on board. i could deploy my chute, but I wanted to save that for earth or maybe just never use it. I like the idea of powered descent and see no reason why my rocket suddenly decides it has to spin and flip when I'm controlling the descent nicely with the application of thrust. I've tried turning off SAS. That cures the spinning problem but the rocket still tends to flip and become uncontrollable. I need it to point in the direction opposite to it's trajectory so that I can apply thrust and slow down. It works on the moon but not on Laythe, maybe because it has an atmosphere. 

To me it feels like the game thinks the rocket should be hurtling through atmosphere when its actually creeping. The game is thinking, I'm flying through an atmosphere in reverse and I'm unstable and stuff so I've got to go into a death spin because I'm the game and i insist that you use a parachute or a plane if you want to land on a planet with atmosphere.

In other words the game is a jerk. 

Edit: No ones here. fine. I'll talk to myself. I've crash landed on Laythe. Stopped the spinning by turning off SAS, but when I got very low the rocket still wanted to flip and it ended up crashing on it's side. My Kerbal survived, but am now stranded. I think the aerodynamics being all wrong are part of the problem, but Space X has similar aerodynamics and they've been landing. It should work. 

Edit 2  I reloaded and lowered myself very fast. Then I gave it high thrust for a brief period and managed to lower my incoming speed greatly. The rocket swung back and forth. I managed to apply thrust at critical points and land on my landing gear at a pretty high speed. Probably around 50m per second. My landing gear consist of long structural beams that radiate out and down and are nearly as wide as the rocket is tall so the rocket bounced and settled down. I destroyed a few short segments of my landing gear but its still usable. Due to high gravity I was unable to exit the craft as I'd never be able to climb back up and the jet pack isn't powerful enough to get me to the capsule. I found that out on a previous crash. I took off after awhile and I don't think there was any damage to the engine nozzles or the fins. However, the rocket flew bizarrely. I think the atmosphere is unusually thick because the rocket heated up really fast. I had to throttle down and I had a lot of trouble going in the direction that I wanted to go. The rocket kept swinging around. Is this caused by gravity from Jool? Jool is very close, quite huge and moving. I managed to get out of the atmosphere and right away the rocket started to behave itself so i think the atmosphere and the game engine are to blame for all the bizarre behavior. 

 

Hey maybe you could try using some airbrakes or something that can create a lot of drag on the  side of the rocket which you want to face retrogade. 

In my experience, atmospheric entry with long crafts can lead to spinning and flipping so some sort of drag is needed to make it controllable. 

The reason SpaceX is able to land their rockets is because they deploy those controllable fins during re-entry in the atmosphere, in ksp, those fins are the airbrakes. 

I really don't know if jool is the one to blame. I am not that experienced in the game. 

Edited by Space boy
Link to comment
Share on other sites

1 hour ago, Chik Sneadlov said:

In other words the game is a jerk.

Do you need to let off some steam, or are you actually asking for help?

In case it is the latter: things (planes, rockets, whatever) tend to "fall" (here "flying" just means "falling horizontally") through an atmosphere with their center of mass in front of their center of drag (or center of lift as long as that dominates the drag). So if you have a rocket that has lots of fins at the bottom then it wants to "fall" nose first! (Or more scientifically phrased: the aerodynamic forces tend to turn it nose-first.) This is fine for lift-off when you want it to go nose-first, but not so good if you want to land bottom first. In theory you can counter-act the aerodynamic forces with enough control authority - e.g. from engine gimbaling, reaction-wheel torque, or even aerodynamic control surfaces (as long as they can rotate far enough to actually provide control) - but this is a bit like trying to drive a nail into a wall head-first. My suggestion would be to remove the all (or at least most) of the fins at the bottom of the rocket, add some high-drag devices at the top (like airbrakes as suggested by @Space boy) and maybe change to engines with more gimbaling range.

I'd also suggest that you have a look at a tutorial about aircraft design, e.g.:

https://forum.kerbalspaceprogram.com/index.php?/topic/47818-basic-aircraft-design-explained-simply-with-pictures/

Yes, I know that you don't want to design an airplane, but the physical realities are the same.

2 hours ago, Chik Sneadlov said:

but Space X has similar aerodynamics and they've been landing.

If you look at the Falcon 9 boosters, then you see that they don't have any fins at the bottom of the rocket, but do have grid fins (which provide high drag when unfolded) at the top of the booster. In addition the center-of-mass of a mostly empty real rocket is very far at the bottom of the rocket because the engines are quite heavy and empty fuel tanks are lightweight (much more so than in KSP, where many people complain about fuel tanks made of armor plate).

Link to comment
Share on other sites

2 hours ago, Chik Sneadlov said:

 

In other words the game is a jerk. 

When you will solve this problem I bet you will be saying "this game is awesome." Keep trying. 

2 hours ago, Chik Sneadlov said:

 Edit: No ones here. fine. I'll talk to myself. 

It can take a  bit of time to get replies. 

Link to comment
Share on other sites

It's also entirely possible your issue is a control point problem.

It would greatly help if we could see pictures of your rocket, and screenshots from your descent and landing attempt.  If your control point is not where you think it is, it can cause all kinds of problems like this since your rocket may be trying to flip.  Although having a lot of fins on the bottom isn't helping you at all.

Link to comment
Share on other sites

3 hours ago, Chik Sneadlov said:

To me it feels like the game thinks the rocket should be hurtling through atmosphere when its actually creeping.

Define "creeping". 100m/s already is a lot.

Link to comment
Share on other sites

Please pardon my rants if that's what you consider them. 

Thanks for the replies. I ended up surviving a landing attempt on eeloo,  the rocket on it's side. It once again was behaving oddly as I lowered it to the ground. It was in once piece and undamaged this time, but when I tried to take off by sliding it up a hill it exploded several times. (reloads)

Talking to myself is actually very helpful. I often figure things out when I try to write about the problem. Of course people on the forum are the best. 

Fins on the top makes sense for Space x. I put my fins on the bottom because I want to do that gravity turn and I want to be able to control spin and that kind of thing as I lift off from kerbin. It worked fine and in the past this kind of thing has given me no trouble on low gravity planets. I think my COG is only a little higher than my COT. That might be part of the problem. Its my first time running with cheats and a rocket that has 0 fuel weight. With the mammoth engines it accelerates like crazy. Pulling like 29gs at times. I'm pretty sure it would kill astronauts. No blood in the brain. The lower part of the rocket has also started to heat up badly even in deep space. In a vacuum. How can that be? Can i make a gravity turn with just fins on the top? Maybe I should put them on top? Air brakes might work but only on planets with atmosphere.

I'm sorry i don't do pictures or video. I've tried it a few times but can't figure out how to set up an account. That's how it went last time. I finally put something on reddit and linked to that and they started griping about me drawing on the image and questioning if i was a robot or something, They have all these rules and they expect you to know everything before you post the first time. I can never remember my passwords for the  sites and the software I have to remember doesn't work and I don't understand that either. 

Kerbal is a great game because of what it can do, but it is also a jerk because of the way things work. Like the editor for example. They give you all these parts and you try to use them the way you want to use them, but they don't snap together. You have to know how to use them or do a lot of trial and error. That's what I did with the small wire frame beam. Tried to attach it to the side of an engine, but it doesn't let me. It attaches to the center of the engine. You can't stack this way. You have to stack vertically from the bottom of a fuel tank or a capsule and then you can break it off and attach it radially to a fuselage. Then you can magically bend all 8 of your assemblies, which is great but also not like anything in reality. You just bend these beams at 90 degrees and they stay welded together and incredibly strong. 

I've got a save before eeloo. I think I'll just orbit the planet because there really isn't anything interesting down there anyway. Obviously I need to change my rocket design. Then I'll beat it back to kerbin and try to land there. That'll take me a day and a half. Either that or go play with some of those robotic parts. That looks interesting. What would I even want to build? I never know what I'm trying to do with stuff like that. Have the same prob in scrap mechanic and factorio. I need to work under the wing of an expert like Manley. 

 

 

 

Edited by Chik Sneadlov
air brakes
Link to comment
Share on other sites

i dunno about the physics, but it will be something to do with centre of mass x aerodynamics guessing (going too fast / too big a lander) ? you still have not said your speed unless i have missed.)

BUT i can tell you; i overcame the problem you are speaking, by putting an inflatable heat shield on the top and bottom like a hamburger, then deploying chutes at ~2000m (obviously might want to take an engineer with you).

 

 

Link to comment
Share on other sites

Turns out it was my rcs nozzles that were heating up in the vaccuum of space from running for so long. I don't remember my exact speed. I only have one crew, a pilot. I can't repack chutes and so i don't want to deploy until I'm back on kerbin. I wanted to do a powered descent, ideally. I'm on my way to eve now. Set a maneuver node and chose warp to. Didn't realize until ten minutes later that it was a about 66 years away at the fastest time acceleration or about an hour in real time. 

Link to comment
Share on other sites

your orbital trajectory may be a "little bit" "all around the houses" due to having "infinite fuel" maybe ?; i don't really think infinite fuel is good unless you are sandbox building spacebases or something IMHO.

i have only ever been to laythe and made it back into low orbit with fuel - laythe is on the outer system (i.e. "very hard").

BUT here is a good video, to see what can be accomplished / landing on laythe if you are interested:

(stolen from this thread -

 

Link to comment
Share on other sites

5 hours ago, k00b said:

i dunno about the physics, but it will be something to do with centre of mass x aerodynamics guessing (going too fast / too big a lander) ? you still have not said your speed unless i have missed.)

BUT i can tell you; i overcame the problem you are speaking, by putting an inflatable heat shield on the top and bottom like a hamburger, then deploying chutes at ~2000m (obviously might want to take an engineer with you).

 

 

That's what I was thinking too about the inflatable heat sheilds. 

Link to comment
Share on other sites

9 hours ago, Chik Sneadlov said:

Please pardon my rants if that's what you consider them.

No harm done.
One problem is that I don't react well to people of whom I think that they are blaming the game for their faults. They may not have meant it that way, but what someone writes and what the readers understand is not necessarily the same.;)

9 hours ago, Chik Sneadlov said:

Kerbal is a great game because of what it can do, but it is also a jerk because of the way things work.

I like KSP because it simulates the physics remarkably well. There are many simplifications in KSP but in many scenarios the simulation is close enough to help you learn something about the real physics. It is one thing reading about the 737 MAX and its big engines that cause trouble, but it is another thing building a plane in KSP that flies fine in one situation and tries to flip around in another. And then figuring out why that happens...
(One reason is that the center-of-drag shifts when the air is coming at the craft from different directions.)

Where KSP does have a big problem is that it doesn't actively teach the physics. It only lets you experiment and shows you what works and what doesn't work. The devs of KSP2 promised that their take will be better at teaching how to get things done, but I'm still skeptical because it is complicated. There is no easy way around that. (Well, except "dumbing down" the physics, which would make me very unhappy.)

(And, yes, I react differently to "The game is a jerk because the physics isn't the way I want it to be." and "The game is a jerk because it doesn't help me to understand the physics." ;))

9 hours ago, Chik Sneadlov said:

Fins on the top makes sense for Space x.

I'm not sure how you meant that. But keep in mind that the grid-fins that SpaceX uses are folded in during ascent and only unfold for reentry.

9 hours ago, Chik Sneadlov said:

The lower part of the rocket has also started to heat up badly even in deep space.

Engines generate heat when they are running. If they are connected to a big fuel tank, then this tank can soak up the excess heat and radiate it away over time. If they cannot do that (or cannot do that fast enough like the nuclear engines) then they heat up, which will increase the rate at which they radiate away heat.

IRL some rocket nozzles are designed to become red hot when the engine is running, so that they can radiate away enough heat.

9 hours ago, Chik Sneadlov said:

Can i make a gravity turn with just fins on the top?

Same as my previous answer: in theory yes, but I wouldn't recommend it. :cool:

A rocket cannot be aerodynamically stable going forward and backward without moving around the center-of-mass or the center-of-drag. You can do the latter by adding high-drag devices that can be triggered like airbrakes of parachutes to the top of the rocket. Moving the center-of-mass can be done by selectively using up fuel or pumping it around. But getting this right on a rocket is quite hard because the center-of-lift that the editor shows is only close to the center-of-drag in typical airplane situations. (Where there are lots of lifting surfaces, in dense air, and at low-ish speeds.)

My suggestion for you would be to leave out most or all of the fins. That will make your craft harder to control during ascent because it will not automatically point itself into the airstream, but it will also not automatically fight you when trying to go backwards during landing. It also helps a lot to not point too much away from prograde or retrograde (during landing) while in atmosphere. That way the turning moment of the aero-forces is small and you need less torque from engine gimbalig or reaction wheels to keep it pointed where you want to.

Link to comment
Share on other sites

I fiddled around with the ship back on earth. Took the fins and 3 radially mounted fuselage pieces off and tried it with fins near the top pointed down and again pointed up and finally just took off all the fins and put all the lights down under the rocket where the engines are attached. That worked the best, but if you come blazing in in reverse on kerbin and then you apply a lot of thrust to where you hover your computer will go nuts and pointing at retrograde won't mean pointing up, the rocket will tend to spin like a top and swing side to side like a pendulum. Turn off SAS and the spinning will stop but you have to just guess at whats straight up. You almost never get stable, steady descent with gentle movements unless you descend very slowly. Even then it tends to spin. So I built another rocket. I was going to make it very squat, but just putting on the science stuff and payload and it was already getting taller. I had wanted to bring the center of mass closer to the center of thrust, but it just wasn't possible unless that's all you want to do. I also tried putting rcs below as well as above the center of mass. That seemed to help some, but again, limited.

Took the new ship to the moon without even going to orbit, thanks once again to cheats for unlimited fuel and electricity. Then I landed near the north pole in the smoothest landing I've ever made in Kerbal. Later I noticed I couldn't deploy my payload or use my science because it told me I had to be stationary to do seismic readings. I noticed the ship constantly rocked, which is not realistic as it's on a tripod of landing gear and tripods do not rock. I thought I'll just move it a few feet and see if I can stop that rocking, but I was careless and oh, I lifted the landing gear thinking it wasn't necessary, but no. It slowly fell on it's side. In my attempt to right it, I snapped off the capsule so that's over. The rest of the ship is roaming around the huge crater on the top of the moon at low thrust. 

I guess the unlimited fuel cheat applies to the personal rcs jetpack. I got myself going about 400 m/sec and am heading out into space if I don't correct. Just surveying the area. Its neat but I wish there wasn't that pop up and there are vents and rocks now. Don't have the equipment to analyze though. 

I've been thinking about building a self righting mechanism. Seems like landing is always risky. 

If you think Kerbal is above criticism, pick up a three legged stool or three legged anything and put it on any kind of ground. Does it wobble once it settles down? No. It can't unless the leg is loose or the ground keeps settling. Kerbal terrain acts weird and its not intended weirdness. 

 

 

 

 

Link to comment
Share on other sites

1 hour ago, Chik Sneadlov said:

but if you come blazing in in reverse on kerbin and then you apply a lot of thrust to where you hover your computer will go nuts and pointing at retrograde won't mean pointing up, the rocket will tend to spin like a top and swing side to side like a pendulum.

This is because, when you try to hover, if you start going back up, then you're now pointing prograde.  Since your SAS is set to retrograde, it will try to flip.

SAS is supposed to reset to Stability Assist when you get below 1 m/s, but if your horizontal velocity is too high, you'll never actually get there.

What you want to do, when coming to a stop to hover or land gently, is switch SAS to Radial Out.  This should keep it pointing pretty much straight up while you hover, unless you're travelling sideways too fast or you're moving upward too fast.

Link to comment
Share on other sites

This thread is quite old. Please consider starting a new thread rather than reviving this one.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...