Fairing Stability Problems

[jakeb1993]

Ok i am fairly sure everyone must be sick of me asking things by now BUT:

This is my first ever off world base ship. This is going to drop a module instead of just landing a ship down and calling it a base. I even used the infernal robotics hinge to create a deployable 'skycrane' thing.

Here is the issue, when i take it off, it wobbles. It wobbles like there is no tomorrow and eventually it flips, not matter the altitude. Now, if i remove the fairing and try again it is better, but drag pulls it down more and eventually it will do the same thing. The thing is, the fairing has auto created some struts, but i can't create any more to the wall lining myself. Not exactly sure how to do this now. Any ideas?

[FullMetalMachinist]

Not sure if you're aware, but there's a nasty bug with the stock fairings that causes the aero forces to be applied way out in front, causing stability issues. Fix it using Claw's Stock Bug Fix Modules. 

That probably won't help you much with your case of the wobbles though. You could try Kerbal Joint Reinforcement. Some very smart people around here think that it's more of a crutch, and that if you need it to stabilize your rocket then it means there's inherently something wrong with your design. I'm not saying they're wrong, or anything against them, but I'm not one of them. I say try it out and see if it helps. 

[N_Danger]

 I've found that the struts need to go from the de-coupler to the payload and then some more struts from the de-coupler to the fairing base. Then struts from the faring base to the rocket. I've never been able to attach a strut from or to a built fairing skin so I don't think that is possible.

I've also, before building the fairing, put struts from the boosters to the payload. It looks ugly but it works(and probably adds drag). I don't use KJR but I do use a lot of struts at times to stabilize the load.

 And the stock bug fix helps out using the stock fairing immensely. 

[Plusck]

I would say there are four or five major problems you need to deal with, and one catch-all problem which relates to all of them:

1 - fairing node strength: nothing to do with the fairing "lift bug", but the fairing attachment nodes are not that strong and need strutting if you have a heavy (and particularly a top-heavy) payload.

2 - lifter TWR: it is really far too high. I would get rid of the Mainsails and replace them with Skippers at the very least (they're also more efficient engines).

3 - thrust vector torque: this is certainly far too high. If I were you, I'd make a custom action group to toggle engine gimbal off. Select an action group, select each of your lifter engines and select the "toggle gimbal" option. You have aero surfaces for control so you really should only need engine gimbal if your ship starts drifting off prograde. The rest of the time it's probably hurting you because of the next thing...

4 - your "control from here" part: I can't see where your control pod is. If it's inside the fairing, you're asking for trouble due to the first point above (fairing node strength). The payload in a fairing will tend to wobble; if your "control from here" part is in that payload, then SAS is going to be overcorrecting all the time, probably ending up by ripping the craft apart.

5 - the fairing lift bug: if you get off prograde in the atmosphere, the torque arising due to lift from the payload will wrench your craft around. Although some people complain mightily about this bug, it really isn't a great problem as long as your gravity turn is done properly: if you stick to prograde all the way up to the upper upper atmosphere, and you do it properly, you will (a) get the most efficient ascent to orbit and (b) never have a problem with the fairing lift bug.

6 - ascent profile: connected to the problem with TWR (with such a high TWR you cannot get the most efficient launch from Kerbin) and the fairing lift bug (as above), and also involving the problem with fairing node strength. If your ascent profile is not good, you will put a lot more stress on your ship, on the fairing, and on the payload, you'll need to overcompensate with thrust vectoring, and you can easily end up shaking your rocket apart.

[Snark]

+1 to @N_Danger's assessment.  I don't think those struts are actually doing anything-- I think that the game doesn't treat the skin of the fairing as an actual structural element, so the struts are basically just a "bridge to nowhere".

You need to have a strut between something that's actually solid (i.e. not the fairing skin) and the fairing-internal payload.

Presumably you want to save mass on the fairing payload, which means you want to strut to it instead of from it, so that you'll drop the strut mass when you detach the payload.  So find some good anchor spots on the ship, inside the fairing, so that they can attach to the stuff inside.

Unfortunately, the game won't let you place a strut directly on top of the fairing base.  What I usually end up doing in cases like this is to put the short, squat conical adapter on top of the fairing base, a decoupler on top of that, and my payload on top of that.  Then I can strut from the conical adapter to the payload.

The good news here is that this should become a moot point in 1.1.  Squad has announced that fairings are getting two things in 1.1 that should make them a lot more useful.  First, they're fixing the "lift bug" so that the fairings won't cause aero instability and make your rocket flip all the time.  Second (and more relevantly to your current problem), they'll add "automatic internal strutting" so that stuff inside the fairing won't wobble-- basically, until you blow the fairing, the stuff inside it is rigidly held in place for you, not extra strutting required.

So your other option besides tinkering with your strut placement is to just wait until 1.1 comes out in the hopefully-not-too-distant future and the problem will go away.

[jakeb1993]

Okay so a fair few replies so here it goes:

 I started with @Plusck's ideas and replaced the engines with skippers to get TWR to 1.99, I placed a OKTO2 inbetween the decoupler and the fairing and on launch clicked on 'control from here', Since i don't have VAB Level 3 yet i couldn't create a custom action group, but i locked the gimbal on the engines. I tried to stick to the prograde  marker going upwards, but sadly at 6000m it flipped.

So i went and downloaded the fix. The good news is this time it worked, but the craft has issues. I had to fight it all the way to avoid another flip - at least the fairing doesn't wobble now - And once i turned off the engines at 35,000m the ship went into a slow backwards roll upon which it will never recover. Quite alot of my craft seem to do that and i don't know why, It flips backwards, like the weight of the front is too heavy and sending it backwards, but the bottom of my rockets is the heaviest. The only way to fix it is to blast the engines once i leave the atmosphere 

[Nich]

in 1.0.5 struts are required I have had payloads flex out of the fairings.  Struts fixed this.  

I agree TWR is too high, and fairing lift bug is probably killing you.

Struts from the fairing base to external tanks will help solidify the shell.

[Plusck]

17 minutes ago, jakeb1993 said:

@PlusckSince i don't have VAB Level 3 yet i couldn't create a custom action group, but i locked the gimbal on the engines. I tried to stick to the prograde  marker going upwards, but sadly at 6000m it flipped.

Ah, sorry about that.

Indeed, if you can't attach the gimbal lock to an action group (but if you have basic action groups you can try using the action group for landing gear, or lights or something), then locking them in the VAB is likely to be fatal since you don't have that control option if you need it. So sure, you avoid SAS overcorrecting and shaking things apart but at the same time you're unable to get back on prograde if ever it drifts off, which with unfixed fairings is highly likely to be fatal.

[FullMetalMachinist]

28 minutes ago, jakeb1993 said:

Quite alot of my craft seem to do that and i don't know why, It flips backwards, like the weight of the front is too heavy and sending it backwards, but the bottom of my rockets is the heaviest. The only way to fix it is to blast the engines once i leave the atmosphere 

The problem here is that (quite understandably) your reasoning is backwards. 

When you have a rocket going through the air, it will rotate around it's center of mass into a position that is aerodynamically stable. This means the "heavy" side will want to be at the front, and "light and/or draggy" end at the back. Having heavy stuff at the front will move the CoM forward. 

Think of a dart or arrow. The tip is heavy and metal, the body is either wood or plastic, and light fins are at the back. Try throwing a dart backwards and it'll self-correct to fly front first. That's because the drag forces have a long lever arm to torque the light back end around, and a short lever arm to work on the heavy front.

So the moral of the story is to try and design your rockets like a dart: heavy at the front, light at the back. 

Edit: relevant tutorial :

 

Edited March 28, 2016 by FullMetalMachinist

[jakeb1993]

1 minute ago, Plusck said:

Ah, sorry about that.

Indeed, if you can't attach the gimbal lock to an action group (but if you have basic action groups you can try using the action group for landing gear, or lights or something), then locking them in the VAB is likely to be fatal since you don't have that control option if you need it. So sure, you avoid SAS overcorrecting and shaking things apart but at the same time you're unable to get back on prograde if ever it drifts off, which with unfixed fairings is highly likely to be fatal.

I am just glad it got up, even if it takes an age to respond. Alas the mission failed anyway, the rocket didn't have enough Dv to make a mun orbit so tried to land hard straight from intercept, turns out 852 Dv for that wasn't enough.

[Plusck]

2 minutes ago, jakeb1993 said:

I am just glad it got up, even if it takes an age to respond. Alas the mission failed anyway, the rocket didn't have enough Dv to make a mun orbit so tried to land hard straight from intercept, turns out 852 Dv for that wasn't enough.

I don't use MechKeb, but it looks like you have 5000m/s before releasing the payload.

That is very hard to get to low Munar orbit, but possible. First, time your launch perfectly: leave when the Mun is about 5° below the eastern horizon. Next, do a perfect gravity turn (turning about 8-10° east when you're at about 100 m/s or less, depending on TWR, then stick absolutely to prograde all the way - if you're not down to 45° between 8-11km altitude then revert and try again), next don't stop accelerating when you reach the upper atmosphere - keep burning slightly left of Munwards, watching the map to make sure you stop at or slightly short of the Mun's orbit. This shoud give an encounter.

Then as soon as you can, drop a maneuvre node and get your encounter as close as possible. Ideally passing the Mun with a Pe of about 20km.

This method is actually only slightly more efficient than doing it perfectly the "normal" way. However, if you don't do it perfectly the normal way, you rack up inefficiencies, especially with all the gravity losses by coasting up from the lower atmosphere to orbit. If you keep burning from a lower altitude, you maximise the Oberth effect, which gives you a touch more leeway to perfect your approach and circularise at the Mun.

[N_Danger]

 35,000 meters for engine cutoff sounds a little low. One advantage of a lower T/W is that you are higher up in thinner atmosphere after cutoff when you coast to AP.  I try to lift off at a 1.5 T/W aiming be at 45,000 to 50,000 meters and be nearly horizontal at engine cutoff.

I also leave the center engine in a cluster with 30% or more gimbal for when control surfaces lose their bite in thin air.  This helped me to control flippy rockets in the past.

I will admit that I haven't done a lot of playing in a contract game so I don't know if this is available to you.

Edited March 28, 2016 by N_Danger
some grammer

[jakeb1993]

17 hours ago, FullMetalMachinist said:

The problem here is that (quite understandably) your reasoning is backwards. 

When you have a rocket going through the air, it will rotate around it's center of mass into a position that is aerodynamically stable. This means the "heavy" side will want to be at the front, and "light and/or draggy" end at the back. Having heavy stuff at the front will move the CoM forward. 

Think of a dart or arrow. The tip is heavy and metal, the body is either wood or plastic, and light fins are at the back. Try throwing a dart backwards and it'll self-correct to fly front first. That's because the drag forces have a long lever arm to torque the light back end around, and a short lever arm to work on the heavy front.

So the moral of the story is to try and design your rockets like a dart: heavy at the front, light at the back. 

Edit: relevant tutorial :

 

Okay, so this poses a new question, how the fudgecake do you build a top heavy rocket when the fuel tanks and engines reside at the bottom?

[Plusck]

1 hour ago, jakeb1993 said:

Okay, so this poses a new question, how the fudgecake do you build a top heavy rocket when the fuel tanks and engines reside at the bottom?

Precisely the problem! Also your fuel tanks drain from the furthest to the nearest, compounding the problem.

The two aerodynamic command pods (Mk1 and Mk1-2) are both pretty heavy, which helps.

If you have severe difficulties with flipping, you can lock the fuel/ox tanks at the top of the rocket and then unlock them shortly before the engines run out of fuel.

Asparagus staging and/or drop tanks help - a fuel line from the top of one stack to the top of the next will draw fuel from the bottom of the source stack up (but not perfectly: for some unknown reason fuel usage is not exactly intuitive and apparently [though I haven't really tested this myself] in a three-tank column with a fuel line drawing from the top, the top two tanks will empty at the same time once the bottom one is empty).

And if all that fails, the only real solution is fins at the base of the rocket, plus lots of torque (a number of medium-sized torque wheels work better than one large one in the current version of the game, also for unknown reasons. I don't mean stack them together - rather put one on the top of a couple of side stacks and one on the main stack), plus carefully following prograde (SAS doesn't always get this right, so manual adjustment to follow prograde while using only "stability assist" may be better than "prograde hold"), plus not going too fast in the lower atmosphere (accepting gravity losses as the price to be paid for your light and draggy payload).

Oh and plus RCS (or Vernors for larger rockets), too. On difficult, big rockets Vernors work very well.

Edited March 29, 2016 by Plusck

[jakeb1993]

So here it is, my new rocket

Now it says that the TWR is 1.32, but when i launch its 1.17. Now of course thiswon't be enough to get me to the Mun, but the good news is that it doesn't flip as the CoM seems to be in the center of the central rocket. Personally i would like those boosters to be between the central rocket and the aux rockets, but i have no idea about going for that.

I am just thinking, all this thrust and i can't get one measly planetbase module to the Mun xD

[Plusck]

Looking at your original pic again, there is another problem that was probably hurting you: the fuel lines from the outer stacks to the core. Since they source from the orange tanks, the top brand adaptor tanks will drain first. This is bad.

You really want to have the brand adaptors drain last. One difficulty is that even if you move the fuel line to draw from the brand adaptors at the top, the side stack engines will drain them quite quickly, while the Twin Boar in the centre (since it is drawing from four stacks) will draw much less from the orange tanks. In all cases with that setup, you'll find it hard to keep the top of the rocket full of fuel without manually locking tanks (which you most definitely don't want to do asymmetrically during flight).

So the best bet would be to replace that 4x symmetry with 2x 2x symmetry, with asparagus-style lines from the top of the first pair to the top of the second pair then to the core. This won't entirely cure the problem (consumption via the fuel lines will never be high enough to drain the orange tanks before the adaptor on each side stack) but will limit the difference. You might even find (and MJ should tell you this in the VAB) that you have enough dv with asparagus staging to be able to lock the tanks on the adaptors on the second pair of side stacks until you get to engine cut-off, which would be ideal.

And if you have 2x 2x symmetry, you could even add a small 2.5 tank above the orange on the first pair of side stacks, which (in addition to allowing you to lock tanks on the second pair of adaptors) should also raise total dv enough to get to low orbit around the Mun...

[jakeb1993]

Okay, so my first ever Asparagus style ship

Here is how it goes:

Tanks West and East supply fuel from their central tanks to the upper central tank (that way the center is as stable as possible during ascent. Upper and Lower tanks drain fuel into the center to try and maintain center of mass.

Tanks North and South Central feed East/West central. Upper and Lower feeds central again.

North/ South drops first followed by East/West. Weirdly Mechjeb can't calculate the Dv after the first two tanks drop. It spent 3 minutes flashing different figures on two stages until it got it's story right. If what Mechjeb says is true, this aint going to the mun.

 

Edit. Just launched it, seemed to be going well until it came to drop the fuel tanks, it usual style  the tanks decided to colide with the center despite using the spaced out decouplers.

Edited March 29, 2016 by jakeb1993

[Plusck]

27 minutes ago, jakeb1993 said:

Okay, so my first ever Asparagus style ship

Tanks West and East supply fuel from their central tanks to the upper central tank (that way the center is as stable as possible during ascent. Upper and Lower tanks drain fuel into the center to try and maintain center of mass.

...

Edit. Just launched it, seemed to be going well until it came to drop the fuel tanks, it usual style  the tanks decided to colide with the center despite using the spaced out decouplers.

No real need for the fuel lines on the side (by abundance of caution I understand, but I'm not sure it will give the result you really want...), but that's a detail.

However - fuel lines don't work that way!

What happens is that the game takes each engine, and works from part to trying to see what fuel is available. If a source has already been noted, it'll be ignored if ever it comes up again. Fuel lines indicate a priority direction (so the engine will draw via a fuel line, rather than from the next connected part upwards).

The lines from the middle tanks upwards will therefore do nothing: by the time the engine starts looking at the top tank, it'll have already marked off the middle tank as "available" on its list, so the top tank will always appear higher on its list of fuel sources. It will not follow the fuel line to attribute higher priority to the middle fuel tank.

On the other hand, the lower fuel line (from bottom to middle tank) might actually force the next stage to draw from the bottom tank first... I have to say I'm not sure about that.

For curing colliding tanks, it's easy. Once you've placed your side stack, just use the "offset" tool (default key = 2) and slide it down the stack a bit. You've already strutted at the base so it should work just fine. The mass of the side stacks will be significantly lower than the decoupler, so it'll be pushed out into the airflow when it ejects.

 

Edited March 29, 2016 by Plusck

[FullMetalMachinist]

2 hours ago, jakeb1993 said:

Okay, so this poses a new question, how the fudgecake do you build a top heavy rocket when the fuel tanks and engines reside at the bottom?

Put very nicely, and as Plusck said, that's exactly the problem. 

The short answer is: staging. If you have a long, skinny rocket with the bottom 1/3 as your first stage then as it's fuel burns, that stage as a whole will become lighter. This moves your CoM forward, increasing your stability. One of the problems (that Plusck already said) is the way the tanks for that one stage drain isn't ideal. Follow his advice on that, it's right on.

The other problem is that if you have 3 or 4 linear stages all on top of the bottom first stage, then it takes a really big engine to lift all of that. That's where auxiliary boosters (either SRB's or asparagus LF-O) come into play. 

Looking at your most recent pic, it looks mostly ok, but there are a few problems. First off, the vertical fuel lines between stacked tanks aren't doing you any good. At least I don't think they are. What I would do instead is move the horizontal ones between radial stages to the top tanks. 

Second, when you mentioned that when you stage away the boosters they hit the center stack. I would move the decouplers down, and move the boosters down on the decouplers as far as you can go. Also make it so when you drop the first pair, that is the East/West pair, because as you start your gravity turn, the West booster is not "on top" of your rocket. So drop those first when you're still mostly vertical. I would also get rid of the fins on the boosters and just use the ones on the orange tank. 

And a side note related to something you said before about the TWR reading differently in VAB than on the pad. The Mechjeb readout is correct, you just have to look closer. There are two columns for TWR, one is TWR and the other is the SLT column. It stands for "sea level thrust" (I think) and it's the thrust that the stage will get when you take air pressure at sea level into account. 

[jakeb1993]

4 minutes ago, Plusck said:

No real need for the fuel lines on the side (by abundance of caution I understand, but I'm not sure it will give the result you really want...), but that's a detail.

For curing colliding tanks, it's easy. Once you've placed your side stack, just use the "offset" tool (default key = 2) and slide it down the stack a bit. You've already strutted at the base so it should work just fine. The mass of the side stacks will be significantly lower than the decoupler, so it'll be pushed out into the airflow when it ejects.

Well i Manged to get around some of it by cutting the engines and then letting them go. But the when its the central stage on its own it wants to flip, i had to fight it all the way up but once more when the engines cut it went into a flip. Its costing me nearly 5000 Dv as well to get into orbit with this thing too. So a mun mission is out of the question

[Nich]

Just an FYI it takes more DV to land on the mun then to land on duna

[Harry Rhodan]

2 hours ago, jakeb1993 said:

Okay, so this poses a new question, how the fudgecake do you build a top heavy rocket when the fuel tanks and engines reside at the bottom?

The easiest way would be using a dedicated upper stage like real rockets do. Asparagus and onion staging will always keep the weight towards the bottom and if you try to get to orbit with only your center stack  the weight will just drift down and the whole thing will fight you. What's the payload mass if I may ask.

Edited March 29, 2016 by Harry Rhodan

[Superfluous J]

2 hours ago, jakeb1993 said:

Okay, so this poses a new question, how the fudgecake do you build a top heavy rocket when the fuel tanks and engines reside at the bottom?

Don't lift a tiny payload with 4 Mainsails

[jakeb1993]

7 minutes ago, Harry Rhodan said:

The easiest way would be using a dedicated upper stage like real rockets do. Asparagus and onion staging will always keep the weight towards the bottom and if you try to get to orbit with only your center stack  the weight will just drift down and the whole thing will fight you. What's the payload mass if I may ask.

Total payload is 12.7 Tonnes. It's not like it is super heavy. I still need to put it in a fairing once my tests are done, but trying to build a 5500 Dv rocket to get to the Mun is proving hard.

Okay, so i did everything you said @FullMetalMachinist but sadly one side still collides and it is still flip happy, i had to fight it even harder as it climbed

[Harry Rhodan]

8 minutes ago, jakeb1993 said:

Total payload is 12.7 Tonnes. It's not like it is super heavy. I still need to put it in a fairing once my tests are done, but trying to build a 5500 Dv rocket to get to the Mun is proving hard.

With a 13t payload you can get around 5000m/s with one orange tank and a Skipper on top of two orange tanks, a Mainsail and two Kickbacks.