Why aren't my struts letting go?

[Loren Pechtel]

New rocket, I'm heading to Minmus for some science and my rocket won't stage.

Based on the discussion in my previous thread I strutted the strap-ons to the main body--and those struts are not letting go when the decouplers fire.  This design won't make it to Minmus with them on, let alone come back home.

MechJeb is confused by the rocket.  It's showing a TWR for stages 3 and 4 but only a delta-v for stage 3.

At least this rocket doesn't wobble.  That doesn't do much good if it can't do the job, though.

 

Edited February 18, 2016 by Loren Pechtel

[Vim Razz]

Are the outer columns supposed to be sitting on radial decouplers?  It looks like they might be directly attached to the inner tanks instead, which can be pretty easy to do with the skinny decouplers until you become totally paranoid about checking and rechecking them.

[Loren Pechtel]

3 minutes ago, Vim Razz said:

Are the outer columns supposed to be sitting on radial decouplers?  It looks like they might be directly attached to the inner tanks instead, which can be pretty easy to do with the skinny decouplers until you become totally paranoid about checking and rechecking them.

I had the same thought after flying it--I took the outer columns off and put them back on--no change.

I also tried taking the struts off--no change.  It's certainly like they're all stuck together but I sure can't see how.

[Snark]

I'd say that almost certainly the answer is what @Vim Razz said.  You're attaching tanks directly together, they're not going on the decouplers.

Also... any chance you could squeeze out a bit of science somewhere and unlock Fuel Systems?  That would be a big win for this craft-- you could use the FL-T800 fuel tank, which would cut your part count in half.  Also, it would give you fuel ducts, which would be a real game-changer-- you could asparagus those tanks and get a lot of extra dV.

One more suggestion:  why not put all 6 of your radial boosters directly attached to the central core, rather than in two layers like that?  Ought to be stronger that way.

Edited February 18, 2016 by Snark

[Loren Pechtel]

8 minutes ago, Snark said:

I'd say that almost certainly the answer is what @Vim Razz said.  You're attaching tanks directly together, they're not going on the decouplers.

Also, what sort of fuel-duct setup do you have on that ship? (or are ducts not unlocked yet?)  Hard to tell from the picture.

I just unlocked the ducts.  The outer columns feed the inner columns, the center is not connected (It's a terrier for use in space, I don't light it on the pad.)  That part is working correctly, I just need to get those spent boosters off the rocket!

Is there any good way to verify that things are properly attached to the decouplers?

[kotomikun]

17 minutes ago, Loren Pechtel said:

Is there any good way to verify that things are properly attached to the decouplers?

Usually Kerbal Engineer will show it all as one stage if they're improperly attached, because that's what it really is. Not sure about MechJeb. Nothing in stock helps there, unfortunately, other than test launches.

Remove each stage and re-attach it by clicking directly on the decoupler, with the little box near the mouse saying "radial decoupler." Then it should work. I don't think I've ever seen struts cause this problem, it just looks that way because the decouplers blow up but don't actually decouple anything.

[Snark]

1 hour ago, Loren Pechtel said:

I just unlocked the ducts.

Excellent, that means you also just unlocked the 4-ton tall, skinny tank, since that's on the same tech node. Suggestion: Replace each pair of 2-ton tanks with a single 4-tonner.  It'll reduce your part count by a lot, and significantly reduce wobble as well.  Will also help a bit with CoM issues.

1 hour ago, Loren Pechtel said:

The outer columns feed the inner columns, the center is not connected (It's a terrier for use in space, I don't light it on the pad.)

Fair 'nuff.  You could still mount all six of the radial boosters to the central core for greater structural strength.  Also, that would allow you to stage them off in three pairs rather than as two sets of three, which should eke out a bit of extra dV.

Also, you can set up the fuel ducts to help with (though not completely solve) the top-down drainage issue for better stability during ascent.  If stack A is draining into stack B, set up the duct so that it's hooked from the top of stack A to the top of stack B.  That way, stack B will be pulling fuel from the bottom of stack A, which helps keep CoM high.  It also ensures that stack B will use zero of its own fuel until stack A is completely empty.

1 hour ago, Loren Pechtel said:

Is there any good way to verify that things are properly attached to the decouplers?

A couple of ways to tell.

First, just eyeballing the separation.  The decouplers have a bit of "thickness" to them, so if the outer thing is on the decoupler rather than being directly attached to the inner thing, there should be about 30 centimeters (maybe? never measured it) of clearance between them.  In your screenshot above, it looks like your outer boosters are touching the inner ones, snugged right up against them, which is the warning sign.

Second, try mousing over the decoupler.  In the VAB (or SPH), whenever you mouse over a part, it highlights green, and so do all of its child parts.  So, if the radial booster is attached to the decoupler (i.e. is a child of it), then when you mouse over the decoupler, both it and the entire radial booster should highlight in green.  On the other hand, if only the decoupler highlights when you mouse over it, then it means nothing's attached to it.

Edited February 18, 2016 by Snark

[Loren Pechtel]

1 hour ago, kotomikun said:

Usually Kerbal Engineer will show it all as one stage if they're improperly attached, because that's what it really is. Not sure about MechJeb. Nothing in stock helps there, unfortunately, other than test launches.

Remove each stage and re-attach it by clicking directly on the decoupler, with the little box near the mouse saying "radial decoupler." Then it should work. I don't think I've ever seen struts cause this problem, it just looks that way because the decouplers blow up but don't actually decouple anything.

That was it.  It took three tries rebuilding the rocket before it "worked"--then when it came time to burn to Minmus the rocket went unresponsive.  SAS wouldn't line it up, I couldn't turn it off and it totally ignored any attempt to steer, I couldn't light the engine.  Full electric charge, it has a reaction wheel.  I've got a pilot, scientist and a tourist on board.

I did try to EVA in Kerbin orbit and was told the hatch was blocked.  (It certainly doesn't look like it to me.  There was a drogue chute near it but not over it.)  Nothing worked after that--could my pilot have ended up not at the controls even though he didn't get out of the capsule?

[Snark]

10 minutes ago, Loren Pechtel said:

That was it.  It took three tries rebuilding the rocket before it "worked"--then when it came time to burn to Minmus the rocket went unresponsive.  SAS wouldn't line it up, I couldn't turn it off and it totally ignored any attempt to steer, I couldn't light the engine.  Full electric charge, it has a reaction wheel.  I've got a pilot, scientist and a tourist on board.

I did try to EVA in Kerbin orbit and was told the hatch was blocked.  (It certainly doesn't look like it to me.  There was a drogue chute near it but not over it.)  Nothing worked after that--could my pilot have ended up not at the controls even though he didn't get out of the capsule?

By any chance is your tourist in the command pod, and your pilot/scientist in the crew cabin?  That would make it uncontrollable (tourist can't fly the ship) and your two crew members would be stuck with "hatch obstructed" (crew cabin doesn't have a hatch).

Though how it could have gotten off the launch pad like that, I don't know.

Certainly from the screenshot it doesn't look to me like your command pod's hatch is obstructed, it ought to be just fine.

You're not running RemoteTech by any chance, are you?

Edited February 18, 2016 by Snark

[Loren Pechtel]

39 minutes ago, Snark said:

Excellent, that means you also just unlocked the 4-ton tall, skinny tank, since that's on the same tech node. Suggestion: Replace each pair of 2-ton tanks with a single 4-tonner.  It'll reduce your part count by a lot, and significantly reduce wobble as well.  Will also help a bit with CoM issues.

Fair 'nuff.  You could still mount all six of the radial boosters to the central core for greater structural strength.  Also, that would allow you to stage them off in three pairs rather than as two sets of three, which should eke out a bit of extra dV.

Also, you can set up the fuel ducts to help with (though not completely solve) the top-down drainage issue for better stability during ascent.  If stack A is draining into stack B, set up the duct so that it's hooked from the top of stack A to the top of stack B.  That way, stack B will be pulling fuel from the bottom of stack A, which helps keep CoM high.  It also ensures that stack B will use zero of its own fuel until stack A is completely empty.

A couple of ways to tell.

First, just eyeballing the separation.  The decouplers have a bit of "thickness" to them, so if the outer thing is on the decoupler rather than being directly attached to the inner thing, there should be about 30 centimeters (maybe? never measured it) of clearance between them.  In your screenshot above, it looks like your outer boosters are touching the inner ones, snugged right up against them, which is the warning sign.

Second, try mousing over the decoupler.  In the VAB (or SPH), whenever you mouse over a part, it highlights green, and so do all of its child parts.  So, if the radial booster is attached to the decoupler (i.e. is a child of it), then when you mouse over the decoupler, both it and the entire radial booster should highlight in green.  On the other hand, if only the decoupler highlights when you mouse over it, then it means nothing's attached to it.

I didn't notice that new fuel tank--it looks small and I didn't realize it was actually big.

I hooked up the fuel about as you say (I wasn't quite at the top in either case, I didn't realize it mattered) and it drained the outer ones faster than the inner but both lowered.

I see what you mean about the hover once I try it.  I've got some color vision issues, while I can see the highlighting it's pretty subtle to me and I wouldn't have thought to call it "green".  Unfortunately, those decouplers are well buried, it's hard to pick them out.  The game really needs a better means of building stuff.

[Loren Pechtel]

25 minutes ago, Snark said:

By any chance is your tourist in the command pod, and your pilot/scientist in the crew cabin?  That would make it uncontrollable (tourist can't fly the ship) and your two crew members would be stuck with "hatch obstructed" (crew cabin doesn't have a hatch).

Though how it could have gotten off the launch pad like that, I don't know.

Certainly from the screenshot it doesn't look to me like your command pod's hatch is obstructed, it ought to be just fine.

You're not running RemoteTech by any chance, are you?

Yeah, if he wasn't in the command pod I wouldn't have gotten into orbit in the first place.  The controls went dead in orbit after the failed attempt at EVA and then setting up the maneuver node for Minmus.

I'm not running RemoteTech.  The only relevant mods are MechJeb (and I have yet to use it to fly anything, just for delta-v and having the orbit info panel open so I don't need to switch between the rocket and the map) and StageRecovery.

[kotomikun]

That's weird... Are you sure it had full electric charge? I don't see any solar panels on there. Also, this may sound dumb, but make sure you aren't timewarping; I often think my rocket's frozen up but it turns out I'm at 5x warp. If none of that works, returning to the space center and reloading the ship at the tracking station often fixes weird problems (sometimes one part will just refuse to move for no reason).

[Snark]

33 minutes ago, Loren Pechtel said:

I hooked up the fuel about as you say (I wasn't quite at the top in either case, I didn't realize it mattered) and it drained the outer ones faster than the inner but both lowered.

The precise location doesn't matter; what does matter is which tank.  If you have stack A and stack B, each of which has an active engine on the bottom, and there's a fuel duct going from the topmost tank of A to the topmost tank of B, here's how it should work, I think:

• A's engine will drain stack A from the top down, and will stop when stack A is completely out of fuel
• B's engine will drain stack B from the bottom up, using none of stack B's fuel until stack A is completely empty.  When the last stack-A tank is empty, stack B will then drain from the top down.

That's not what you observe?

[Loren Pechtel]

8 minutes ago, kotomikun said:

That's weird... Are you sure it had full electric charge? I don't see any solar panels on there. Also, this may sound dumb, but make sure you aren't timewarping; I often think my rocket's frozen up but it turns out I'm at 5x warp. If none of that works, returning to the space center and reloading the ship at the tracking station often fixes weird problems (sometimes one part will just refuse to move for no reason).

I'm using the flat panels, they're not obvious.  Charge was 450 out of 450.  I was definitely out of warp--I was watching the clock creep up on my maneuver node and trying to get ready.

2 minutes ago, Snark said:

The precise location doesn't matter; what does matter is which tank.  If you have stack A and stack B, each of which has an active engine on the bottom, and there's a fuel duct going from the topmost tank of A to the topmost tank of B, here's how it should work, I think:

• A's engine will drain stack A from the top down, and will stop when stack A is completely out of fuel
• B's engine will drain stack B from the bottom up, using none of stack B's fuel until stack A is completely empty.  When the last stack-A tank is empty, stack B will then drain from the top down.

That's not what you observe?

No.  I didn't look at the fuel levels in the individual tanks, just the fuel bars shown on the screen.  A drained a lot faster than B but B did drain.  I think I had about 3/4 in the inner ones when I jettisoned the outer ones.  Tomorrow I'm rebuilding with the bigger tanks and I'll see what happens.

[Snark]

8 hours ago, Loren Pechtel said:

No.  I didn't look at the fuel levels in the individual tanks, just the fuel bars shown on the screen.

Ah, then it's working. This is normal behavior for asparagus-type setups where you have two stacks, both with running engines, and there's a fuel duct running from one to the other.

Yes, the fuel bar display is kinda misleading. The B stack isn't actually draining, when you see this-- it has 100% fuel in it, even though its fuel bar doesn't show full. Its fuel bar shows partly empty because it can"see" the A stack's empty space through the fuel duct. It's behaving exactly as designed. Try clicking on the individual tanks and you'll see what I mean.

You'll also notice that when you stage away the empty A stack, the B stack's fuel bar snaps back up to 100% full, because you just ditched the empty tank space.

Edited February 18, 2016 by Snark

[AlphaMensae]

Another way to check if you have correctly attached stuff to the radial decouplers is to adjust the camera view in the VAB until you're looking straight down on the stacks and see if they really are attached.  Those basic TT-38Ks can be really finicky, and I always have to look down to see if things are attached right.   The TT-70s are so much nicer to use.  

[Loren Pechtel]

20 hours ago, Snark said:

By any chance is your tourist in the command pod, and your pilot/scientist in the crew cabin?  That would make it uncontrollable (tourist can't fly the ship) and your two crew members would be stuck with "hatch obstructed" (crew cabin doesn't have a hatch).

Though how it could have gotten off the launch pad like that, I don't know.

Certainly from the screenshot it doesn't look to me like your command pod's hatch is obstructed, it ought to be just fine.

You're not running RemoteTech by any chance, are you?

Stranger and stranger.  I rebuilt the rocket with the bigger fuel tanks and it flew like a charm.

That is, until it's time to leave orbit.  I've been collecting EVA data while flying over various biomes--and now I have the same problem: The rocket's controls are dead.

Could the EVA have shuffled the order of the people in there?  How would I tell?  And if so how can I get him back in the cockpit?

[Snark]

7 minutes ago, Loren Pechtel said:

Stranger and stranger.  I rebuilt the rocket with the bigger fuel tanks and it flew like a charm.

That is, until it's time to leave orbit.  I've been collecting EVA data while flying over various biomes--and now I have the same problem: The rocket's controls are dead.

Could the EVA have shuffled the order of the people in there?  How would I tell?  And if so how can I get him back in the cockpit?

You can tell who's in what pod by clicking the "Transfer Crew" button for each crewable part, and it'll show you the list of all the kerbals who are in that part.

If you're not running RemoteTech, and you're not out of electricity, then I'm running out of ideas.  Post a screenshot of the ship in its uncontrollable state?

[Loren Pechtel]

25 minutes ago, Snark said:

You can tell who's in what pod by clicking the "Transfer Crew" button for each crewable part, and it'll show you the list of all the kerbals who are in that part.

If you're not running RemoteTech, and you're not out of electricity, then I'm running out of ideas.  Post a screenshot of the ship in its uncontrollable state?

The pilot is in his seat:

and here's the whole rocket, admittedly quite dark as the tail is pretty close to Kerbol (but the batteries are fully charged) and I have no way to fix that:

 

Also, when images are attached are they being stored locally or pulled from the space I have them stored?

[Snark]

49 minutes ago, Loren Pechtel said:

The pilot is in his seat:

Also, when images are attached are they being stored locally or pulled from the space I have them stored?

Hrmmmm.  Yeah, clearly Jeb is in the command pod, and you've got electricity.  At this point I have no clue what could be going on.

Is it possible MechJeb could be up to some sort of mischief?  I've never run MechJeb myself, so I have no clue what its failure modes might be.  Is it possible to turn it off and see what happens?

About the only other thing I could suggest would be to pop open the Alt+F12 debug window and click on the "Debug" tab to look at the log and see if maybe something is throwing exceptions all over the place.  Other than that, I'm about out of ideas.

[Loren Pechtel]

1 hour ago, Snark said:

Hrmmmm.  Yeah, clearly Jeb is in the command pod, and you've got electricity.  At this point I have no clue what could be going on.

Is it possible MechJeb could be up to some sort of mischief?  I've never run MechJeb myself, so I have no clue what its failure modes might be.  Is it possible to turn it off and see what happens?

About the only other thing I could suggest would be to pop open the Alt+F12 debug window and click on the "Debug" tab to look at the log and see if maybe something is throwing exceptions all over the place.  Other than that, I'm about out of ideas.

Bingo!

[Error]: Serialization depth limit exceeded at 'Contract.Agents::Agent'.  There may be an object composition cycle in one or more of your serialized classes.

And the same thing with 'Contracts::ContractParameter' less frequently.  I guess I have mod that wigs out after my pilot goes EVA.  Time to disable things and see what happens.

 

Edit:  No.  The error still shows up with all mods removed.

Furthermore, after removing everything I found that while I didn't have SAS I did have manual steering back.  I then tried putting everything back--and I still had manual control.  I burned (for Mun, after redesigning I added a bit more delta-v to visit both worlds.)

About 2/3 of the way out my intercept orbit suddenly no longer was an intercept orbit and I decided to burn a little bit.  Oops--dead controls again.

Now I find that simply exiting to the main menu and reloading the game fixes the controls, still no SAS, though.  I'm slowing turning the rocket to prograde, it crosses the marker--and won't stop.  Dead controls again.  I try simply going to the space center and then back--controls work again.

Edited February 19, 2016 by Loren Pechtel

[Plusck]

It really does sound like a mod that's doing this. Specifically it sounds very much like RemoteTech - although I've never tried it. Are you really, really sure you don't have it installed? Like triply sure by going into your GameData folder and checking what each subfolder is...

In stock, if your design works once it will always work. Common reasons for "loss of control" are leaving warp on and not noticing, running out of electricity, and that's about it.

RemoteTech sets a limit on how far you can go while still retaining remote control. To go further you need a constellation of communications satellites and bigger antennae. Some other mods may do this but I'm really not knowledgable about mods in general... But from what you're describing something in the game seems to be applying this rule without you being aware of it until problems occur.

What I'd recommend is to make a clean unmodded install (while keepig the old one - if you use steam just move your game folder to a different location where Steam can't get at it), copy your saves folder across to the new one and see what the game says about unavailable parts when you try to return to a savegame.

People on this forum tend to be helpful, so you could let somebody else do all the work by putting your persistent.sfs file up on dropbox (or similar) and posting the link. ; )

[Loren Pechtel]

6 hours ago, Plusck said:

It really does sound like a mod that's doing this. Specifically it sounds very much like RemoteTech - although I've never tried it. Are you really, really sure you don't have it installed? Like triply sure by going into your GameData folder and checking what each subfolder is...

In stock, if your design works once it will always work. Common reasons for "loss of control" are leaving warp on and not noticing, running out of electricity, and that's about it.

RemoteTech sets a limit on how far you can go while still retaining remote control. To go further you need a constellation of communications satellites and bigger antennae. Some other mods may do this but I'm really not knowledgable about mods in general... But from what you're describing something in the game seems to be applying this rule without you being aware of it until problems occur.

What I'd recommend is to make a clean unmodded install (while keepig the old one - if you use steam just move your game folder to a different location where Steam can't get at it), copy your saves folder across to the new one and see what the game says about unavailable parts when you try to return to a savegame.

People on this forum tend to be helpful, so you could let somebody else do all the work by putting your persistent.sfs file up on dropbox (or similar) and posting the link. ; )

When I was thinking it was a guilty mod I took everything except the Squad folder out of the GameData folder.  There couldn't have been any mods installed at one point.

Furthermore, I've managed to take this craft into orbit about Mun for some time and now have just done orbital insertion over Minmus--it's gone dead about 10 times, every time going back to the space center fixes it.

And I seem to have run out of science space , the Minmus part of the trip is a total bust.  And talk about a wild return:  Something was wrong with the final decoupler, it fired but the booster didn't fall off.  I couldn't manage to orient it properly but I kept steering in the direction it was currently turning, rather than a deadly prograde it kept swinging around, spreading the heat.  Total destruction: the drogue chutes.  The mains survived, though, and landed it.  It of course fell over and exploded but both the command pod and the crew capsule were intact.  Everyone came home and I got almost 1000 science.

Edited February 19, 2016 by Loren Pechtel

[Snark]

22 minutes ago, Loren Pechtel said:

And I seem to have run out of science space , the Minmus part of the trip is a total bust.

Can you elaborate on that?  What do you mean by "run out of science space"?  It's true that the command pod is limited to holding only one copy of any given science result, but that's generally not a practical limitation because you don't get much benefit from having multiple copies anyway.  As long as each science result is a different one, there's no upper limit to how many science results you can stuff into a command pod.

If a purpose of this trip was to gather new science from Minmus, and you didn't already have those Minmus results on your ship, you should be able to stuff 'em in there with no problem.

What's the issue you're having there?

[Loren Pechtel]

18 minutes ago, Snark said:

Can you elaborate on that?  What do you mean by "run out of science space"?  It's true that the command pod is limited to holding only one copy of any given science result, but that's generally not a practical limitation because you don't get much benefit from having multiple copies anyway.  As long as each science result is a different one, there's no upper limit to how many science results you can stuff into a command pod.

If a purpose of this trip was to gather new science from Minmus, and you didn't already have those Minmus results on your ship, you should be able to stuff 'em in there with no problem.

What's the issue you're having there?

I'm not sure but when I re-entered the capsule I was told I had to dump the experiment and nothing was learned.  Was my orbit too high for biome-specific EVA reports, perhaps?  Anyway, I unlocked a lot of stuff and I'll go back with a hopper to get the surface data.