A very odd physics problem , work this one out.

[ASnogarD]

Yeah yeah I know I complain a lot, but after investing way too much time I think I have cause

I spent a 4 - 5 hours working on a lifter to get my test load of 145 tonnes into LKO, the lifter I made gets to a height of 200 000m straight up... I cant gravity turn because the craft develops a nasty spin when I detach my first stage which is a heavy 9 mainsail engine with 9 orange tanks of fuel, the 2nd stage is still controllable but the spin gets worse...and by the time its just the payload its completely unusable...even with mass RCS thrusters.

The craft is about 840 tonnes in weight, the payload is 143 tonnes approx. and to troubleshoot the spin problem I thought the top of my craft was causing some sort of drag, so I swapped out the Rockomax brand adaptor 02 for a Rockomax brand adaptor (the flat adaptor for the more streamline version)... my craft fell apart at the pad, not once ...every time, so I put the old adapter back on and the craft flew to 200 000 km but with the usual spin issue.

I swapped the adaptors again, and again the craft fell apart on the pad...

How much does that adaptor weight, hmmm... the flat one is 0.08 , the adaptor that causes my ship to fall apart is 0.1.

Aha you noob the extra weight is causing your ship to fall apart...stop raging at Squad and L2P...

... so I added the short big grey tanks...each at 4.5 tonnes, on the 3 orange tanks of the payload... in other words adding a massive 13.5 tonnes to my craft, this ship should fall apart for sure... *sigh* nope, craft was just fine... well too heavy to fly but didnt fall apart, so its not weight of the payload...

... changed the adaptor again and craft fell apart.

Why ?

Screenshot of the craft...

My point ? The calculations are so odd, they defy logical deduction to troubleshoot any craft other than a basic rocket (I cant say a thing about spaceplanes / planes as I have never been in the SPH). It seems cause and effect is not accurate... my other tri engine craft dont spin, and why does a particular adaptor in that position on the top of the craft cause pylons at the bottom to fail ?

Extra question: Why would the payload which has 3 full orange tanks attached in symmetry to the centre stack with a nuclear engine spin so much that even mass RCS cant stop ? There is no issues with the weight to one side...or if there was then the craft should flip rather than spin logically.

I am 90% sure its the payload that is causing the spin the whole craft experiences but the winglets keep the spin in check until thin atmosphere where the tri mainsail engine 2nd stage sort of keeps the spin manageable due to thrust vectoring... but the payload nuclear engine cant deal with it.

(Its spinning not flipping by the way)

PS : Sorry if I seem to complain a lot, I do spend a lot of time going over videos, checking other peoples designs, watching tutorials... my other craft fly well but I really wanted to make a relatively low part count heavy lifter to get at least 100 tonnes easily to LKO , as a standard to limit my station / base designs... currently I am limited to about 40 tonnes.

I use cross strutting, taking pains to keep symmetry as much as possible even with the struts, use grey tanks to prevent overheating, asparagus when possible, stitch strutting when needed, limit thrust and pilot the craft manually (MechJeb2 is there for the information, I like to see when I hit a certain target apoapsis without jumping to map mode)... mods is Mechjeb2, Kethane, Chatterer, Engineer and LSI alpha

[Whackjob]

Regarding spin: I believe you fell victim to the same thing that occasionally gets me. One or more struts may be touching or conflicting. This creates a torque as the physics engine tries to move them apart. They can't, since they're immoble, so you end up with a continuous spin that doesn't go away until the parts are shed.

My suggestion: Re-strut everything.

[ASnogarD]

I actually did pull a lot of struts out prior to trying to change the adaptor, the payload had struts attaching each tank to each other and from the top of each tank to the main body of the payload craft, and the bottom of the tanks to the body as well... I took the all out and the tanks sagged away from the body so I added 1 set of struts from the body to the tank to stop the sagging... other than that set the payload has no other struts on it.

The struts you see now are from the lifter to the payload... ie when the payload is released there is only 1 set of struts and they only go the short distance from tank to body in a straight line.

I guess a test sans payload would eliminate the payload as the culprit but I am pretty sure it is the cause of the spin... but it is the odd case of why does that adaptor cause my ship to fall apart ?

EDIT: Not disregarding Whackjobs advice, and will take that into account on future designs... may even re-re-design this one with that concept in mind.

Edited July 8, 2013 by ASnogarD

[JaySmoka]

... One or more struts may be touching or conflicting. This creates a torque as the physics engine tries to move them apart. They can't, since they're immoble, so you end up with a continuous spin that doesn't go away until the parts are shed.

That's an intersting statement... I was wondering why some time my rocket roll and other time not... I use the same designs for my rocket, just adding fuel tanks and engines to have more weight. I'll try restruting.

[Johnno]

Just from that picture I'd say you need more control surfaces, more struts, more RCS and more boosters. You're at a very optimistic 17% payload to total craft mass with a single (?) nuclear engine for orbital insertion. When you say 'mass RCS' I'd expect to see 6 or 8 RCS thrusters at the top and bottom of your core, possibly a few extra on the outer tanks as well (try turning a full red tank in space with just two RCS blocks... it's not going to be easy). You simply won't have enough RCS thrust to turn that thing properly.

I'd suggest triple up on winglets as well, possibly quadruple, I can only see 3. And struts too. Also aim for around 4000-4500 m/s delta-v prior to releasing your payload.

If you can't turn in atmosphere you don't have enough control surfaces, which is why I mentioned more winglets. More engines with gimbal would also do the trick. I often lift heavy stuff and it doesn't really matter if it spins, just keep an eye on your navball and keep manouvering in the right direction no matter the spin.

[Whackjob]

That's an intersting statement... I was wondering why some time my rocket roll and other time not... I use the same designs for my rocket, just adding fuel tanks and engines to have more weight. I'll try restruting.

Aye. Sometimes all it takes is for one strut end to touch another.

#EDIT: I put this mammoth station in orbit as a single chunk (except for command pod docked at the end of the center tank) and did it without control surfaces or even RCS!

I prevented the collision-roll by hand stitching the struts on that bad boy. One at a time. Slowly. Onerously. But the payoff is there.

Edited July 8, 2013 by Whackjob

[FITorion]

I only see one asas... and no sas to provide the torque needed.

I also see almost no RCS... not massive RCS... woefully inadequate RCS.

[ASnogarD]

The RCS is there to try arrest the spinning in orbit, not turn the craft... I am trying to troubleshoot the spinning issue.

The craft is fine until 1st stage separation (about 15 000 m at about 400 m/s ) , then its a slow and annoying spin until my apo is at 200 000 m , then I separate the 2nd stage from the payload... the payload spins out of control and blows 100 units of mono prop without managing to stop the spin.

So I tried to adjust the payload by messing with the struts, didnt help so I tried changing the payloads profile... which caused the odd issue of the adaptor causing the ship to fall apart.

Power to orbit is fine, given I can perform a reasonably decent gravity turn there should be enough delta V to get the payload to a point it can circlise the orbit... it has plenty of fuel even without using the cargo.

The issue is the spin on the payload, which is what I cant work out... the payload is balanced as far as I can tell with the mass of the 3 orange tanks spread out evenly, and the bottom of the payload craft is heavy with a large grey tank and nuclear engine.

The craft isnt behaving in a manner that would indicate its top or bottom heavy, and I cant see how one side can be heavier as the 3 tanks are symmetrically attached (unless having odd number of tanks would cause spin in orbit ? Why ?). The spin isnt manageable at all...its like if I had attached a engine horizontally and fired that at full thrust (obviously I didnt).

EDIT: Replying to FITorion : there is 18 blocks of RCS thrusters on the payload alone , there is also a capsule with a Kerbil inside so that provides torque for the payload (usually) so SAS would be redundant in this version of the game (I would imagine it is ?).

... and still doesnt explain why changing that adapter would cause my craft to fall apart on the pad without even firing the engines.

Edited July 8, 2013 by ASnogarD

[Whackjob]

Collision spin. It has to be collision spin. If I were at home and not the office, I'd have you send me the craft file to troubleshoot.

[ASnogarD]

Well knowing your penchant for struts the craft would probably look like a spider shat on it, and kill my PC loading the result into the VAB (just kidding... probably ).

[Johnno]

If it's the payload alone that spins then it's a glitch (collision issue like Whackjob mentioned). Rebuild your payload. But you're talking about several different issues. You mentioned not being able to do a gravity turn because of spin but then say it's slow and annoying. Then say that it's the payload alone that spins too fast to stop. Two entirely different issues. And the craft falling apart depending on an adapter swap can also be a collision issue alternatively just introducing a weak spot somewhere.

For a more thorough troubleshooting by others on the forum you'd have to make your craft file available to download. We can give you advice but we can't troubleshoot your craft without clearer information. Your "why does it break?" question isn't something we can answer simply by looking at one picture, we'd have to build the ship ourselves if you don't provide the craft file to see exactly where it breaks and then figure out why. At which point you'll likely be told 'moar struts'.

[arq]

Another thing that can cause spin is not enough struts on outer stages. The thrust of the engines can cause them to twist slightly and cause unstoppable spin. However, if you have cross-strutted as much as you say, that is probably not the issue. Usually if I put one strut at the top of 3 x32 tanks and one at the bottom, going inward and with a mainsail at the bottom that is usually plenty to keep stuff stable, if I add a cross-strut every other ring.

[Whackjob]

Came into the thread to help, left with my first forum signature. I call that a win!

[ASnogarD]

I hope this works... dont usually upload files

http://www.megafileupload.com/en/file/433193/Lifter-with-Fuel-Proto-WIP-craft.html

I taken Mechjeb2 off, and there shouldnt be any mods on there but installed on my client is : Mechjeb2, Chatterer, Kethane, Engineer.

If you are going to fly it, 1 is the control group to disable gimbals and as its TWR is a bit over 2 I launch at 75 % thrust...any more and it drives its arse through its payload. It has no batteries or generators (all from the capsule), and 100 units of mono prop only... just to see if I could use RCS to stop the spin of the payload.

Its still a usable lifter, I mean granted I could just take off a orange tank, 2x symmetry it instead of 3x and that should be more than sufficient as a 100 ton lifter ...and about 274 parts total (IIRC) including the test payload so pretty slim on the parts for such a heavy lifter. I mean to cut it down a bit more too... taking struts off and checking if they are redundant or not.

Thats my goal, a relatively slim lifter that isnt part heavy (or rely on mods) that can lift a decent amount ... 143 was the upper limit, and it nearly does it.

I cant do a efficient gravity turn with the slow and annoying turn my piloting skill isnt too hot to manage all those corrections....and by the time I get to the payload the spin is just insane.

Its the weird issue with the adapters that really stumped me... and made me wonder how many of the illogical structure fails were caused by 'weird' issues.

[Johnno]

Well I gave your lifter a look. First off I gave it a few test flights, it really needs more struts. A lot more struts. The hardpoints seemed to be the most prone to failure, I'd suggest swapping those for radial decouplers. I tried the adapter change as well and only a single hardpoint failed, but I noticed something in doing so. Your launch stabilizers are mounted way too low. You should at least mount them at equal height to the hardpoints/decouplers, otherwise the entire weight of the rocket is resting on the few hardpoints that have stabilizers on them.

As for strutting, it didn't seem like you had stitched the small gray tanks to the jumbo tanks, that will also induce shaking. Locking gimbals on outer engines was a good start but the inner engines are too unstable without more struts.

And once I'd strutted it properly I gave it another shakedown flight, gravity turn was fine even though with a bit of spin, payload did spin while under acceleration but the RCS could actually handle it. Think there's some imbalance there, probably because of your tanks swaying since they're not lined up in symmetry nor strutted properly (and ASAS overcompensating). Ultimately, the lone nuclear engine really doesn't have thrust to get you into orbit. You need more delta-v in earlier stages if you want the whole payload to reach orbit.

More struts, more boosters.

[Whackjob]

MOAR STRUTS

[Johnno]

Noticing the launch stabilizers being low got me thinking so I did some more testing. The adapter swap isn't actually about mass being an issue, it's height. I simulated the height of the adapter while using the mass of the lower adapter, a small I-beam, 2 octo struts and to make sure it wouldn't wobble to cause more issues I attached 4 struts between the RCS tank and the adapter below. Lo and behold, I can actually see the struts on the tanks snapping, one by one, without touching the rocket while it's standing on the launchpad.

So actually, the 'odd physics problem' is in fact by your own design. It's actual physics at work. The added height just below the RCS tank and pod at the top gives them enough extra force when swaying in the wind that it's enough to start snapping the struts on your tanks at the bottom because the rocket isn't rigid enough. The combination of the extra height and extra mass from the taller adapter, no matter how insignificant they may seem, are in fact perfectly matched to hit a destructive resonance on your rocket. It's purely down to your design.

Isn't physics fun?

Edit: I did a second test to confirm, I had the same setup as in my previous test but added a couple small batteries to get the mass up to 0.1 instead of 0.08. I got the exact same result as with just swapping adapters, a single hardpoint failed dropping a tank+engine stack.

So there you go, mystery solved. The extra height and mass, even if seemingly insignificant, breaks your rocket.

Edited July 8, 2013 by Johnno

[strongest_2hu]

The short adapter is terrible and always clips into the stuff around it and you should never use it.

[ASnogarD]

First up Johnno , thanks for taking so much time to try help out.

I agree the adaptor issue is due to height, but not because its is swaying in the wind (I dont believe wind is even in the game), I believe its more a force that cascades from the top of your craft downwards adding more force to the connections in a unrealistic simulation of taller rockets exert more downward pressure on its components.

Unrealistic because it seems to have 'tiers' of height, and doesnt seem to take weight into consideration.... basically if you cross into a new height 'tier' you will get more pressure no matter what structure or how heavy the structure was that got you to that height.

This would explain why the connections were 'fine' when I added 13.5 tonnes extra to the craft in a manner that did not raise the overall height of the craft.

I did strut the small grey tanks to the orange tanks before but while that stopped the engines from doing the samba on the launch pad, it also caused the stack to fall off on the pad, so I removed those struts... I spent a few hours trying to fix the design without resorting to massive strutting of everything.

Believe me I wasnt like 10 min design and toss my hands up and come rant, I spent hours... add struts here, launch, observe the effect, remove strut if ineffective, repeat.

I think Squad will either have to strengthen the connections the decouplers / structural pylon (engine mount part) to unrealistic proportions to offset the 1 connection limitation, or work out how to allow for multiple connections (docking ports dont count)... and Squad definitely has to fix the connections between components. A Rocket doesnt typically act like a tower of plates a waiter/waitress is hauling off to get washed.

Would be nice if Squad could also add a method to watch replays of launches, and visually show stresses on the craft as it flies so the player could look at the replay and study why the craft fails... getting told a link failed between a orange tank and some other tank on a rocket with something like 18 orange tanks doesnt help that much. Think of a colour overlay over the craft in flight that shows green for structurally intact going through yellow, orange , red colours showing the crafts stress points.

Note: This isnt a rant at Squad, the physics are a mess due to PhysX which is just meant to simulate debris in explosions, not calculate Newtonian laws in real time. The part connection issue is also understandable as I dont think Squad intended us to make such large unusual designs... the connections would be mostly fine (except the wobbly rocket syndrome) if we stuck to standard rocket design doctrines.