Building very heavy rockets

[Spricigo]

11 hours ago, si2504 said:

[at some time after launch, something] was causing flips

I see that you have an hypotheses. However, with the limited info you provided so far, it seems more like hopeful guesses  than an informed assessment of what is really happening.  I understand if you  prefer to figure it yourself but is not unlike that you are so focused in something that is not the actual issue and missing something other person would spot in a glance. The keyword there: spot.  a good picture of your craft in the VAB or the launch pad can go a long way toward a solution.

While KSP have a few visual cues, not everything is made explicit by those. There is no indication of Center of Drag or a easy way to figure out when the airflow velocity and/or angle of attack becomes too much for your controls to handle. As said:

2 hours ago, swjr-swis said:

Heavy stuff at the front, draggy stuff at the back, basically.

the trick part is to notice where the draggy stuff is.

 

[paul_c]

"Pointy end up, flamy end down".

Make the front pointy.

[Krazy1]

On 3/4/2021 at 2:59 PM, Spricigo said:

my issue is actually getting the damn things on the launch pad without them bouncing about and blowing up!!!

You need struts and you need enough of them in the right places. But you need the right mounting too. It helps to use beams on radial decouplers to help separate the boosters so they don't smash into the core when they detach. If you have radial stacks that are not detachable, then use I-beams to mount them. Butt-joints are not suitable for rockets this big. See how I did it for  >Kisyphus< I didn't use mods or rigid attachment or autostruts or even launch clamps (this time).  It takes patience and some iteration until you get good at it. Study working examples to get some ideas. (Honestly if I built this again, I'd use TweakScale just to make it easier and keep the framerate higher with lower part count.) 

 

 

 

 

[Spricigo]

I appreciate the advice, but...

1 hour ago, Krazy1 said:

On 3/4/2021 at 7:59 PM, Spricigo said:

Please, don't misattribute the quote. This thread is already confuse enough.

 

[si2504]

On 3/8/2021 at 9:09 PM, paul_c said:

si2504, this is a bit confusing because you've jumped on someone else's thread, and also not really given much info apart from your big rockets wobble. Then you've misattributed a shift of Centre of Mass forwards, with some kind of instability. I dare say, what is actually happening is that the weight/mass balance isn't the issue, but the drag is higher at the front than the rear. As speeds build, the aerodynamic forces destabilise the rocket (as swjr-swis mentions above).

It would help to post some pics of the design.

Of course, stage 2 needs to have its aerodynamics balanced as well as stage 1 (and stage 3, etc).......

Also, AFAIK KSP doesn't simulate "boil off" of fuel over time, tanks are modelled as perfectly sealed (unless you have a weird mod in place.....)

well all I know is I rebuilt the exact same rocket I was having issues with getting into orbit, with one that had a lower CoM and it flew like a dream. Also how does drag become an issue when you are out of the atmosphere? I will post a picture of the rocket that was causing trouble and see if I can demonstrate what was happening.

[paul_c]

It doesn't when you're out of the atmosphere, but you didn't say you were - you said by the time the first stage was empty.

Looking forward to the pics!

[swjr-swis]

4 hours ago, si2504 said:

Also how does drag become an issue when you are out of the atmosphere?

If your rocket is flipping once out of the atmosphere, that indicates either an offset thrust vector (and insufficient gimbal to compensate), or a control point that isn't pointing in the same direction as your first stage.

[si2504]

14 hours ago, swjr-swis said:

If your rocket is flipping once out of the atmosphere, that indicates either an offset thrust vector (and insufficient gimbal to compensate), or a control point that isn't pointing in the same direction as your first stage.

Tbf, the 1st stage engines always seemed to go crazy out of the atmosphere when the 1st stage was nearly empty. Sometimes I'd end up cutting my first stage short to avoid the craziness it caused, because a lot of the time due to the severe gimballing it was causing the rest of the craft to explode due to the instability. I can never understand it though as I always check that the thrust is going directly through the CoM, Guidance units are the correct way around etc. 

[paul_c]

Just rewind a bit.......

When you say "1st stage" - what do we mean? Are you launching from KSC, and if so how much deltaV is in the first stage?

The "traditional" amount you might use is  ~1000m/s if its SRBs, maybe a bit more say up to 1500-2000m/s if its a liquid fuelled 1st stage. If you're doing a gravity turn to build up horizontal velocity too, that won't get you out of the atmosphere.

From memory, my most recent launch with a ~1600m/s 1st stage decoupled it at around 25km. With the other stages on top and a sensible engine in the 1st stage, it might start out at TWR 1.3 and end up with TWR ~3 - certainly not enough to "go crazy" but a nice kick to get uphill. And I've always made my rockets more/less balanced, in terms of side-to-side mass variation. If not cylindrical, then obviously an amount of thought put into the design to balance the weight on each side to match.

In a gravity turn, or any launch really, you shouldn't need to be gimballing except when changing direction - and that should be minimal (it doesn't make sense to burn in one direction, change drastically, then burn in another).

[rynther]

Just BTW, here and there, the game will glitch, and stuff just won't work, sometimes you need to re-start the game.

Not sure if that has anything to do with the OP or not, but it seemed like something to mention.

The Rhino engines came up in conversation, and I gotta say, I haven't found a good use for them, for the big fat rockets, you're generally gonna be better off running vector engines, given the "mount anywhere" nature. My favorite for the 5m rockets is to put 9 vectors on it. (it would be interesting to know the actual payload weight and destination.)

[king of nowhere]

4 hours ago, rynther said:

Just BTW, here and there, the game will glitch, and stuff just won't work, sometimes you need to re-start the game.

Not sure if that has anything to do with the OP or not, but it seemed like something to mention.

The Rhino engines came up in conversation, and I gotta say, I haven't found a good use for them, for the big fat rockets, you're generally gonna be better off running vector engines, given the "mount anywhere" nature. My favorite for the 5m rockets is to put 9 vectors on it. (it would be interesting to know the actual payload weight and destination.)

for the glitch, you deal with it two ways. one part is autostruts, and the other part is saving often. yeah, with big ships you have glitches.

 

for the rhino, i never had much use for them before making the really big ships. but now, i use them often, they have a good vacuum Isp and a high thrust, making them good vacuum engines for large ships when wolfhound clusters aren't enough anymore.

[rynther]

19 hours ago, king of nowhere said:

for the glitch, you deal with it two ways. one part is autostruts, and the other part is saving often. yeah, with big ships you have glitches.

 

for the rhino, i never had much use for them before making the really big ships. but now, i use them often, they have a good vacuum Isp and a high thrust, making them good vacuum engines for large ships when wolfhound clusters aren't enough anymore.

As far as game glitches go, the size of the rocket doesn't really matter, here and there, you just can't stage a rocket, even a fairly small and simple one (less than 50 parts, under 100t to start) and generally it happens when the game has been running a long time. (8+ hours)

So far, I haven't felt the need for 100t+ transfer systems, but I don't have a network of stations that would be able to utilize the fuel. My current fuel taxi is something like 60t, so a wolfhound and a few srb's to kick it to high apoapsis is more than enough for my very small orbital base so far.

I think my only real complaint about engines, is that there are no true 5m engines,  making any 5m upper stage inherently unstable.

 

(I think the last time I played with super huge rockets was like 0.45 or so, and there wasn't really any goal other than how many tons it could get to orbit without exploding)

[rynther]

I've heard autostruts mentioned before, what/where are they, and what's the main difference to well placed regular struts?

[king of nowhere]

well, i've been running big missions with kerbalism. you need life support, and radiation shielding, and more electricity because it must keep life support going, and food and water and oxygen and other stuff, and everything can break so you need multiple copies of everything... well, you need big ships there, and big rockets to move them

7 hours ago, rynther said:

I've heard autostruts mentioned before, what/where are they, and what's the main difference to well placed regular struts?

you set them by right clicking on the part (you can do it in the vab but also in flight), they basically create virtual struts. they work more or less like regular struts. they have no weight, and you don't see anything. if they have any downside, i am not aware of it - except they glitch with robotic parts, because they hold still something that's trying to move, but that's a user mistake. 

for the really complex stuff, i put both autostruts and regular struts.

[rynther]

That sounds a bit like cheating

[paul_c]

I've never used autostruts - doesn't seem kosher when there's normal manual struts - even they do a little bit of magic when they magically detach from decoupled stages!

[rynther]

I can understand the attraction, especially for big multi-layer assemblies, but as a designer, I feel like it's a save weight easy button.

That being said, I got a pretty good run down on how they work, and I'll have to play with them a bit, apparently there's a weirdness to autostruts where it comes to docking ships, and the change in tension causing "kraken attacks",  sounds like a fun way to detonate an orbital base.

BTW, do any of you wreck stuff just for giggle and snort value?

[Curveball Anders]

1 hour ago, rynther said:

That sounds a bit like cheating

You can't cheat in a single player game.

Unless you mean cheat on your own rules.

[rynther]

37 minutes ago, Curveball Anders said:

You can't cheat in a single player game.

Unless you mean cheat on your own rules.

Something like that, clearly the devs put it there, so it's not outside the game boundaries, but then, starting your career with 5000 science points is also an option.  

I care more about building good rockets than manually flying everywhere, so I tend to be picky about the details, but I agree with the implication that there's no wrong way to play in a sandbox, other than perhaps throwing the sand at the other kids at the park.

[Fraktal]

4 hours ago, paul_c said:

I've never used autostruts - doesn't seem kosher when there's normal manual struts - even they do a little bit of magic when they magically detach from decoupled stages!

You could always just use both. I personally use autostruts solely to flex-proof fuel tank stacks. Core stack? Autostrut-to-root the bottommost tank in each stage to combat metronome flexing from gimbal and aerodynamics during the initial gravity turn. Radial stack? Autostrut-to-grandfather on whichever tank the engine is mounted on to combat flexing under high thrust.

For attaching radial stacks to the core stack, I use normal struts.

[SunlitZelkova]

On 3/26/2021 at 4:58 PM, rynther said:

Just BTW, here and there, the game will glitch, and stuff just won't work, sometimes you need to re-start the game.

Not sure if that has anything to do with the OP or not, but it seemed like something to mention.

The Rhino engines came up in conversation, and I gotta say, I haven't found a good use for them, for the big fat rockets, you're generally gonna be better off running vector engines, given the "mount anywhere" nature. My favorite for the 5m rockets is to put 9 vectors on it. (it would be interesting to know the actual payload weight and destination.)

I'm not sure if this was directed to me or not, but I will answer just in case.

The payload is the LK-700 (which includes the capsule, landing gear, TLI stage, and lunar crasher stage) replica for a Munar landing. Mass is 94 tons. I have tried the things people have suggested here but it doesn't want to fly, even with Mainsails.

I have also tried mounting it (LK-700) on different rockets (like my smaller design based on the UR-700) but it doesn't work.

Perhaps it is time to sacrifice accuracy and just build something regardless of design, so long as it can lift the LK-700?

[king of nowhere]

13 hours ago, rynther said:

That sounds a bit like cheating

well, i do agree on a certain level, and i make a point of using some manual struts even if autostrutting would be more convenient.

however, i still use autostruts when it's necessary. mostly for docking large ships, as those tend to blow up after docking if you haven't autostrutted them. also, when i use large ships, to reduce part count; when you are using a 1000-part ship lag is a problem, my cpu can take all the help it can get. and in such big ships, the few quintals saved on struts won't make a difference anyway

[rynther]

1 hour ago, king of nowhere said:

well, i do agree on a certain level, and i make a point of using some manual struts even if autostrutting would be more convenient.

however, i still use autostruts when it's necessary. mostly for docking large ships, as those tend to blow up after docking if you haven't autostrutted them. also, when i use large ships, to reduce part count; when you are using a 1000-part ship lag is a problem, my cpu can take all the help it can get. and in such big ships, the few quintals saved on struts won't make a difference anyway

This is a sandbox, you make whatever castles you like my man. 

I make satellites with rosettes of shrimp missiles to blow up other satellites, just because the mood strikes me to do it.

[rynther]

2 hours ago, SunlitZelkova said:

I'm not sure if this was directed to me or not, but I will answer just in case.

The payload is the LK-700 (which includes the capsule, landing gear, TLI stage, and lunar crasher stage) replica for a Munar landing. Mass is 94 tons. I have tried the things people have suggested here but it doesn't want to fly, even with Mainsails.

I have also tried mounting it (LK-700) on different rockets (like my smaller design based on the UR-700) but it doesn't work.

Perhaps it is time to sacrifice accuracy and just build something regardless of design, so long as it can lift the LK-700?

Lifting 94 tons should be pretty easy as a project, but you are trying to make a working replica of a specific rocket if I'm not mistaken.

A seven pack of BFSRB's is enough to ssto that much, depending on how "bulky" the scaffolding is, but it might be a bit too bumpy a ride. (Not specifically familiar with LK-700)

For motors, there are only 2 that have maximum ASL thrust, the mammoth, and (yeah, don't laugh) the vector. Why the vector? because you can cluster it big time, and the mammoth is the highest single engine thrust in the game. (Neither allow for use in second stage as described, as there is no decoupling option.) You can squeeze 7 vectors onto an S3 gas can, with room for 11 on the S4's (9 is generally enough) It has the highest thrust potential per surface area taken in the game, and gets top marks for ISP ASL. I would say try those 2 before scrapping the design.

 

[rynther]

Just for the giggle and snort value, 11 vectors strapped to the bottom of 2(ish) full size S4 tanks, at 620 tons gross vessel mass, has a TWR of 1.81, max acceleration of 17m/s, and 5168 ASL delta v. (raw tankage, no payload factor, but you get the idea)

 

Edit: Oopsie, wrong delta v number

Edited March 29, 2021 by rynther
Wrong delta v number