The only thing thing that I care to see in .24

[CleverClothe]

Real rockets do not have struts inside the tanks. The tanks are just made out of something stronger than aluminum foil, and are attached to things with more than a bit of rubber cement.

Note the ribbing is not a support structure, it is a baffle to help combat slosh.

Here's another drawing, this one's not the whole craft but shows the makeup of the first stage tank.

From your last image 1. and 7. Hundreds of ribs to support the rocket between the fuel tanks (the fuel tanks themselves are quite strong) and large rings to connect and support the weight above.

Now, due to the limits of Unity, KSP rocket components have a single very strong point of contact. Real rockets have dozens to hundreds set in a ring. But even those are not needed in KSP unless you have a very weak component or a very heavy load.

[PDCWolf]

From your last image 1. and 7. Hundreds of ribs to support the rocket between the fuel tanks (the fuel tanks themselves are quite strong) and large rings to connect and support the weight above.

That's why we don't need struts. Those things are already supposed to be inside the tanks.

Now, due to the limits of Unity, KSP rocket components have a single very strong point of contact.

You mean the same limitation that lets us only dock 1 docking port at any given time? oh wait, we can connect 2 parts by any number of docking ports at the same time.

[draeath]

From your last image 1. and 7. Hundreds of ribs to support the rocket between the fuel tanks (the fuel tanks themselves are quite strong) and large rings to connect and support the weight above.

Now, due to the limits of Unity, KSP rocket components have a single very strong point of contact. Real rockets have dozens to hundreds set in a ring. But even those are not needed in KSP unless you have a very weak component or a very heavy load.

Those are slosh baffles, are they not?

[KSK]

Those are slosh baffles, are they not?

I think the horizontal rings inside the tanks are the slosh baffles. If you look at the space between the tanks and the outer skin though, it's nothing but struts (in KSP terms). This shows the construction of the first stage in a bit more detail for folks that are interested.

If you think of KSP rockets as badly skinned rather than over-strutted, they become a bit more realistic. After all - these are kerbal rockets. Why bother putting that heavy, speed sapping metal skin around a rocket stage when you're going to ditch the thing 2 minutes into the flight!

I'm also with Sal on this one - my designs don't tend to need many struts for whatever reason, especially after the inline remote guidance components were added (before that I relied on a probe core in a cage of struts). Maybe I'm just building too small!

Edited January 4, 2014 by KSK

[Seret]

Wobbliness is half the fun.

[PDCWolf]

If you think of KSP rockets as badly skinned rather than over-strutted, they become a bit more realistic. After all - these are kerbal rockets. Why bother putting that heavy, speed sapping metal skin around a rocket stage when you're going to ditch the thing 2 minutes into the flight!

I'm also with Sal on this one - my designs don't tend to need many struts for whatever reason, especially after the inline remote guidance components were added (before that I relied on a probe core in a cage of struts). Maybe I'm just building too small!

Wobbliness is half the fun.

Overall, I'm convinced the obsession with disasters and perception of Kerbals as worthless engineers only caring about explosions is destructive for the game. KSP deserves much more than being a glorified disaster simulator where rockets falling apart and crews being killed is the prime entertainment and the only expected result. The achievements of players who strive to be successful, who create beautiful, well-engineered, reliable designs, should never be devalued, and the opinion that going to space is impossibly hard deserves to be crushed and disproved over and over again. Kerbals are capable engineers and it's up to the player to utilize their technology well.

This same mindset is harming the game in many other areas as well. The bugs of the physics system aren't there because we thought they would be fun and don't deserve to be defended as some players surprisingly do.

yeah, message too short.

[kzauner]

No. Stitch-strutting to combat what should be a strong joint (given that we're slapping fuel tanks together in a game abstraction instead of properly engineering each stage as a single structure) is in no way a reflection of my building and flying skills, it is a reflection of how bad the game handles those joints.

I think that the game models real-world physics not so badly. I see your point here, but to be honest - the problem are not the joints per se, to my mind. Most problems arise from large-small-large diameter constructions (in 'real world' there would be a load bearing hull-structure). This is one example, where i see no other way than simulating this structure with struts (this is btw. the reason I see use of the poodle-engine - she saves me often from running into strutting-out-issues).

Other Problems are just construction-issues. Large things wobble; especially with the forces we are talking about. This not really a game-issue, you can see the same IRL with bridges, towers, sky-scrapers, planes and lot of other stuff.

Too badly the editor has no information-display concerning lateral forces, maximum loads, and the like. And concerning vertical load something like max-g load (one has to find out about these things, and guessing future constructions). Especially the Mainsail is a beast here, throttle-down is a must if you don't want to go crazy with strutting. Again, this very real-life-esque. Shuttle e.g. had a max-load around 3g, IIRC. And there was lot of throttle up/down during ascend (to around the lowest setting of about 2/3). And guess what: since i follow similar regimes, my rockets started to be a lot more stable.

Mainsail liquid boosters are an example where i say yes to struts (as a rule of thumb i use 2 per booster in a V-shape for taking the load at the bottom). I see no real problem to reach around 100t LKO capcity; which is around the same capacity which has proven doable on earth.

[CleverClothe]

You mean the same limitation that lets us only dock 1 docking port at any given time? oh wait, we can connect 2 parts by any number of docking ports at the same time.

Not docking ports, the connection joints between parts that are snapped together. In the editor, these are highlighted by a green and black sphere.

[ddavis425]

Wobbliness is the biggest problem with the game IMO. For those of you saying "it's part of the KSP charm :)", why would you come to accept a long-standing critical bug because it amuses you? It's not a feature, it's a major issue.

This thing has loads of struts all over it and is using KJR.

Not docking ports, the connection joints between parts that are snapped together. In the editor, these are highlighted by a green and black sphere.

As far as I know docking port connections are the same thing as parts connected in the editor. The only difference is that you can bypass the parent-child part tree system with docking ports.

Edited January 5, 2014 by ddavis425

[epicbuilder333]

Exactly how big do you want your rockets?

http://imgur.com/a/ouzzA

SWEET Jesus, how do those things flight!?!?

[agentKmurph]

I still stand by my statement. I think part of the problem is that KSP rockets look OK in the editor when you snap them together, so people think they will work just like they look. But real rockets have a lot of strutting that is built in to their structure. That is not something that is obvious until you try to build rockets yourself. Then when your rocket fails, the game doesn't give you a lot of information or help about what to do.

I went through this same process when I picked up the game.

"real rockets have a lot of strutting that is built in to their structure." IN TO - Right, the structure is inside. Why then do we have to recreate it externally with struts? The superstructure should for the most part already exist. I'm not inclined to the "modular parts are why joints are so weak" argument because that methodology is only used to make this into a game rather than a CAD program. I hope I am not to far out of line assuming that these parts are being properly bolted, welded, and riveted to each other.

Now radial parts, odd placements, extremely massive stuff. Of course that should all require strutting. But a simple linear Atlas, Titan I, Ariane, Saturn, Redstone, or Delta style booster shouldn't bend and oscillate like a tree in the wind. That's really my only point here.

And of course real rockets have a lot put into their structure; I don't think that knowledge is quite as esoteric as you make it sound.

Since we are limited to only attaching at a single point, you need to add struts to bind those three tanks at the bottom together. I would also relocate that reaction wheel at the top to some point in the middle. This will reduce the bendy behavior quite a bit. Personally, I'd also remove that second mid engine and coupler and just make that a two tank stage. Should be a net gain in DeltaV (I have not done the maths, just saying.)

The bottom stage getting weird spread or oscillations isn't a problem since I am using the non thrust vectored LVT30 engine and 1.25m tanks. It most bends at the decoupler between the first and second stage. Removing the third engine would be a theoretical dV gain but the TWR would be unacceptably low given that the first stage drops at the start of the gravity turn. I'll try the reaction wheel relocation though, thanks for the tip.

[PDCWolf]

Not docking ports, the connection joints between parts that are snapped together. In the editor, these are highlighted by a green and black sphere.

The only difference between how docking ports connect and how parts connect is that the docking ports overcome the really badly planned tree hierarchy and their connections are created dynamically, something that wouldn't be too difficult to translate into the editor scene with tanks attached via multiple nodes. Or you can just use a stronger single joint like KJR does.

As per internal structure, one assumes tanks and whatever the hell they are connected to are joined by correct procedures like welding, screws, or that multiple tanks just create a bigger one instead of thinking that they are joined by weak spaghetti magic. Wobble is a bug, not a feature. On top of that (And implying it is not a bug) it's a non-intuitive mechanic that you come to realize when it's already too late. Once again iit's-not-a-feature, it's a bug that must be squashed.

[Sof]

It is my experience that a well designed rocket in KSP is rather stable, even without many (if any) struts. In any case, it is a lot better now that it used to be.

The rockets definitely shouldn't wobble when stationary on the launch pad. There is a limit to how much you could reduce that by before just letting anything with an engine and fuel get into orbit, but structural weakness shouldn't be due to bugs.

[John FX]

Wobbliness is the biggest problem with the game IMO. For those of you saying "it's part of the KSP charm :)", why would you come to accept a long-standing critical bug because it amuses you? It's not a feature, it's a major issue.

http://i.imgur.com/mAvfZ0c.png

This thing has loads of struts all over it and is using KJR.

As far as I know docking port connections are the same thing as parts connected in the editor. The only difference is that you can bypass the parent-child part tree system with docking ports.

This thing looks very similar to yours but is rock solid and is not using KJR. It has struts of course but as all the weight is central it doesn`t need whackjob levels. It doesn`t turn very fast but once stage one is done it`s fairly nippy.

[ddavis425]

This thing looks very similar to yours but is rock solid and is not using KJR. It has struts of course but as all the weight is central it doesn`t need whackjob levels. It doesn`t turn very fast but once stage one is done it`s fairly nippy.

Yours is stable because it doesn't get top-heavy, that rocket I posted has a center of mass right between the first and second stages even fully fueled. KSP physics doesn't like top-heavy rockets.

[OtherDalfite]

I completely agree with the OP. Wobbliness is one of the factors that keeps me from having fun, and many others in KSP. The game would be much better if the idea that rockets falling apart even when they are designed just like real ones. We shouldn't have to design our rockets to conform to KSP's physics, we should have to design our rockets to conform to REAL physics, or at least something remotely close. In KSP's current state, you are just stitching masses together. The larger masses, the more instability you have (ie: wobblieness).

[CleverClothe]

"real rockets have a lot of strutting that is built in to their structure." IN TO - Right, the structure is inside. Why then do we have to recreate it externally with struts?

Due to the aforementioned single contact points (joints) between parts. If this were LEGO, you could connect the pieces of your construction at many points and provide cross support. You can do the same in KSP, but it is done by a dedicated part (the strut). This is done to limitations of the game engine (i.e. hierarchical part tree and single points of contact).

Yours is stable because it doesn't get top-heavy, that rocket I posted has a center of mass right between the first and second stages even fully fueled. KSP physics doesn't like top-heavy rockets.

REAL physics doesn't like those type rockets too much either.

[Whackjob]

Struts = Happy

Or something.

[LostElement]

I hate wobbly rockets. I use Kerbal joint reinforcment, but sometimes it isnt enough. Struts dont even exist in IRL rockets. It's all super bolted or welded together.

[1096bimu]

I hate wobbly rockets. I use Kerbal joint reinforcment, but sometimes it isnt enough. Struts dont even exist in IRL rockets. It's all super bolted or welded together.

and they would break if not controlled with extreme caution. Not like in KSP where you could do a couple of flips during gravity turn and still enter orbit intact.

[Guest]

The main problem (which KJR fixes) is not joint strength. It's joint rigidity. The rockets are wobbly, because the connections, instead of being rigid, are flexible. A real rocket doesn't bend much under force, but at some point, it snaps. Which KSP is perfectly capable of simulating. Structural failures can and do happen with KJR. It's just that they happen where you'd expect them, when you would expect them. That means under high aerodynamic loads and too strong control inputs.

Another problem is lack of proper aerodynamics, which means we can do flips and violent turns without any problems. Using FAR, you do need to precisely control the rocket, lest the aerodynamic forces tear it apart, KJR or not.

[Wahgineer]

I should be able to build a Saturn V and, like NASA, not have to worry about it becoming a Giant Flying Spaghetti Monster. Tall, spindly rockets shouldn't need a load of struts just to attempt to stay together. The tanks should be more than enough to hold it all together. Like Bac9 said, a bug is not a feature to be defended, it is one to be squashed.

[John FX]

Yours is stable because it doesn't get top-heavy, that rocket I posted has a center of mass right between the first and second stages even fully fueled. KSP physics doesn't like top-heavy rockets.

You need to address your ship design then in that particular case. Don`t ask for the game to be changed so you can fly bad designs.

I should be able to build a Saturn V and, like NASA, not have to worry about it becoming a Giant Flying Spaghetti Monster. Tall, spindly rockets shouldn't need a load of struts just to attempt to stay together. The tanks should be more than enough to hold it all together. Like Bac9 said, a bug is not a feature to be defended, it is one to be squashed.

Apollo had to have a big thrust plate for example that the engines pushed onto for the first stage to stop the engines ripping the ship apart. I have to do the same. This is only right. If you do build like NASA then you will not end up with a top heavy rocket for starters. If you do it will not be tall and spindly either. You will either have short spindly rockets or tall fat ones.

The only top heavy ones there are the atlas V and the SLS and the SLS looks more squat than mine. One more thing to note is the slightly different first stage tank sizes between the 70T SLS and the 130T SLS. I`ll come back to that.

This has zero struts, is as tall and spindly as I could make it (TWR=1 on the pad) and is rock solid. If your rocket is taller or more spindly it is because you have designed it that way.

[imgur]rQovf[imgur]

You can always break stuff in a physics game. Don`t ask for the game to be changed so you can fly bad designs.

Having said that, I agree with Op that some joints need looking at and maybe the whole joint system. (detaching NERVA anyone?)

The elegant solution is to let us build rockets exactly like NASA and let us have procedural tanks, fairings and wings.

It would solve the problems of the OP without creating game imbalance as you would not need to stitch together 5-10 tanks for the first stage, you just have one which will not flex, bend or burst apart. Invisible auto stitching between vertical parts would eliminate unrealistic movement and voila. Problem solved.

On the plus side, your ship also has about 200 less parts as a result.

[Monkeh]

I see both sides of this argument and feel affinity to both as well. I find strutting to be a minor inconvenience, something that I'm just used to doing I suppose. It doesn't take long to turn on symmetry and stick a strut or two to the larger weighted parts of my rockets.

Building efficiently and small requires very little in the way of struts, it's the big rockets that need lots of them and that feels right to me. All that weight and all that thrust need reinforcement, you can't expect a fuel tank that weighs tens of tonnes and being subjected to a few KN of thrust, to sit, unmoving, when attached via a small piece of a decoupler! Thinking it should definitely isn't realistic either.

I found the invisible struts in the B9 pack, which also happen to be mighty strong, so it's become even less of an annoyance now, here's my lovely craft that used a fair few invisible struts but flew like a baws.

Here's my minimal Mun land and returner which had 4 struts on the first stage tanks I believe.

I honestly feel the balance between annoyance and realism in this case is just about right. If you build big then you have to pay for it by spending a few minutes strutting it all. However, saying that, when pieces of my rocket fall off for very little reason, or the thing splits in half because the engine gimbal was too strong, it can be annoying.

For those who wonder why their engines wobble so much under the fuel tanks, turn off the gimbal. Less engine wobble will make things much more durable.

[numerobis]

The single point of attachment is not at all a limitation of the game engine, the proof being that we can get multiple points of attachment by adding struts. It would be nice if node attachment were implemented by a ring of attachments. This is relatively easily implementable if you require the top and bottom part attachments to have the same diameter (which, bonus, means adapter parts become necessary rather than aesthetic). There's some runtime cost from adding a bunch of extra rigid body constraints, and it will make things work differently than they do now, so while it's easy to implement a demo it may take a while to make it work right.

Edited January 6, 2014 by numerobis