Help: correct way to use EAS-4 Strut Connector to increase strength of my rocket?

[tfabris]

When I make rockets that are fairly tall (approximately Saturn V size and shape... ish...), I keep running into a problem where the rocket tends to bend and sway during its initial ascent and gravity turn. For the first couple of stages, there can be enough oscillation to make it hard to complete the gravity turn accurately, or even cause the ship to rip itself apart.

I'm trying to use EAS-4 Strut Connectors in my design to add rigidity to the structure. But I have run into problems where the very pieces I'm trying to tie together, will refuse to tie together. When I place the end of the strut connector (the second click), the endpoint snaps back to a spot very near the start point. It's as if the construction system doesn't allow me to tie pieces that are separated by two seams.

For instance, maybe I want to use a strut connector to tie two stacked stages together to make them more rigid. I try to tie the fuel tank on top to the fuel tank below it. It won't let me, possibly because there's an engine and a stack separator between them. I can tie the stack separator to the fuel tank below it, but my rocket still bends and wobbles.

What am I doing wrong? Is there a different way to make my rockets more structurally sound?

[M5000]

The method you are using to tie your rockets together at the seams is, intuitively enough, called "stitching" by many of the community members here. It's a sound and common practice many users utilize to add structural rigidity to their rockets, though I too have experienced difficulty with getting them to attach sometimes.

Here's something to take note of: You should only need about 3x or 4x symmetry for optimal stability. Anything much over that is overkill and is more parts than you need. However, look at how you're placing the start point. If you place it in some places, where it is clearly above the surface of the part, it, for some reason, will not connect. For another, equally weird reason, putting the initial anchor below the surface of the tank, decoupler, part, whatever you're connecting it from will often get it to work just fine.

You may have to rotate them about the central axis a ways, as some positions simply aren't friendly. I believe this may have something to do with the "collider" meshes not lining up perfectly (read: having lower polygon counts) with the actual visual models you see, so some points are slightly below and some are directly on the surface.

Here.. Take a look at this image, I just now made it just for you.

Welcome to the forums, by the way.

-M5K

[tfabris]

Thanks, this looks like the information I'm looking for. I'm still not clear on how to get it to *do* what is pictured in the screen shots. Do I just experiment with different placements until I get lucky and it looks like the screen shot?

[sephirotic]

@M500 Your advice is very good, but it doesn´t help when we have a 2.5m decoupler on the middle, since they always block the way of the struts.

When making single column large and heavy vertical rockets, i recommend doing this:

Works flawlessly.

But, lately i thought: Well, why the hell not just add some extra lateral fuel tanks and use then as intermediate structural linkage for struts?

Like this:

Also, is important to know that struts only "transfer" strength, it doesn´t create firmness in nodes that doesn´t have.

I mean, there is no point in doing something like this:

If individually each booster is instable by themselves, you will not make then more strong. Is interesting to always connect strut indirect between two pieces that are connected together by a third piece.

Edited September 16, 2013 by sephirotic

[M5000]

@M500 Your advice is very good, but it doesn´t help when we have a 2.5m decoupler on the middle, since they always block the way of the struts.

*snip*

Hmm.. Odd.. I usually can find SOME way to get it to behave and eventually connect, no matter the situation... Ah yes! You know what I do? I go strutting overload and attach to the decoupler, then to the next part. A bit hard to do, but certainly possible and it is structurally sound.

Also, is important to know that struts only "transfer" strength, it doesn´t create firmness in nodes that doesn´t have.

I mean, there is no point in doing something like this:

*snip*

If individually each booster is instable by themselves, you will not make then more strong. Is interesting to always connect strut indirect between two pieces that are connected together by a third piece.

I might argue with you a bit there as well. While you're definitely correct in saying that a strut placed as you have shown will not make it less likely to fall off, I have observed sometimes, with very large boosters with high thrust, if you are doing several radially, they may (not always) begin to resonate back and forth and eventually tear loose, even if the thrust vectoring is disabled. I don't know what causes this, but adding struts between boosters can dampen lateral oscillations.

Thanks, this looks like the information I'm looking for. I'm still not clear on how to get it to *do* what is pictured in the screen shots. Do I just experiment with different placements until I get lucky and it looks like the screen shot?

Yeah you really gotta play with it, but do note how that first strut "anchor" looks. See how the failed one is very visible, whereas the successful one is below the surface of the model a bit more? That's the key, I think.

I mean, you can even get struts' second anchor to attach to parts that the first anchor can't attach to (landing legs, rover wheels...et cetera..), but it's hard to do reliably.

[sephirotic]

You know what I do? I go strutting overload and attach to the decoupler, then to the next part.

Ow, you use the decoupler itself as an intermediate structure for the struts? Never thought of that. I did try some years back attaching struts to the 2.5 decouplers to try and firm then with no avail and after the release of the small cubic struct i started doing that on the picture above and all my problems ended so i stopped trying to mess with struts in the decouplers.

I have observed sometimes, with very large boosters with high thrust, if you are doing several radially, they may (not always) begin to resonate back and forth and eventually tear loose, even if the thrust vectoring is disabled. I don't know what causes this, but adding struts between boosters can dampen lateral oscillations.

Ow, i have experienced those kinds of harmonic oscillations in some rare occasions, but i noticed they occurred to me when i was using the Subassembly mod. For some reason even tough it preserved the struts (even after loading and going back to hangar) they acted as there were no strut at all. Every time i used the subassembly i then removed all the struts and placed the struts on the same place all over again and such oscillations on the boosters stopped.

Tough indeed in this case you exemplified putting struts directly between boosters would reduce their wobbliness, just putting a couple of regular struts connecting directly to the main tank would have the same effect and be even more firmer still, so in this case wouldn´t help much.

My point was about that struts is about optimal placement, quantity doesn´t help. There are several small little rules i follow when struting a very large vessel, in the beginning i relied more on quantity, now I do on quality.

An example of such a large rocket with several boosters.

[sgt_flyer]

between boosters, X shaped strutting usually works pretty well (like on the real soyuz, there's some X shaped struts at the base of the rocket between the core stage and the boosters)

[DerekL1963]

(deleted message)

[allmhuran]

The problem is that fuel tanks aren't actually perfectly round, they are composed of polygons like everything else in a computer game. This means that there are little polygon edges on the surface. You can see these easily if you zoom in and go to surface attach mode (ie, not angle snap) and slide a part over the surface of a fuel tank.... as it crosses the edges of the polygons the attachment angle will suddenly shift.

For similar reasons, several fuel tanks have little sticky-outy-ends that make it difficult for a strut to pass over them.

This also means sometimes the strut's connection point is actually "buried" inside the surface a little bit.

In any of these cases, when you try to attach the other end to something else the line for the connector actually collides with the "origin" fuel tank before going anywhere else, and you end up with a strut that starts and ends in virtually the same spot.

One solution to this is to provide some stand-off for your struts... attach a cubic octagonal strut to the fuel tank, and then attach the strut connector to that. Another way is to just make sure you have a lot of angle between the direction the strut is going, and the direction the fuel tank is curving.

[tfabris]

Yeah you really gotta play with it, but do note how that first strut "anchor" looks. See how the failed one is very visible, whereas the successful one is below the surface of the model a bit more? That's the key, I think.

Thanks very much for the tutorial on strut placement, and thanks everyone for an interesting discussion. With this information, I was able to successfully place the struts to prevent my rocket from bending in the middle. It takes trial-and-error to figure out how to place both ends of the strut.

[tfabris]

By the way, how do you change the thread title from "Unanswered" to "Answered"? When I posted the original question, it made me choose between the two. Now I don't see a way to change it to the other thing.

[Catapadurariu]

I am more of a simple man ...

If it looks stupid but works, it ain't stupid !

However, this is mostly used on the stages that get ejected anyway, otherwise it would up the part count.Make sure to start the strut from the cubic strut so that when it gets ejected away, the 1 count goes away with it, too