Quick question; can you strut rockets with the girder pieces?

[thiosk]

Consider the following case:

Does this act as a substitute for struts in this case, or are they not acting as structural units?

My tests have suggested both, so I ask ye of the kraken hivemind fer yer insights.

[Murph]

In short, no.

That's not to say they can't be used to provide some extra structural integrity. Where they can be useful, is to, for example, add 4 radially to a decoupler, effectively making it wider. You can then use struts from above and below onto that, possibly allowing struts where they would not normally be possible, and creating some nice triangular bracing. A rule of thumb, however, is that you need to strut twice, once going up, once going down, as the girders will pivot on their attachment. when pushed up/down by the struts.

So, they can provide a useful strut attachment point, which then needs to be braced itself, but don't provide any structural integrity on their own.

Edit: In the case of the example shown, 2 stacked Jumbo 64s, I would rarely strut that joint directly, but it's personal preference how you achieve structural integrity. I typically strut from the boosters attached to the lower Jumbo, up to the in-line stack decoupler above it. As long as there's at least 3 or 4 boosters spread evenly around the Jumbo, that tends to be sufficient. The boosters need struts anyway, or they will pivot on their attachment or decoupler, so I more or less kill 2 birds with 1 stone. It does leave a little flex on the Jumbos after the boosters decouple, but that should be reasonably high in the atmosphere, with less aerodynamic load, less gravity load, etc. I then only strut the Jumbos directly if it actually proves to be necessary after testing.

Edited December 5, 2013 by Murph

[thiosk]

Thank you for the clarification.

My ships are getting bonkers in scale, so coming up with viable part-reducing approaches that ensure stability are important.

[Murph]

Thank you for the clarification.

My ships are getting bonkers in scale, so coming up with viable part-reducing approaches that ensure stability are important.

You're welcome. Don't go over the top with struts. A small, occasional amount of flexing or wobbling can be ok. It's when it either forces you off course, or breaks apart that it's an issue.

[Kasuha]

Girder segments (or more often "cubic octagonal strut" parts, i.e. the small girders) can be used to provide elevated anchor points for struts. Such as around the center piece in this picture:

[larsmaehlum]

Girder segments (or more often "cubic octagonal strut" parts, i.e. the small girders) can be used to provide elevated anchor points for struts.

This is almost mandatory on longer rockets. A big payload on top of multiple orange tanks get very wobbly without them.

[DrMonte]

I usually try to avoid using large Modular Segments. You might think that they does not weight much but those six already adds ~3% of an orange tank mass. Usually you can achieve the same results with the small ones which are almost weightless.

[Monkeh]

If you get it just right then you can just stitch two orange tanks together with an 'x' shaped pair of struts. Otherwise you can use the tiny girders with struts for very low mass, see below.

Works for any kind of inline stabilization.

[J.Random]

the small ones which are almost weightless.

They ARE weightless in flight. In fact, any part with the word "strut" in it will be weightless. I'm not sure why Squad did this: even if it's some sort of a workaround for some problem, it could be done with "subcategory" config value instead of parsing the part's name.

[DrMonte]

They ARE weightless in flight. In fact, any part with the word "strut" in it will be weightless. I'm not sure why Squad did this: even if it's some sort of a workaround for some problem, it could be done with "subcategory" config value instead of parsing the part's name.

That's kind of weird information. From where do you get it?

[Monkeh]

They ARE weightless in flight. In fact, any part with the word "strut" in it will be weightless. I'm not sure why Squad did this: even if it's some sort of a workaround for some problem, it could be done with "subcategory" config value instead of parsing the part's name.

Does their weight change when you launch then?

The wiki has the weights of struts as:

Strut connector 0.05

Cubic octagonal strut 0.001

Octagonal strut 0.001

Which granted is very low but hardly weightless.

http://wiki.kerbalspaceprogram.com/wiki/Parts#Adapters.2C_Couplers_.26_Struts

[maltesh]

In many cases, if you rotate the strut piece to be flat-side-out before placement, you can stitch two parts of the same diameter together without outrigger pieces.

Can make moving the struts after you placed them a bit annoyiing, though, because the part that's best to click on to grab the strut is now inside the tank.

Edit: Yes, the cubic octagonal strut is a physicsless part. The simulation doesn't consider it to have mass, regardless of the claim in the VAB, which results in all sorts of crazy things that can be done with it, including using stack separators to make one hit the Mun in 13 seconds. Back in 0.18, one of the reasons that some stations would jitter apart on docking was because a player had used a cubic octagonal strut to mount a docking port radiall, though that may have been fixed by now. In /most/ situations however, a player doesn't wind up using sufficient COSes in critical places to realize that this is so.

There are a couple other physicsless parts. I think the RCS ports were similar,but I do not recall what all the other ones were, because the Cubic Octagonal Strut, by way of being a structural part, was the easiest one to accidentally build yourself into trouble with.

Edited December 5, 2013 by maltesh

[DrMonte]

Does their weight change when you launch then?

The wiki has the weights of struts as:

Strut connector 0.05

Cubic octagonal strut 0.001

Octagonal strut 0.001

Which granted is very low but hardly weightless.

http://wiki.kerbalspaceprogram.com/wiki/Parts#Adapters.2C_Couplers_.26_Struts

We also have "Landing Gear" called "LT-X Landing Strut" which has quite high mass.

[J.Random]

We also have "Landing Gear" called "LT-X Landing Strut" which has quite high mass.

Part name, not title.

Proof: http://imgur.com/bzw5B2m in the VAB, http://imgur.com/w2lHRlW at the launchpad. 16 "struts" => 16 kg lighter than it should be. It also may lead to physics Kraken if one of those "struts" is a root part for your rover, for example.

[DrMonte]

Part name, not title.

Proof: http://imgur.com/bzw5B2m in the VAB, http://imgur.com/w2lHRlW at the launchpad. 16 "struts" => 16 kg lighter than it should be. It also may lead to physics Kraken if one of those "struts" is a root part for your rover, for example.

It seems we just need to learn to read From wiki:

"Currently (0.20.2), the Octagonal Strut has "PhysicsSignificance = 1", making it massless and dragless."

"Currently (0.20.2), the Cubic Octagonal Strut has "PhysicsSignificance = 1", making it massless and dragless."

Though EAS-4 connector - "Currently (0.21.1) it has a mass of 0.05 and is dragless."

[Rhomphaia]

Part name, not title.

But then there is the SmallGearBay landing gear, which doesn't feature the word strut anywhere, nor does it have the PhysicsSignificance line, yet is massless and dragless. though it will influence the CoM display in editors it has no effect in flight.

[J.Random]

EAS-4 are massless even in VAB/SPH. Not sure about gear bay, I think it has mass and drag when it's lowered and doesn't have any when it's stowed. I suspect some 1337 spaghetti programming there, but again, I may be completely wrong.

[Rhomphaia]

EAS-4 are massless even in VAB/SPH. Not sure about gear bay, I think it has mass and drag when it's lowered and doesn't have any when it's stowed. I suspect some 1337 spaghetti programming there, but again, I may be completely wrong.

No difference whether up or down, at least in my tests strapping them to SRBs

[pizzaoverhead]

Edit: Yes, the cubic octagonal strut is a physicsless part.

HarvesteR provides a good explanation of this here.

It sounds like it has been fixed in 0.23:

That's the one indeed. In can see the little cube strut dangling there off the broken port.

Turns out, while there wouldn't be any errors explicitly with this bug, the physicsless parts would get handled as if they were normal ones, and they'd get joints added to them. Joints that require rigidbody components, which got added with no initial setup to these objects.

To get an idea of what this means, consider that by default, rigidbodies have 1 full unit of mass, respond to physx gravity (which we don't use) and that these parts were parented to another object. You can get a sense of the chaos that would come from it.

It's one bug I'm very glad to see fixed. Definitely.

Cheers

[thiosk]

Interesting that the struts are in fact weightless.

I've used the internally clipped struts for linking tanks, the octagonal struts for creating in line stabilizations, and have followed the advice "MOAR STRUTS" to the letter.

I probably have 250 or so struts on this rather ambitious launch (i'm launching a too-big-refueling station in one go, to reduce part count and wobble)

I get failures of essentially every component going around, from the strutted tanks to the engines. I suppose 5 orange tanks + station is just too much. The asparagus stages start collapsing if I fire at full thrust, and the payload (specifically an in line sr. sized clampotron) fails despite 15+ strut connections to various girder segments once I get ready for first stage drop.

Edited December 5, 2013 by thiosk

[numerobis]

In current stock KSP, parts called "strut" are indeed massless, but making things massless isn't done by parsing the name. It's done with PhysicalSignificance = 1 in the part.cfg for some parts, and in the relevant modules for other parts. For example, the EAS-4 (which is a StrutConnector part) and the small gear bay (which has a ModuleLandingGear module) do it in their respective OnStart functions, which are invoked when the spacecraft launches. The small gear bay throws off the CoM marker, a perennial problem for plane builders: in the SPH, the gear is accounted at 0.5t, which is a lot -- but in flight they're massless.

I've definitely used strut cubes to make attachment points to which I sent some struts, for tall floppy rockets. A very useful technique, if one that uses a lot of parts. I'm currently consumed by the attempt to make a very low-part setup with a huge number of tanks.

Edited December 5, 2013 by numerobis

[Smidge204]

For larger (thicker) rockets, sometimes you need a bit more clearance. My favorite technique (the horizontal struts are not entirely necessary, but they can help if they fit):

Edit: For part count/mass/size reasons, it's better to put the trusses on parts of the rocket that will fall away when staged so you ditch them, too. Obviously you don't want to do that if they are still providing structural support, but in a case like above - imagine there's a decoupler between the tanks, and you can see how the trusses would be dropped when their job is done.

Interesting that the struts are in fact weightless.

Weightless, but not free. They add to part count and physical simulation complexity. Try to use as many struts as you need, and not more.

=Smidge=

Edited December 5, 2013 by Smidge204

[DMagic]

The small cubic struts are physics-less too, which can occasionally cause issues. Using them as strut anchors shouldn't be a problem, but connecting other things to them might not be a good idea.

There is also a mod that comes with a really great part made specifically to be used as a strut anchor.

http://forum.kerbalspaceprogram.com/threads/34664-Large-Structural-Station-Components-20-Compatible-Release-Thread