Something doesn't seem right here....

[Temeter]

Nice findings! I also found that struts aren't that insignificant in terms weight anymore. Didn't knoiw their weight stays after cutting them, though. Might be interesting to look if they are still adding weight after reloading (which causes stuff like decoupler hinges to disappear).

Btw, maybe the initial slowness of the non-strutted rockets is caused by some velocity-change when the strut is cut during liftoff?

[sal_vager]

This is probably why most airplanes don't use interplane struts or wires anymore.

[basic.syntax]

That was a fun experiment, I'll be paying close attention to asymmetrically placed struts in future craft.

The only difference between the two sides below, is double strutting on one of them.

[Red Iron Crown]

Interesting, I wouldn't have though a strut end would have such a significant effect. Good on you for tracking down the cause of the difference, Levelord.

[Levelord]

To illustrate how significant and detrimental struts are to efficiency, I have made a comparison.

Here we use 11 struts for a total weight of 0.55 tonnes, compared to the second craft which is carrying a FL-T100 tank with 0.5625 tonnes of fuel. That means the strutted craft is still 0.0125 tonnes lighter. Let's see how this goes...

Launch conditions:

It becomes immediately clear that weight for weight, the struts slow you down more than actually carrying rocket fuel. This is a significant finding and now makes it more crucial that crafts now use as few struts as possible.

The un-strutted rocket on the left here, leaving the strutted one in the dust.

[Temeter]

Looks like drag to me.

[urablahblah]

So does this phenomenon extend to other formerly "physicsless" parts?

[Arcwolf]

I'm assuming this phenomenon would also extend to the use of Fuel Ducts. Curious to note that the strut and the fuelLine are the only two parts to use dragModelType = SPHERICAL

Would a MM cfg to changing that to default "fix" this drag oddity?

[Red Iron Crown]

I'm assuming this phenomenon would also extend to the use of Fuel Ducts. Curious to note that the strut and the fuelLine are the only two parts to use dragModelType = SPHERICAL

Would a MM cfg to changing that to default "fix" this drag oddity?

I doubt it. Spherical drag model just means that drag is the same no matter the part's orientation, this is an appropriate simplification for struts/fuel lines as their shape is not fixed. Dialing down minimumDrag and maximumDrag values would likely be better.

[Sobriquet]

Wow, nice find. A shame to, because I love me some struts.

Do the struts and fuel lines drag the same regardless of their length?

[Col_Jessep]

Oi! I was strutting all kinds of stuff on my SSTOs to make them more rigid. Thanks for the science!

[KerikBalm]

:/ Hmmm, I guess my SSTOs must have a lot of room for improvement then:

I'm sure some of those can go. I threw a lot on there only considering the mass as not that much... but if they cause drag too... yeeesh.

But at large sizes... I really really need them for my center-loader/twin boom designs...

I don't want to be limited to mk3 bays (well, those designs already work, so I'm not)

So then is the drag independent of the length of the connection? Even after the connection is severed, there is still drag it seems... :/

Boo... I'm going to have to look into removing struts, and making these designs flimsier

[Deddly]

As I suspected. Well done for confirming it with some solid testing!

[Tw1]

Well, that's a pitty. Who doesn't like a few struts in place?

[Guest]

Well, that's a pitty. Who doesn't like a few struts in place?

This revelation isn't going to change the way I build things in the slightest.

[Red Iron Crown]

This revelation isn't going to change the way I build things in the slightest.

Same. Since 0.23.5's Great Joint Strengthening I don't think I've used more than a half dozen struts on anything. The drag is significant on a contrived test craft but doesn't make much difference for more practical things.

[WinkAllKerb'']

drag vector applied to strut main root not splitted on separation. And there no way to workaround that engine related i mean (and aside *.cfg drag value adjustement ?)

(this also mean when you use strut and decoupler you have all interest to have the strut roots on dropped part)

[Chris_2]

Oi! I was strutting all kinds of stuff on my SSTOs to make them more rigid. Thanks for the science!

Still think you should do that. The drag from the control surfaces of a wobbeling craft is probably a -lot- higher than the drag from these struts...

So, what about struts inside cargobays? They don't experience drag, do they?

[WinkAllKerb'']

hum struting from inside collider (and/or bays) you mean chris using vab sph camera trick ?

this kind of things impact vectorings calculation // with a craft "coating" ?

Edited May 23, 2015 by WinkAllKerb''

[Chris_2]

hum struting from inside collider (and/or bays) you mean chris using vab sph camera trick ?

this kind of thing ?

Either that, or really struts inside cargobays (to strut down cargo ). I think that the former do generate drag, and the latter don't...

[Brotoro]

I noticed the powerful drag of struts when I was building 1.0 spaceplanes. I had a spaceplane that could reach orbit, but it was very flexible (it would sag while sitting on the runway). So I added some struts to make the airframe more rigid...and found out that the plane could no longer get to space.

I think this is very unfortunate. It's true that I can get away with a lot fewer struts nowadays than I could in earlier versions of KSP, but I still need struts to brace very wobbly payload stacks. Because these struts are being used to overcome a limitation of the program (in real life we could build more rigid airframes without external struts), it seems unfair of Squad to hit us so hard with the dragginess of the struts.

I would recommend that Squad reduce the drag of struts.

[WinkAllKerb'']

They don't experience drag, do they?

air intake don't grab air inside cargo bay closed i suppose it's mainly the same for strut. On the other hand it might be also linked with the 3dtools face properties of a tris (strut attach on a specific face and they are unable to cross a collider surface (even from inside, so if a part to part linkage can be done you cannot do multi part crossover linkage). So attached face within the two of a tris surface probably matter.

Edited May 23, 2015 by WinkAllKerb''

[cybersol]

In my career save, I had only unlocked clamp-o-tron juniors. I built a rocket with them stacked end-to-end that had crazy tear itself apart wobbles in atmosphere with SAS. The wobbly version had a strut cage around 3 sets of 4 modular girders transferring the load up the rocket. The middle set of 4 girders was offset 45* from the ends, and I was using 2 sets of 8 struts in an X pattern to transfer the load between the ends and the middle. It was all a wobbly mess when using SAS, even just in stabilization mode. I was confused because these types of strut cages worked well in earlier versions. In total frustration and because I was already hitting part count limits, I ripped out all the struts. I then put the 3 sets of 4 modular girders all aligned at the same angle on the rocket, and simply connected between the ends and the middle with 2 sets of 4 straight struts. The 8 struts now being aligned with the airflow totally got rid of the wobbles!

TL;DR I think the drag on the X version of struts was creating significant cross forces that greatly enhanced the SAS wobble.

[SAI Peregrinus]

Some quick testing indicates that nuFAR fixes this issue.

[GusTurbo]

I don't think struts should behave this way. Their effect seems way outsized.