[1.4] SpaceY Heavy-Lifter Parts Pack v1.17.1 (2018-04-02)

[Krombopulos Michael]

Is there an easy way to attach an srb to 2 decouplers? With 1 its too wobly, but when I try to attach an srb to 2, usually the decoupler on top isn't attached. Like this: https://imgur.com/a/NERlA

[blowfish]

52 minutes ago, Krombopulos Michael said:

Is there an easy way to attach an srb to 2 decouplers? With 1 its too wobly, but when I try to attach an srb to 2, usually the decoupler on top isn't attached. Like this: https://imgur.com/a/NERlA

Welcome to the forums!  There's no way to attach to 2 decouplers unfortunately.  Your best bet is to use struts.  Or to enable advanced tweakables in the game settings and use auto-struts.

[Krombopulos Michael]

1 hour ago, blowfish said:

Welcome to the forums!  There's no way to attach to 2 decouplers unfortunately.  Your best bet is to use struts.  Or to enable advanced tweakables in the game settings and use auto-struts.

Well dang. I just started playing again after a long hiatus and I could have sworn I used to use 2. Thanks anyway

[SmashBrown]

3 hours ago, Krombopulos Michael said:

Well dang. I just started playing again after a long hiatus and I could have sworn I used to use 2. Thanks anyway

Id recommend kerbal joint reinforcement and using one of the nose cones with small engines in them for decoupling.

[NecroBones]

7 hours ago, Krombopulos Michael said:

Well dang. I just started playing again after a long hiatus and I could have sworn I used to use 2. Thanks anyway

 

You can kind-of fake it by putting the SRB on one decoupler, and then strutting it to a second one. Or, just strut it directly to the inner stage, and the struts will detach when the decoupler is separated.

[WuphonsReach]

On 1/11/2017 at 11:45 PM, Krombopulos Michael said:

Is there an easy way to attach an srb to 2 decouplers? With 1 its too wobly, but when I try to attach an srb to 2, usually the decoupler on top isn't attached. Like this: https://imgur.com/a/NERlA

With the new auto-strut, you don't need KJR any longer.  Selective use of "auto-strut to heaviest part" (usually on the heaviest fuel tank in each stage) do well enough.

Radial-attached SRBs should be mounted to the radial decoupler, then one or two regular struts used at the upper end of the stage, attaching it to the fuel tank of the inner core.  I prefer two struts per SRB as it keeps the wobble to zero and it looks nicer then a single strut.

If you're having trouble with the top of the SRB striking the inner core when detaching, mount it lower on the decoupler (or use the decouplers with the built-in separatron rockets, plus maybe the nose cones with built-in separatrons).  I usually turn down the nose-cone separatrons to 20-30% power and only 3.2 of fuel.  Just enough to push the nose away, but without making it tumble.

Fins at the bottom of the SRB also help with stability on detach.

[Vahal]

How I usually attach SRBs: 
 

Spoiler

U place the decoupler, then place the booster.

Use the translation thingy to move up the decoupler on the core AND on the booster so when you'll decouple, the force will be applied on the top of the booster and it will be ejected away from the core. Plus when using the decoupler combo from Necrobones, this is even more efficient since there is an engine to push the booster far away.

And to secure the booster, you place a strut à the bottom of the core to fix the bottom of the booster.

Edited January 14, 2017 by Vahal
failed at link images... -_-

[TiktaalikDreaming]

All of a sudden I'm wondering about extra long radial decouplers.

[blowfish]

2 hours ago, TiktaalikDreaming said:

All of a sudden I'm wondering about extra long radial decouplers.

It wouldn't actually help with wobblyness at all.  All joints between parts in KSP are actually a single point.

[TiktaalikDreaming]

4 minutes ago, blowfish said:

It wouldn't actually help with wobblyness at all.  All joints between parts in KSP are actually a single point.

Yes, BUT, it'd give an excuse to increase the attach node size, which the stiffness is based on.

[NecroBones]

1 hour ago, TiktaalikDreaming said:

Yes, BUT, it'd give an excuse to increase the attach node size, which the stiffness is based on.

 

As far as I know, node size has little to no effect on stiffness (I could be wrong, but that seemed to be the case). Mass of the parts has a huge impact on stiffness though.

[TiktaalikDreaming]

5 hours ago, NecroBones said:

 

As far as I know, node size has little to no effect on stiffness (I could be wrong, but that seemed to be the case). Mass of the parts has a huge impact on stiffness though.

The size setting did have an effect way back when.  Not sure nowadays.

 

Tested.  Once for regular and once for an identical stock part.  No noticeable change in stiffness, but the radial decoupler with node size increased didn't break.  It MIGHT suggest a higher breaking point. 

Still, a long decoupler, with a matching collider and an attach node in the middle somewhere, *should* do something to stop too much movement in or out. 

AND, I first tried the test with KJR still on.  And, well, that does seem to hold everything very nicely in place.

So, all stock parts, using the "Hydraulic Detachment Manifold" mounted near the top so the boosters would force inwards as much as they could.

Followed by

Taking the Hydraulic Detachment Manifold, quickly editing a copy of it's config to just have ,8 at the end of the attach_node definition (8 is huge,WAAAY beyond what that should be) and retrying, the boosters still wobbled about.  They didn't seem to quite hit that wobbles causing wobbles feedback loop though, and got all the way to booster exhaustion.

 

Only one test of each so far, and attach points were done by eye, and aren't the same from craft to craft. 

 

 

And, further testing suggests all sorts of things.  No idea anymore.  Sometimes there's no difference, other times it seems a larger node size makes things worse.  I suspect there's too many small inputs causing large outputs to say anything, and I'll defer to the wisdom of @NecroBones re there being no detectable effect.

Edited January 16, 2017 by TiktaalikDreaming

[WuphonsReach]

I've moved away from KJR in KSP 1.2.2 for stability.

The auto-strut feature is my preferred method.  I'll usually auto-strut the heaviest tank on a stage to the "heaviest part" (which is usually the heaviest tank on the prior stage).  I try to avoid auto-strut on the final payload, but might use the "heaviest part" method on the fairing base.  If I have to use auto-strut on the payload, I'll pick the heaviest part and auto-strut that back to the "heaviest part" (usually the fuel tank of the previous stage).

I'll still use traditional struts (mostly for radially attached boosters, where a pair of struts at the booster's nose connecting it to the core is enough).  I haven't had experience with trying to auto-strut radial boosters to the core, I think they'd still wiggle too much around the CoG of the booster.

[NecroBones]

9 hours ago, TiktaalikDreaming said:

The size setting did have an effect way back when.  Not sure nowadays.

Tested.  Once for regular and once for an identical stock part.  No noticeable change in stiffness, but the radial decoupler with node size increased didn't break.  It MIGHT suggest a higher breaking point. 

<snip>

And, further testing suggests all sorts of things.  No idea anymore.  Sometimes there's no difference, other times it seems a larger node size makes things worse.  I suspect there's too many small inputs causing large outputs to say anything, and I'll defer to the wisdom of @NecroBones re there being no detectable effect.

 

Heh, thanks for playing with it a bit. I didn't test it that extensively, but the only thing that was obviously making big differences for me was the starting mass of the parts. Tanks get really stiff joints, for instance, but small inline decouplers, docking ports, etc can be quite floppy, despite having the same node size. I'm wondering if there's some goofy equation that takes all of these things into account, and some have a greater impact than others at different times, depending on size of the part and other factors? I wish I had a view into what they're really doing. And I've been begging for a "stiffness multiplier" setting forever. We can change the breaking strength directly in the CFGs, but not the stiffness, which is really a shame.

 

[TiktaalikDreaming]

6 hours ago, NecroBones said:

 

Heh, thanks for playing with it a bit. I didn't test it that extensively, but the only thing that was obviously making big differences for me was the starting mass of the parts. Tanks get really stiff joints, for instance, but small inline decouplers, docking ports, etc can be quite floppy, despite having the same node size. I'm wondering if there's some goofy equation that takes all of these things into account, and some have a greater impact than others at different times, depending on size of the part and other factors? I wish I had a view into what they're really doing. And I've been begging for a "stiffness multiplier" setting forever. We can change the breaking strength directly in the CFGs, but not the stiffness, which is really a shame.

 

Yeah.  I have previously (0.8, 1.0.5 etc) seen differences in stiffness with node sizes, but only on fairly large parts now that I think back.  Maybe there's a bunch of algorithms and it takes the worst.  :-)

[TiktaalikDreaming]

A bit more experimentation and it's looking like a lot to do with collider sizes.

First up, there's at least two algorithms (of sorts) and the worst one will be the one causing the problems, because there's always two joins.  None of the Squad surface nodes have sizes.  Which should have been a hint.  :-)

Some of my really bad results, the joint I was theoretically messing with wasn't having any trouble.  The decoupler's node, is the decoupler attaching to the main body of the craft.  The join that usually (not always) has trouble is the surface attaching of the booster to the decoupler.

So, I messed with the TT-70 decoupler, the one that sticks out on framework, and made it longer.  The interesting thing with this decoupler is the collider exists only on the outside portion of the part (ie, where you attach the booster), but not where it itself glues onto the main body.

So, my stretched decoupler showing collider (orange highlighted thing) in blender;

 

With everything else the same, no node size value, same mass, everything, the longer collider makes a huge difference.  There's still a visible torquing going on, but much less.

I'm not sure how the length is making the attachment to the core vehicle stiffer.  It doesn't twist enough to get that collider to contact the main body.  But it seems likely the extra stiffness of the booster to decoupler at least is based on the size of the collider, or possibly the connected surface area, or something.  There's most definitely an effect.
Long decoupler result;

 

Stock decoupler, just before a booster left the building;

And it's likely that increasing the mass to match the larger part, and extending the collider (or adding a second) so it contacts the body of the craft, can all help add some stiffness.

[NecroBones]

1 hour ago, TiktaalikDreaming said:

With everything else the same, no node size value, same mass, everything, the longer collider makes a huge difference.  There's still a visible torquing going on, but much less.

 

Huh! I never even thought to play with collider sizes for that. I could see that being taken into account to adjust stiffness too. Bigger->Stiff. Smaller->Floppy. (insert "that's what she said" comment here?)

 

Unfortunately it's hard to use this to make some parts exceptionally strong, since an oversized collider is usually out of the question.

[TiktaalikDreaming]

7 hours ago, NecroBones said:

Unfortunately it's hard to use this to make some parts exceptionally strong, since an oversized collider is usually out of the question.

Yeah.  It's nice to know the collider is being used, but it doesn't allow much custom tweaking.  I should try a non-collider invisible box as well.  I know KJR uses whatever's there, as adding an entity that is neither visual or collider is the fix for some weird stuff with large tension forces (aka parachutes).

It's also possible the collider is only consulted for joint stiffness for surface attach joins.

Gawd,... I'm digging myself an experiment rabbit hole.

[WuphonsReach]

Wish list items:

A longer version of the S223 with two or three more segments.

A taller version of the Fenrir S321 with two or three more segments.

[blakemw]

I've noticed a problem with the A3-12 Adapter Structure, when a (2.5m) part is attached to the middle node, the undersurface of the attached part is not treated as being "shielded" from drag so you get a draggy surface exposed to the wind. This is especially obvious when the Adapter is used upside down because then the exposed surface will be facing the wind and the drag generated will be very extreme.

This problem doesn't seem to occur when a 1.25m part is attached to the top node, also the adapter part itself and the part attached to the bottom node work properly in terms of drag, it's only a problem with the middle node not acting as a drag-shielder.

[NecroBones]

3 hours ago, blakemw said:

I've noticed a problem with the A3-12 Adapter Structure, when a (2.5m) part is attached to the middle node, the undersurface of the attached part is not treated as being "shielded" from drag so you get a draggy surface exposed to the wind. This is especially obvious when the Adapter is used upside down because then the exposed surface will be facing the wind and the drag generated will be very extreme.

This problem doesn't seem to occur when a 1.25m part is attached to the top node, also the adapter part itself and the part attached to the bottom node work properly in terms of drag, it's only a problem with the middle node not acting as a drag-shielder.

 

Yeah, I can't say I'm surprised. Unfortunately, this is just how KSP works now. I created those parts back when we had the old aerodynamic model, and it was OK to create these sorts of adaptive parts. But the new aerodynamic system isn't cooperative with it, so I've stopped designing parts that way. But we still have these. I'm afraid that the fix for this is to use FAR.

 

On 1/28/2017 at 7:41 PM, WuphonsReach said:

Wish list items:

A longer version of the S223 with two or three more segments.

A taller version of the Fenrir S321 with two or three more segments.

 

Certainly doable. Bigger is better, right?

 

[NecroBones]

 

Updated:

 

1.16 (2017-01-30) - SRB Update.
 - Corrected smoke emitter location on the 3.75m SRB, 21m length.
 - Added a 30m long 3.75m SRB.
 - Added a 29m long 2.5m SRB.

 

[Codraroll]

Did anything ever happen to the idea of more radial engines? I once used a 3.75 m storage bay to haul a rover to the Mun, and remember being frustrated that the relatively weak Thud engines were my only means of creating a lander with a door on ground level. Bigger and stronger radial engines would have been a godsend for such missions.

If radial engines are being considered, I propose the names "Chicken" and "Rooster" for a small-sh (2x Thud size) and a large one (4x Thud or so), keeping with the naming scheme of flightless birds.

[DStaal]

30 minutes ago, Codraroll said:

Did anything ever happen to the idea of more radial engines? I once used a 3.75 m storage bay to haul a rover to the Mun, and remember being frustrated that the relatively weak Thud engines were my only means of creating a lander with a door on ground level. Bigger and stronger radial engines would have been a godsend for such missions.

If radial engines are being considered, I propose the names "Chicken" and "Rooster" for a small-sh (2x Thud size) and a large one (4x Thud or so), keeping with the naming scheme of flightless birds.

Agreed. 

Though really, what we need are *ramps*, so we can get down out of the storage bays.

[Loren Pechtel]

1 hour ago, Codraroll said:

Did anything ever happen to the idea of more radial engines? I once used a 3.75 m storage bay to haul a rover to the Mun, and remember being frustrated that the relatively weak Thud engines were my only means of creating a lander with a door on ground level. Bigger and stronger radial engines would have been a godsend for such missions.

If radial engines are being considered, I propose the names "Chicken" and "Rooster" for a small-sh (2x Thud size) and a large one (4x Thud or so), keeping with the naming scheme of flightless birds.

If you can live with something very draggy you can build a rocket by hooking the rover to the bottom of a fairly big tank and then radially attaching it's boosters to that.  I've been hauling some big rovers that way.  I turn way late on the launch, probably costs an additional 200-300m/s getting up off Kerbin but it works.

As for the Thud--I've found one occasion to use it.  I mounted some on a science rover that I sent around Kerbin--their purpose was to make the rover able to hop to get the flying low over science.

Edited January 30, 2017 by Loren Pechtel