Massless parts

[LongbowEOD]

So I'd long heard that certain parts, such as struts and landing gear, are functionally massless. That is, they have a mass listed in the VAB or SPH, but that mass is not calculated when physics take over. Is there a comprehensive list of parts that have this feature? I ask because I just noticed that the small science experiment packages (seismic, temperature, gravioli, and air pressure) seem to share it, which was news to me. I'd been using experiments to try and balance out some of my designs (using m4v's excellent RCS Build Aid), but it was anomalously not working.

I did a bit of poking around to find some parts that might be massless, but this list is not comprehensive (or necessarily even correct). So far I've got:

FTX-2 External Fuel Duct

Both RCS Ports

EAS-4 Strut Connector

Cubic Octagonal Strut

Octagonal Strut

OX-STAT Photovoltaic Panel

Illuminator Mk 1 and Mk 2

Communotron 16 and 88-88 (but the Comms DTS-M1 does have mass)

Small Gear Bay

Double-C Seismic Accelerometer

Presmat Barometer

GRAVMAX Negative Gravioli Detector

2HOT Thermometer

Is there something in the part configs that makes them massless?

Edited April 11, 2014 by LongbowEOD
Forgot to put the Small Gear Bay in my list

[Javster]

The small cubic struts are massless???????????? Hooray!!

[stupid_chris]

PhysicsSignificance = 1

This line makes the parts massless. The Z-100 battery also shares this, as well as the ladders. I've changed the behaviour of some parts in the Stock Rebalance project (link in my sig). Some parts like the battery /should/ have a mass, and those who don't have their mass changed to zero so they don't affect the CoM.

[AlexinTokyo]

Additionally (at least):

The 3.75 m decoupler from 0.23.5 (definitely a bug)

Both radial battery backs

The 'flat' tiny-to-small adapter

The small hardpoint

All the ladders

The .cfg files for all these parts contain

PhysicsSignificance = 1

which tells the physics engine to ignore them.

[Red Iron Crown]

I don't understand why there are massless parts at all. Is it a concession to some shortcoming of the engine?

[Ziff]

I don't understand why there are massless parts at all. Is it a concession to some shortcoming of the engine?

Certain parts like ladders and the handrails make sense to be massless. It would suck to have to balance a craft with useless rails on the opposite side just to balance the center of mass. It would take away from the design look as well. As for the batteries, that's probably just something that was overlooked.

[AlexinTokyo]

I think the idea is to allow you to put one only of a given part off center without unbalancing the rocket. So you don't have to add two ladders in symmetry, for example.

I also think that as of 23.5 there are too many by far.

[Red Iron Crown]

Hmm, I could see that I guess. Personally, I think balance could be achieved by offsetting ladders with battery packs, etc; just another part of the challenge of design.

I'm fairly sure that the delta-V calculator mods, like Kerbal Engineering Redux and MechJeb, count the mass of the massless parts when calculating. Weird.

[Claw]

MJ counts some parts and not others. A lot of the parts listed above changed in the last version (0.23.5). I'm also guessing it's to help with balance.

Although my thought is that the mass of things like science parts is so light, it doesn't really effect big designs anyway. And for small probes, part of the challenge is to manage placement of those small pieces. So having no weight makes designing small probes almost a non-consideration.

[LongbowEOD]

I'm fairly sure that the delta-V calculator mods, like Kerbal Engineering Redux and MechJeb, count the mass of the massless parts when calculating. Weird.

That's actually what led me to this problem in the first place: MechJeb, at least, does NOT add the mass of these massless parts.

Oddly, I was not able to find the "PhysicsSignificance = 1" line in the Small Gear Bay .cfg file.

[Kasuha]

I think there's point behind making these parts physically insignificant, you don't always need two batteries or lights and it's good if you can attach just one without making your ship unbalanced.

On the other hand I also think that it's unfair that it makes them massless. Maybe something could be made to distribute their weight over remaining "massy" parts of the ship proportionally to their (dry) weight so the ship's mass is the same as if these parts are massy but it stays balanced as long as the "massy" part of the ship is balanced. This redistribution could be recalculated every time the ship splits to two.

[Red Iron Crown]

Massless batteries strikes me as a bit exploitable. Imagine an ion powered ship with only one solar panel or RTG, with hundreds of massless batteries spammed over it to give enough capacity for a burn.

Seems to me the RTG should be massless if balancing ships is the goal, it's given me more trouble with balancing small craft than any other three parts combined.

[PlonioFludrasco]

Oh, seriously RCS ports and little batteries are massless? nice!!

Anyway, I'll look at the configs and make some tests.

[stupid_chris]

Hmm, I could see that I guess. Personally, I think balance could be achieved by offsetting ladders with battery packs, etc; just another part of the challenge of design.

I'm fairly sure that the delta-V calculator mods, like Kerbal Engineering Redux and MechJeb, count the mass of the massless parts when calculating. Weird.

No they don't, MechJeb and engineer specificall check for this.

That's actually what led me to this problem in the first place: MechJeb, at least, does NOT add the mass of these massless parts.

Oddly, I was not able to find the "PhysicsSignificance = 1" line in the Small Gear Bay .cfg file.

Because it's in the module

Massless batteries strikes me as a bit exploitable. Imagine an ion powered ship with only one solar panel or RTG, with hundreds of massless batteries spammed over it to give enough capacity for a burn.

Seems to me the RTG should be massless if balancing ships is the goal, it's given me more trouble with balancing small craft than any other three parts combined.

Indeed, reason why I fixed it in the Stock Rebalance project. Probably an error, if not, really dumb.

[Claw]

I think there's point behind making these parts physically insignificant, you don't always need two batteries or lights and it's good if you can attach just one without making your ship unbalanced.

On the other hand I also think that it's unfair that it makes them massless. Maybe something could be made to distribute their weight over remaining "massy" parts of the ship proportionally to their (dry) weight so the ship's mass is the same as if these parts are massy but it stays balanced as long as the "massy" part of the ship is balanced. This redistribution could be recalculated every time the ship splits to two.

I thought it would be nice to make something like a procedural probe core. Something that you could add more battery power, torque, etc... (up to some max) and it would scale mass automatically. Same sort of idea with a science module. Something that allows the tweakable inclusion of the small science parts.

Sort of gets around the balance problem while still forcing design choices and mass considerations.

[Dunbaratu]

I don't understand the concept of zero mass parts and how on earth that doesn't completely break the physics engine, because of:

F=ma

Wouldn't a massless part be accelerated up to an infinite speed the moment it collides with anything?

[stupid_chris]

I don't understand the concept of zero mass parts and how on earth that doesn't completely break the physics engine, because of:

F=ma

Wouldn't a massless part be accelerated up to an infinite speed the moment it collides with anything?

They're not actually totally massless. IIRC, the game adds a mass of 0.005t (5kg) at the CoM of the ship for non significant objects. So if your ship is only made out of physicsless parts, it will behave as if it weighed 0.005t.

[kurja]

No they don't, MechJeb and engineer specificall check for this.

Engineer absolutely does show increased ship mass when adding trusses, batteries, lights, landing gear, etc massless parts. VAB and SPH center of mass indicators also move about when attaching massless parts.

I just tried making a 2,99 ton ship with 3kN thrust, so it can just barely take off. Added whacktonian amounts of ladders and such, engineer shows it weighs way more but takes off the pad just as well =D

The in-flight engineer seems to be showing the 'corrected for massless parts mass', but the VAB engineer counts mass of all parts, physics significance on or off.

Edited April 11, 2014 by stupid_chris

[stupid_chris]

Engineer absolutely does show increased ship mass when adding trusses, batteries, lights, landing gear, etc massless parts. VAB and SPH center of mass indicators also move about when attaching massless parts.

I just tried making a 2,99 ton ship with 3kN thrust, so it can just barely take off. Added whacktonian amounts of ladders and such, engineer shows it weighs way more but takes off the pad just as well =D

The in-flight engineer seems to be showing the 'corrected for massless parts mass', but the VAB engineer counts mass of all parts, physics significance on or off.

Then engineer is dumb. I know for a fact that MechJeb doesn't since I got the how-to-do-it for realchute from MJ's source. And yes, the mass indicator is broken on that part, but I beleive m4v's RCS balancement plugin fixes that.

[Jouni]

The massless RCS thrusters feel almost cheaty. I tried to build a RCS-powered SSTO, but my initial design was 200 m/s short of reaching orbit.

[SRV Ron]

The OX Stat, battery ?, four instruments, Commotron, all have a mass of 0.005. This shows up when making small probes and flying them. The one LADEE placed in Mun orbit came in at 0.07 tons. Overall, their mass is so small compared to fuel cans, engines, and other components that it tends to be ignored.

[Padishar]

Then engineer is dumb. I know for a fact that MechJeb doesn't since I got the how-to-do-it for realchute from MJ's source. And yes, the mass indicator is broken on that part, but I beleive m4v's RCS balancement plugin fixes that.

MJ only started doing it correctly a few days ago and KER (the test development version) did the same.

You shouldn't need to change RealChutes as the fix was purely for the VAB rather than in flight (or do you do in VAB calculations?).

[Claw]

MJ only started doing it correctly a few days ago and KER (the test development version) did the same.

You shouldn't need to change RealChutes as the fix was purely for the VAB rather than in flight (or do you do in VAB calculations?).

MJ was smart enough to know some parts were weightless, but it seemed hard coded. If you modified a .cfg, MJ didn't seem to catch on.

KER didn't recognize that the small gear bay was massless but MJ did. That's actually why I chose MJ over KER several months back when I was first deciding. So I know that at least that part was right.

Personally I went in and modified the landing gear .cfg file so that the mass was zero. That way the CoM in the SPH stopped lying to me.

Edited April 12, 2014 by Claw

[cantab]

The massless RCS thrusters feel almost cheaty. I tried to build a RCS-powered SSTO, but my initial design was 200 m/s short of reaching orbit.

Agreed. I'm left wondering why on Kerbin Squad made them massless in .23.5. They weren't in .23. I just threw together the same basic design: an OKTO2, the biggest RCS tank, and as many place-anywheres as I could fit on the bottom and 4-ways on the side without clipping. In .23 it got off the ground and climbed slowly and steadily, exhausting its fuel at around 1000 m. In .23.5 it went like a bat out of hell and reached over 25 km apoapsis.

[Padishar]

Yes, MJ used to use zero mass for the landing gear in the VAB/SPH and KER didn't. Also, both mods didn't correctly account for other parts with PhysicsSignificance set in the VAB/SPH. Both now should be working "correctly"...

I think it might be better for parts with PhysicsSignificance set to 1 to have their mass added to the mass of the parent part instead of just ignored. This would avoid the worst of the imbalance issues (assuming you attach stuff to something whose COM is already aligned with the whole vessel COM) without adding more issues by the parts being massless...