Kerbal Space Program 1.7.3 is live!

[Noir]

13 minutes ago, Acvila said:

just decrease motor size, my plane consumption is 0.03. decrease motor size in editor and don't increase torque if rotor is spinning at full velocity

God damn your right.

That's not how engines work. Fuel consumption should be based on RPM, not torque, at least with internal combustion engines. I -kinda- understand why the electric engines would, but not the fuel based ones. And torque should be based on RPM. Both RPM and torque should be based on how much fuel is being pushed into the motor.

Decreasing the torque on an engine that is already at max RPM down to 1-5 kN brings the fuel usage down to 0.2, less than the 0.4 listed value... And that was with 2 engines...

I don't know anymore..

Either do as Acvila said, or lower the torque once in flight and at max RPM.

Edited July 12, 2019 by Noir
Wat

[wookiee_goldberg]

7 hours ago, 5thHorseman said:

* Add ability to switch secondary docking ports (docked) to primary via the PAW.

What does this do, other than the obvious (switch which docking port says 'undock')? In other words, under which circumstance would I need or want to use it?

It might have implications on whether or not "Undock" action in an action group is allowed to do its job. And also maybe if you have a docking port that's buried inside a surface and hard to click, this makes it so you can easily click Undock on another docking port that is more visible and easier to get to.

7 hours ago, 5thHorseman said:

* Convert same vessel docking connections to real docked connections whenever the vessel undocks or decouples elsewhere on the vessel.

What is the reason for this? Is it that a ship needs one "real" docked connection and the rest are "fake" ones? If you have 10 docking ports connected and undock one of them, do the other 9 become "real" ones or does the game just make sure there is always one "real" one? And finally, does this matter to the players in any way or can we just ignore it?

I think it's more like the "Kanadarm" situation, and docked-part-loop behavior (same-vessel connections). If you are in a MK3-based spaceplane docked to a space station through the shielded nose docking port, that is a "real" docking connection. When you use your Kanadarm to grab a payload from your cargo bay, that would be a "fake" docking connection. But then when you decouple the payload from the rear wall of the cargo bay, the fake connection becomes a real one, so that it's securely attached to the end-effector of your Kanadarm with all the normal functionality of a docked connection. It should be possible for a craft with 10 docking ports connected to stuff to all use "real" docking connections as long as there are no loops created by the docking ports - just imagine a station core with 10 docking ports and 10 capsules docked to those ports, those would all be "real" docking connections.

 

[Acvila]

5 minutes ago, Noir said:

God damn your right. That's not how engines work. Fuel consumption should be based on RPM, not torque, at least in internal combustion engines. I -kinda- understand why the electric engines would, but not the fuel based ones.

Either do as Acvila said, or lower the torque once in flight and at max RPM.

decreasing motor size in editor give you also a lighter motor. 

[wookiee_goldberg]

 

9 minutes ago, KerikBalm said:

It seems they also changed something so that you can't do regenerative breaking, and excess power is dumped (which is why I need to monitor rotor RPM, any more throttle than needed to maintain RPM just wastes energy)

looking at the .cfg here:

They rid of an input/output system that can reclaim energy, and have one where excess/negative power is dumped.

For me, this is a big BOOOO

You can't deploy a rotor as a windmill now, for example when landing a rapier powered jet that is nearly out of EC

I think it was already like that before - i remember seeing those things commented out even in 1.7.1 when BG was just released. Looks like something they thought about doing, but decided against.

Just now, Noir said:

God damn your right. That's not how engines work. Fuel consumption should be based on RPM, not torque, at least in internal combustion engines. I -kinda- understand why the electric engines would, but not the fuel based ones.

Either do as Acvila said, or lower the torque once in flight and at max RPM.

Torque-based fuel consumption is correct even for combustion engines. If you have something that's spinning at a certain RPM but doesn't need much torque to sustain it, then that shouldn't use much fuel (coasting downhill in a car). If you need to add torque to sustain your RPM, that should use more fuel (climbing a hill in a car). If you have a manual transmission car with cruise control on and a loud enough engine, you can hear the volume of the engine getting louder and quieter as you go up and down hills with varying torque demand, while the RPM of it (and the car's speed) remain constant.

 

[Noir]

1 minute ago, wookiee_goldberg said:

Torque-based fuel consumption is correct even for combustion engines. If you have something that's spinning at a certain RPM but doesn't need much torque to sustain it, then that shouldn't use much fuel (coasting downhill in a car). If you need to add torque to sustain your RPM, that should use more fuel (climbing a hill in a car). If you have a manual transmission car with cruise control on and a loud enough engine, you can hear the volume of the engine getting louder and quieter as you go up and down hills with varying torque demand, while the RPM of it (and the car's speed) remain constant.

That makes complete sense, but unless my understanding of combustion engines is completely wrong, or I'm overthinking it, the new rotor engines don't make all that much sense realistically.

[wookiee_goldberg]

Just now, Noir said:

That makes complete sense, but unless my understanding of combustion engines is completely wrong, or I'm overthinking it, the new rotor engines don't make all that much sense realistically.

Back in 1.7.1 / initial BG release, i did notice that the concept of "how much torque is needed to do this thing" was just following the maximum setting rather than using a smaller number based on actual needs... so maybe that's still in-play here. Like as if you were driving around with your gas pedal floored, and using your brakes to moderate your speed. I was asleep when the update dropped and then at work, so I haven't gotten any play-time with the new stuff (and updated old stuff) just yet.

[mystifeid]

Alternators for the turboshaft engines would be nice. I add the torque limit for the turboshaft engines to the main throttle action group and seem to be able to cruise at 225m/s in the four engine plane I made with the torque set at 25%. So it's great I'm not running out of fuel so quickly but the Mk1 cockpit only has 50ec and SAS eats that before fuel is consumed. Looks like I'm adding batteries or ec generation.

[KerikBalm]

The prop engines still need their consumption reduced.

They aren't competitive with jet engines.

Observe the prop RPM, its not that my throttle setting or engine size is too high:

compared to the same craft, but with panthers in dry mode (not even wheesleys)

Faster, much higher top speed, lower fuel consumption, much easier to use.

 

[Noir]

18 minutes ago, KerikBalm said:

-snip-

 

I got a dual engine setup to use only 0.02. Have you tried decreasing/increasing the prop pitch/authority limiter further? My torque is using 4.5 kN,  and when I lower it the craft will pick up speed. But if I lower it to much, the engine with lose RPM.

I am using FAR, and I don't think the new prop are playing nice with FAR, but the same logic should still apply in soup stock aero.

Edited July 12, 2019 by Noir

[mystifeid]

When I lower the turboshaft torque to 25% the rpm drops from 460 to 415 but visually the props look like they are hardly spinning.

[Kerbart]

4 hours ago, 5thHorseman said:

If you have SAS on

In KSP? I haven’t tried it out TBH, but that’s not how it works in reality.

Why would the ship counter rotate once a part rotates at a constant rate? Yes, there is torque generated by the rotor, but it’s balanced by the resistance it has to overcome. If it weren’t balanced there wouldn’t be a constant rate of rotation. But I can see how KSP doesn’t model internal friction, and the the vessel would indeed counterrotate.

i heard the responsible NASA engineer once explain that it’s a lot easier to keep the solar panels rotating on the dark side, because there’s no need to activate the servos to stabilize the ISS—only when they’re started and stopped. Of course, putting them in “vane mode” lowers atmospheric resistance so it’s not always done like that, but that’s a different story.

[klgraham1013]

Wow.  Does KSP really use 6GB of ram now?  All the people with potatoes are going to start complaining again.

[Maxsimal]

17 hours ago, Bubbadevlin said:

"re-balanced all rotors"

Are there numbers for this somewhere, or am I going to have to look in game?

Either way, great patch! So many QoL additions and cool new mechanics!

This actually goes beyond just the numbers.  We changed how the PhysX joint is configured, the rotor behavior is more stable now - if you watch the RPM figures they will hold a lot steadier than before.  

[dok_377]

1 hour ago, klgraham1013 said:

Wow.  Does KSP really use 6GB of ram now?  All the people with potatoes are going to start complaining again.

*Laughs in EJ's computer*

[Yargnit]

16 hours ago, Wallygator said:

why do the turboshaft engines show up under "robotics"?

They are basically the electric motors, just with their resource's changed from EC to LF/Air as far as the game in concerned. Really they'd fit equally in either category, but this makes them a bit easier to find because that tab is less crowded.

7 hours ago, Klapaucius said:

Has anyone else noticed issues with power consumption in the new update? My frog and another rover, which worked fine before now suck down power like there is no tomorrow. The fuel cells cannot at all keep up whereas previously, both had reasonable range. My frog makes 3 jumps and then collapses. My Kant rover dies after about 30 seconds.

The usage model has changed, you'll see higher peak usage in some cases, but other cases will see much less EC used.

5 hours ago, KerikBalm said:

Looked in the files, there is a "reference consumption velocity" and for the large rotor, its set at 230... I think this is 230 RPM. Thus if spinning at 460, then it seems the velocity is 2x what the "reference" is, perhaps that is why they consume 2x as much as they did before.

On the other hand, my old quad tilt rotor design does perform better in some aspects. I replaced its blades (servo+strut+wing segment, 3 parts each, 6 parts per rotor) with 2x medium helo blades each (4 parts per rotor), and reduced the engine size of each rotor to 50%, it still used a bit more fuel than before, but it lifted more and flew faster.

I think changing the reference velocity to 460 instead of 230 will do what needs to be done. Maybe increase the efficiency value (its at 0.75 for the heavy rotor).

As for the LF engines, to compere with turbofans, its base fuel consumption needs to be halved, at least (combined with the change of reference velocity to 460)

The current fuel use is correct per intended design. ReBut yes you are correct with the reference of 230 meaning the listed fuel usage in the editor is for 50% max RPM.

5 hours ago, KerikBalm said:

It seems they also changed something so that you can't do regenerative breaking, and excess power is dumped (which is why I need to monitor rotor RPM, any more throttle than needed to maintain RPM just wastes energy)

looking at the .cfg here:

They rid of an input/output system that can reclaim energy, and have one where excess/negative power is dumped.

For me, this is a big BOOOO

You can't deploy a rotor as a windmill now, for example when landing a rapier powered jet that is nearly out of EC

This was never actually a working thing, it's been commented out the entire time.

4 hours ago, wookiee_goldberg said:

Back in 1.7.1 / initial BG release, i did notice that the concept of "how much torque is needed to do this thing" was just following the maximum setting rather than using a smaller number based on actual needs... so maybe that's still in-play here. Like as if you were driving around with your gas pedal floored, and using your brakes to moderate your speed. I was asleep when the update dropped and then at work, so I haven't gotten any play-time with the new stuff (and updated old stuff) just yet.

In 1.7.1/1.7.2 the torque and rpm sliders both effected max rpm and torque. (so 50% torque 50% rpm meant max rpm and max torque were both 25%) Now you can reach max rpm at 1% torque if there is no resistance, or you can have max torque at 5 rpm if that's you're desire. 

4 hours ago, KerikBalm said:

The prop engines still need their consumption reduced.

They aren't competitive with jet engines.

Observe the prop RPM, its not that my throttle setting or engine size is too high:

 

compared to the same craft, but with panthers in dry mode (not even wheesleys)

 

Faster, much higher top speed, lower fuel consumption, much easier to use.

 

In some scenarios (higher speeds specifically) the jets will be way better. Doing helicopter type stuff the turboshafts with blades can actually lift more than jets with the same fuel consumption. Basically when you're blades can be flat it will take way less torque to keep them spinning so you can generate more low speed power with low fuel consumption. When you're trying to go fat and have to pitch your blades significantly you'll need more torque and fuel so jets will be more efficient.

4 hours ago, Noir said:

I got a dual engine setup to use only 0.02. Have you tried decreasing/increasing the prop pitch/authority limiter further? My torque is using 4.5 kN,  and when I lower it the craft will pick up speed. But if I lower it to much, the engine with lose RPM.

 

I am using FAR, and I don't think the new prop are playing nice with FAR, but the same logic should still apply in soup stock aero.

The blades are very unlikely to work with FAR do to the way they are configured. You can't spin any part faster than 477rpm in kerbal without krakening (1 radian per physics tick). This is a pretty hard game engine limitation. The blades have to 'fake' the lift of spinning much faster than they really are to create enough thrust to work effectively. (This is why before this everyone had giant unwieldy looking props)

To do this, the prop tells the game that it's located much farther from the rotor shaft than it really is so it's effective velocity is much faster. (If you turn on aero forces you can see this point as purple lines). Since FAR models it's own aero it won't be reading this offset, so it just seems it as a small slow spinning prop, rather than the large slow spinning prop the game is tricked into thinking it is under stock.

 

Hope this clears up some of the prop questions.

[_stilgar_]

I have a problem with new version - every few minutes I get a message from surface experiments saying that they've added 0.005 science to my account. Warp for few hours and I get spammed with 50 such messages. 

Does anyone else get that "screen message" spam? Is there a way to turn it off? 

[Superfluous J]

3 hours ago, Kerbart said:

Why would the ship counter rotate once a part rotates at a constant rate?

What force stops it?

If you rotate something 100rpm and your station is 100x the mass of the thing, your station will rotate 1rpm the other way.

[wookiee_goldberg]

43 minutes ago, Yargnit said:

The usage model has changed, you'll see higher peak usage in some cases, but other cases will see much less EC used.

In 1.7.1/1.7.2 the torque and rpm sliders both effected max rpm and torque. (so 50% torque 50% rpm meant max rpm and max torque were both 25%) Now you can reach max rpm at 1% torque if there is no resistance, or you can have max torque at 5 rpm if that's you're desire. 

Yeah overall it's much better than before (now that i've been able to play with it some this morning). The "current torque usage" is still kind of broken though.. when current_rpm is less than target_rpm then it's fine for current_torque to equal max_torque because you're making best-effort to spin your thing up to speed. But once current_rpm >= target_rpm, then a PID controller should kick in and manage the torque output (and resource usage) below the "max torque %" setting to govern the current_rpm as closely as possible to the target_rpm - after all, max torque is only supposed to specify the MAX torque available, not the current torque demanded. So if you can hit peak RPM with 2% of max torque then having the "max torque" slider on 100% should still only use 2% until the situation requires more. The max torque% slider should just affect how quickly/violently the motor will try to respond to being under-speed.

Edited July 12, 2019 by wookiee_goldberg

[Yargnit]

9 minutes ago, wookiee_goldberg said:

Yeah overall it's much better than before (now that i've been able to play with it some this morning). The "current torque usage" is still kind of broken though.. when current_rpm is less than target_rpm then it's fine for current_torque to equal max_torque because you're making best-effort to spin your thing up to speed. But once current_rpm >= target_rpm, then a PID controller should kick in and manage the torque output (and resource usage) below the "max torque %" setting to govern the current_rpm as closely as possible to the target_rpm - after all, max torque is only supposed to specify the MAX torque available, not the current torque demanded. So if you can hit peak RPM with 2% of max torque then having the "max torque" slider on 100% should still only use 2% until the situation requires more. The max torque% slider should just affect how quickly/violently the motor will try to respond to being under-speed.

yeah, you have to throttle the torque manually once you hit target rpm. Also, if you can only his say 350 rpm at 100% torque, make sure you set your RPM limit to 350. you'll use less EC/fuel with the RPM limit set to 350 than 460, even if it takes max torque to reach the 350. 

[wookiee_goldberg]

1 minute ago, Yargnit said:

yeah, you have to throttle the torque manually once you hit target rpm. Also, if you can only his say 350 rpm at 100% torque, make sure you set your RPM limit to 350. you'll use less EC/fuel with the RPM limit set to 350 than 460, even if it takes max torque to reach the 350. 

Yeah that's weird too. Both of those things should be based on the current RPM and torque, not the maximum RPM and torque. The way those things are currently kind of bypasses the law of conservation of energy (maybe this is what feeds the Kraken).

[Maxsimal]

Some improvements may be made in how the LF rotors work, calculate their consumption and communicate it to the player in the future. 

However, the overall fuel usage is unlikely to change.   This craft that I threw together can just about circumnavigate Kerbin - fewer passenger compartments and more fuel would easily allow it to do so.  That's more than enough efficiency for Kerbal - we understand that a real long-haul turbo prop would be able to go 



Keep in mind that propellers are harder to use the jet engines - that's just part of how we implemented, and also mirrors reality.

[wookiee_goldberg]

3 minutes ago, Maxsimal said:

Some improvements may be made in how the LF rotors work, calculate their consumption and communicate it to the player in the future. 

However, the overall fuel usage is unlikely to change.   This craft that I threw together can just about circumnavigate Kerbin - fewer passenger compartments and more fuel would easily allow it to do so.  That's more than enough efficiency for Kerbal - we understand that a real long-haul turbo prop would be able to go 



Keep in mind that propellers are harder to use the jet engines - that's just part of how we implemented, and also mirrors reality.

The consumption when torque is limited to perfectly maintain RPM is totally fine - that's where it ought to be. Only the excessive consumption when the torque is set to 100% (but RPM is already maxed out) would need to change. I had a thing that could max out the RPM in flight with just 3% torque, and the consumption was great - increasing torque slider to 100% caused 33x as much fuel consumption with no change in the flight of the vessel.

[Bartybum]

This update is so great. Really enjoying trying to learn about variable pitch and constant speed propellers by messing around.

A small thing I'd like to see though - regarding action grouping, it'd be really nice to be able to reverse the deployment direction for blades and control surfaces via action group, as this would allow quick reverse thrust on landing. It can already be done by incrementally limiting authority, but a quicker method would nonetheless be nice.

Another thing would be to have a 0.625m equivalent air breathing turboshaft for smaller aircraft, or at least more 0.625m LOX tank options to use with electric rotors and fuel cells.

Edited July 12, 2019 by Bartybum

[KerikBalm]

@mystifeid I forget the term for this, but that is because the period of the rotation is similar to a multiple of the frame rate.

With some helos, thete are cool videos in teal life where it looks like the blades are stopped.

9 hours ago, Noir said:

I got a dual engine setup to use only 0.02. Have you tried ...

I am using FAR...

1) Its not about getting a low fuel consumption rate, its about the relative fuel consumption of a prop and a jet when nothing else changes, and firmly in the subsonic realm. Sure I could drop from 160 m/s to 115 m/s, lower blade pitch, and use less fuel. Thats not the point. The point is that they are not competitive with turbofans even at low speeds.

Also, I'm pretty sure that I could have done better with the old rotors and a fuel cell.

They could have just given us a bimodal fuel cell that can us lf +ox, or lf+intake air.

The old heavy rotor+ fuel cell combo had a lf consumption rate of less than 0.02/ sec. Even including ox, it was less than 0.045 units/ sec max (since 1 fuel cell powered 1 heavy rotor with power to spare) and that's for a lot of torque.

2) you can't really compare far to non far

*edit* this post was supposed to be made much earlier in the thread, but it didn't send when I thought it did, and thus I sent it much later

Edited July 12, 2019 by KerikBalm

[Kerbart]

1 hour ago, 5thHorseman said:

What force stops it?

If you rotate something 100rpm and your station is 100x the mass of the thing, your station will rotate 1rpm the other way.

When you start at 0 and speed up to 100 rpm your station will rotate, so that you will have to stop.

once you’re at a constant rate there are two options:

1) your bearings are frictionless. No additional power is required and the whole thing just keeps rotating by itself. There are basically two masses; the space station, and, say, the habitat ring (or whatever rotates) and as there is no interaction between the two, one doesn’t influence the other. Some will  argue that this is not a realistic scenario and I tend to agree with that.

2) there is friction in the rotation cuff/bearing system. Let’s for argument sake, put this at a concrete value of x Nm. To make sure your ring doesn’t speed up or slow down, you drive it with a rotor that provides torque in the opposite direction of the friction; so that would be -x Nm. The total amount of torque that is applied to the space station is x - x Nm, or 0. Without a net torque, the station will not rotate.

For clarity, in scenario (2) when your turn off the rotor, the ring will slow down due to friction, and this friction will apply torque to the space station and will make it rotate.

So you will need SAS or RCS when starting or stopping a habitat ring (or solar panels), to keep the station from spinning (and the SAS system will have to be reset at one point). But in a “stationary” situation (ring spinning at a constant rate) you will not.