[0.25]KSP Interstellar (Magnetic Nozzles, ISRU Revamp) Version 0.13

[forsaken1111]

With all the posts about people adding this mod and then having problems with previous craft, do you think it would be worth it to make a new type of solar panel which works with your mod and leave vanilla panels alone? Some kind of concentrating solar panel which integrates with your thermal mechanics. You could even make it glow when it begins to overheat as a visual representation for the player, and you wouldn't have to worry about breaking existing craft for others by making their static panels explode or not work.

[RadHazard]

Yeah, all the solar panels except the OX-STAT have had their solar curve tweaked.

is there a reason why the OX-STAT panels are treated specially? it seems like a bug that waste heat has no effect on them, and their power output doesn't scale either according to this post.

or is there a need for stock panel behavior?

Fractal_UK, is this intentional?

This was true a few versions back when I posted this, but has since been changed.

[veldril]

Hi. I'm a new user of this mod and have some question about the DT Vista engine that I could not find in this thread or wiki.

According to the description in wiki, the engine should vary it thrust and ISP depending on the level of thrust I currently use. Yet when I try varying the % of thrust, the ISP change but the amount of thrust remain constant at 1,100 kN (also listed in the wiki but not sure if it is correct or not). Is this behavior of the engine correct? If so then what is the different or benefit of using max thrust level compare to lower thrust level of this engine?

Thank you very much in advance.

[Fractal_UK]

Is it possible to have the panels destroy themselves if they overheat?

The ability to auto-retract is important, as it saves the panel from destruction from overheating....but if the panel can't retract, why not have it blow up?

That would alert the user that the panel failed, and would encourage the use of radiators, and the auto-retracting panels would become more valuable since a probe using auto-retracting panels could be restarted by a kerbal in EVA after it cooled.

what do you think Fractal_UK?

It's possible but I'm not 100% sure about what I want to do about it, exploding panels and such like seems a bit excessive so it might be more on a minor malfunctions sort of level if I do anything. If ships are able to operate fine with just those small solar panels and 100% WasteHeat, it doesn't matter all that much. I might up the passive dissipation a bit so that it's rarely a problem, those small panels are so tiny they wouldn't need much cooling (except perhaps on a similarly tiny probe).

With all the posts about people adding this mod and then having problems with previous craft, do you think it would be worth it to make a new type of solar panel which works with your mod and leave vanilla panels alone? Some kind of concentrating solar panel which integrates with your thermal mechanics.

It's important to have the WasteHeat mechanics on solar power in general because it's in theory possible to generate a huge amount of power using low-orbit solar arrays and microwave transmitters, besides it works well with everything else generating heat and it makes sense. Real satellites powered by solar panels most certainly have to worry about dissipating heat.

Thinking about it though, I should probably add passive convective dissipation (in addition to passive radiative dissipation) so that people's existing panels in atmospheric environments don't overheat because that's a bit silly.

Edit: Passive convection done.

Edited December 11, 2013 by Fractal_UK

[Deredere]

Real satellites powered by solar panels most certainly have to worry about dissipating heat.

I'm sure if you thought about that you'd realize it's not true. A solar panel doesn't absorb any more heat energy per unit area than any other material; in fact, it absorbs less, because some portion is converted to usable energy and transported elsewhere.

If an asteroid won't overheat, neither will a solar power system.

Closer to the sun, this is less true, but even then whether something overheats has nothing to do with solar power.

[forsaken1111]

It's important to have the WasteHeat mechanics on solar power in general because it's in theory possible to generate a huge amount of power using low-orbit solar arrays and microwave transmitters, besides it works well with everything else generating heat and it makes sense. Real satellites powered by solar panels most certainly have to worry about dissipating heat.

Thinking about it though, I should probably add passive convective dissipation (in addition to passive radiative dissipation) so that people's existing panels in atmospheric environments don't overheat because that's a bit silly.

Edit: Passive convection done.

Er... my point was to remove the power curve changes you made to vanilla panels as well. Leave them unchanged, and add your own which operate in your updated manner. Unless I am misunderstanding you and you're saying that the stock panels provide massive power in low solar orbit even without your changes?

The passive convection idea is good though. Would you also consider adding some extra heat convection to the radiators based upon your current speed through atmosphere? It seems to me that flying within a certain range of speeds would make your plane's radiators cool quite a bit faster than sitting still. Obviously above a certain speed that effect would be lost by the compression heating of the air.

Might be too complicated...

Edit: And yes I realize that real satellites have to deal with heat but I think come confusion here is coming from you modifying stock parts. Just my 2 cents though, it is your mod obviously.

Edited December 11, 2013 by forsaken1111

[123nick]

you should add extra hybrid aluminum rockets, liquid methane tanks and engines, and more uranium and throium contianers of standard stack sizes. right now the amount of engines that use special fuels is far far FAR to limited

[forsaken1111]

you should add extra hybrid aluminum rockets, liquid methane tanks and engines, and more uranium and throium contianers of standard stack sizes. right now the amount of engines that use special fuels is far far FAR to limited

This is a good point. I'd really like an aluminum hybrid engine which didn't include a tank for example.

[simmy2109]

I'm sure if you thought about that you'd realize it's not true. A solar panel doesn't absorb any more heat energy per unit area than any other material; in fact, it absorbs less, because some portion is converted to usable energy and transported elsewhere.

If an asteroid won't overheat, neither will a solar power system.

Closer to the sun, this is less true, but even then whether something overheats has nothing to do with solar power.

I disagree. Solar panels DO absorb much more energy, thermal and otherwise. You can think about it as simply as considering their color (which is a function of their absorbency). Solar panels designed for space operation tend to be be a dark shade of blue. That is because that color corresponds to maximum energy absorbency from the Sun. Some of that extra energy is converted to usable electrical energy, but a significant part of that energy is absorbed thermally. Now in real life... is that significant? Idk the answer to that. For example, I am almost certain that the massive panels on ISS aren't themselves actively cooled (although the station is). It may actually be the case that the panels are quite warm and that a good bit of that heat transfers its way to the main structure of ISS (which is quite possible, given the lack of convection cooling on those panel structures).

Long story short, dark colored panels certainly absorb more thermal energy per area than a white colored fuel tank.

I do think it's worth considering though... currently I don't think KSPI models any passive thermal radiation. In reality, a craft in space will radiate some heat without need for discrete 'radiators'. In reality, it is possible to put a low-power satellite into orbit with a small amount of solar panels that can passively radiate all waste heat without need for radiators (I am actually designing such a satellite irl). However, this is a game, not reality. In reality, you also need heaters on spacecrafts to keep components warm, potentially in addition to radiators on other parts. It would be easy to change... just give all crafts a base thermal radiation rate (maybe dependent on size). It's really a design decision about how you want the gameplay to work (I actually don't think anything is "wrong" with the current model). With a small amount of passive radiation, it should possible to put low power satellites into orbit with no radiators at all. I was surprised I couldn't do that actually when I first started using this mod.

[Donziboy2]

If I am not mistaken it is possible to turn the WasteHeat off. Correct me if im wrong Fractal_UK. If it is indeed possible, you have a choice, either play it the way Fractal_UK intended it or turn it off.

[Fractal_UK]

I do think it's worth considering though... currently I don't think KSPI models any passive thermal radiation. In reality, a craft in space will radiate some heat without need for discrete 'radiators'. In reality, it is possible to put a low-power satellite into orbit with a small amount of solar panels that can passively radiate all waste heat without need for radiators (I am actually designing such a satellite irl). However, this is a game, not reality. In reality, you also need heaters on spacecrafts to keep components warm, potentially in addition to radiators on other parts. It would be easy to change... just give all crafts a base thermal radiation rate (maybe dependent on size). It's really a design decision about how you want the gameplay to work (I actually don't think anything is "wrong" with the current model). With a small amount of passive radiation, it should possible to put low power satellites into orbit with no radiators at all. I was surprised I couldn't do that actually when I first started using this mod.

They already have this, the rate is very small because components that aren't designed to radiate heat aren't very effective at it but the passive radiation is there. It depends upon craft size so large craft with little power generation capacity are the place to be for passive cooling.

I also just added today convective passive cooling, which is way way more significant and will make a big difference toward stopping overheating in atmospheres.

[DaMichel]

With a small amount of passive radiation, it should possible to put low power satellites into orbit with no radiators at all. I was surprised I couldn't do that actually when I first started using this mod.

Exactly the point i raised some time ago. Since then i just added radiators to every craft. It's not ideal but i guess i won't uninstall Interplanetary so soon anyway

But from a realism point of view, i was very surprised to read that apparently the cooling system on the ISS is mostly there to get rid of the heat from the solar panels. They even call it "Photovoltaic Radiator Panels". Go figure.

http://www.space.com/21059-space-station-cooling-system-explained-infographic.html

[AdmiralTigerclaw]

They already have this, the rate is very small because components that aren't designed to radiate heat aren't very effective at it but the passive radiation is there. It depends upon craft size so large craft with little power generation capacity are the place to be for passive cooling.

I also just added today convective passive cooling, which is way way more significant and will make a big difference toward stopping overheating in atmospheres.

I forget, are you still enacting passive cooling based on vessel mass?

[forsaken1111]

I forget, are you still enacting passive cooling based on vessel mass?

Until there is a decent way to compute a craft's surface area I think mass is the only way to do it

[RabidMonkey]

Until there is a decent way to compute a craft's surface area I think mass is the only way to do it

I would say part count could be a good factor to consider as well. Lots of little parts would make surface area increase pretty quickly.

[forsaken1111]

I would say part count could be a good factor to consider as well. Lots of little parts would make surface area increase pretty quickly.

Not really. A single big orange tank has more surface area than 4 little lights.

Though now I wonder if there is a way to count the polys in every part on the vessel and base it on that. That might give you a better, though still not 100%, indicator.

Edit: Calculating the area of each poly on the mesh of each part? Might be a bit cpu intensive but you only have to do it once each time the vessel changes. (Staging, decouplers, crashes...)

Ooor do it once for each part the first time the mod loads and then add them up on vessel load?

...

area ABC = 0.5 * | AB X AC |

maybe this? http://forum.unity3d.com/threads/117388-figuring-out-the-volume-and-surface-area

public float Area () {

        Vector3 result = Vector3.zero;

        for(int p = mVertices.Length-1, q = 0; q < mVertices.Length; p = q++) {

            result += Vector3.Cross(mVertices[q], mVertices[p]);

        }

        result *= 0.5f;

        return result.magnitude;

    }

Edited December 11, 2013 by forsaken1111

[RabidMonkey]

Not really. A single big orange tank has more surface area than 4 little lights.

I agree - that's why I said "a factor to consider" instead of "the determinant". If your idea with poly counts works, then I think that would be the best indicator. I was thinking more along the lines of including part count as a factor along with mass. Something like (mass * (1+(part count * [shrinky factor of like .01?]))), where the "shrinky factor" would make it nigh-useless to spam lights on an orange tank.

[simmy2109]

Given our recent discussion on the realism of solar panels and overheating.... I found the timing of this quite ironic.

http://spaceflightnow.com/station/exp38/131211coolant/#.UqkNuvTdB8E

[Eadrom]

So then. 0.23 next week (17th).

Fractal, what are your thoughts on release schedule for KSPI? 0.8.3 bug patch to "finish" off KSPI for 0.22 or are you going to put out a 0.9 to go with 0.23? I imagine it's going to be a bit of work to get the tweakables implemented for everything.

[Fractal_UK]

I agree - that's why I said "a factor to consider" instead of "the determinant". If your idea with poly counts works, then I think that would be the best indicator. I was thinking more along the lines of including part count as a factor along with mass. Something like (mass * (1+(part count * [shrinky factor of like .01?]))), where the "shrinky factor" would make it nigh-useless to spam lights on an orange tank.

You're probably right that there are many better schemes that I could use for passive dissipation - the problem is temperature. Unless you're getting the part to really high temperatures, it's just not going to dissipate a large amount of heat - indeed the whole point of having radiators is that your spacecraft doesn't reach some ludicrous temperature. My fear is that by making a more comlicated simulation of passive heat loss, I'll start with the answer 2 and end up with the answer 2.173 when the answer really needs to be 500 to make a difference.

I'm trying to adopt a policy of, if it doesn't matter, don't simulate it accurately because then you won't see a performance hit for calculations that ultimately make no difference. Instead, you'll have a nice quick calculation that gets the job done and gives you an answer fairly close to the correct one.

Interstellar does quite a lot of calculations so I need to place fairly stringent requirements on what is important and what is not, otherwise slowdown will occur - I already worry about this because most of my testing is done on quite a powerful machine, so I worry I'll end up releasing an update that will kill framerate elsewhere. If you're reading this - tell me straight away if this ever happens!

[Fractal_UK]

So then. 0.23 next week (17th).

Fractal, what are your thoughts on release schedule for KSPI? 0.8.3 bug patch to "finish" off KSPI for 0.22 or are you going to put out a 0.9 to go with 0.23? I imagine it's going to be a bit of work to get the tweakables implemented for everything.

I have already come to the conclusion that 0.8.3 is not worth investing the time in - it's essentially done but I'd rather spend the time I might spend on finalising a 0.8.3 release on content for 0.9. So, I will release version 0.9 after 0.23 is released. I've been doing a lot of preparatory work to ensure that all the basics for 0.9 are in place, when 0.23 releases, I'll be performing some quick tweak(able)s and I hope to be able to release the 0.9 update fairly soon after. Pencil in the 18th for KSPI 0.9 and by some combination of sleep deprivation and caffeine, I'll try to make it happen.

[BlazingAngel665]

I have already come to the conclusion that 0.8.3 is not worth investing the time in - it's essentially done but I'd rather spend the time I might spend on finalising a 0.8.3 release on content for 0.9. So, I will release version 0.9 after 0.23 is released. I've been doing a lot of preparatory work to ensure that all the basics for 0.9 are in place, when 0.23 releases, I'll be performing some quick tweak(able)s and I hope to be able to release the 0.9 update fairly soon after. Pencil in the 18th for KSPI 0.9 and by some combination of sleep deprivation and caffeine, I'll try to make it happen.

Your dedication to this project is amazing. I commend you sir!

[Eadrom]

I have already come to the conclusion that 0.8.3 is not worth investing the time in - it's essentially done but I'd rather spend the time I might spend on finalising a 0.8.3 release on content for 0.9. So, I will release version 0.9 after 0.23 is released. I've been doing a lot of preparatory work to ensure that all the basics for 0.9 are in place, when 0.23 releases, I'll be performing some quick tweak(able)s and I hope to be able to release the 0.9 update fairly soon after. Pencil in the 18th for KSPI 0.9 and by some combination of sleep deprivation and caffeine, I'll try to make it happen.

Awesome! I'll have a job interview in the morning and get some to come home and unwind that night with KSP .23 and KSPI .9 while re-watching the KerbalKon vids.. The next 7 days are turning out to be pretty damn awesome!

What kind of coffee do you like to drink? Cream or sugar?

[simmy2109]

You're probably right that there are many better schemes that I could use for passive dissipation - the problem is temperature. Unless you're getting the part to really high temperatures, it's just not going to dissipate a large amount of heat - indeed the whole point of having radiators is that your spacecraft doesn't reach some ludicrous temperature. My fear is that by making a more comlicated simulation of passive heat loss, I'll start with the answer 2 and end up with the answer 2.173 when the answer really needs to be 500 to make a difference.

I'm trying to adopt a policy of, if it doesn't matter, don't simulate it accurately because then you won't see a performance hit for calculations that ultimately make no difference. Instead, you'll have a nice quick calculation that gets the job done and gives you an answer fairly close to the correct one.

Interstellar does quite a lot of calculations so I need to place fairly stringent requirements on what is important and what is not, otherwise slowdown will occur - I already worry about this because most of my testing is done on quite a powerful machine, so I worry I'll end up releasing an update that will kill framerate elsewhere. If you're reading this - tell me straight away if this ever happens!

Yeah I agree that the detailed 'accuracy' does not matter. For most applications, the base passive radiation will be insignificant compared to the total radiation needed. Personally, I think that the base passive radiation should be high enough for a small satellite with a couple of standard extending panels to radiate sufficiently to not need radiators. Once you move on to bigger things, the passive radiation will be insignificant to your needs anyways. On the other hand, perhaps you like it the way it is, where radiators are needed pretty much from the beginning.

Edited December 12, 2013 by simmy2109

[Myrten]

I have already come to the conclusion that 0.8.3 is not worth investing the time in - it's essentially done but I'd rather spend the time I might spend on finalising a 0.8.3 release on content for 0.9. So, I will release version 0.9 after 0.23 is released. I've been doing a lot of preparatory work to ensure that all the basics for 0.9 are in place, when 0.23 releases, I'll be performing some quick tweak(able)s and I hope to be able to release the 0.9 update fairly soon after. Pencil in the 18th for KSPI 0.9 and by some combination of sleep deprivation and caffeine, I'll try to make it happen.

Sounds like a good decision, but can you put changes already made to \develop on Github?