Realism Overhaul

[Starwaster]

At ground level it won't be, in space the difference is noticeable as the rocket it much smaller compared to the planet at the same heigth and those are, of course, your references in space. It is not like it is negligible, as a craft at 64% equates to a planet that is actually 156% of the craft scale. That is a huge difference!

Apparently they are not, NathanKell stated they are 64%.

jebs yardstick says they are. is it lying?

[Guest]

I agree that there should be some sort of automation for that rescaling, otherwise it'll be a day job for all of us to manually rescale all the parts in all mods (not to mention that a fair amount of them prohibit such distribution).

Note, licenses are not a problem for this. Config files referencing the models and textures can be written independently of the mods containing them and even put into their own folders. MODEL calls can be used for that, AFAIK there's no mod with a special license provision against them. Not to mention we have Module Manager, while I'm not terribly fond of it, it'd probably be the best option for manual alteration of various mods. Again, no mod forbids writing MM configs for it.

Still, an automatic rescaler would probably be the best option. I've had bad experiences with rescaling parts that use MODEL calls, so KOSMOS for example could cause us a lot of grief. I've had one of their parts "come apart" when rescaled (basically, it seems that parts got rescaled, their positions did not), so a plugin would probably be a much less painful option.

Edited October 26, 2013 by Guest

[TheCanadianVendingMachine]

Can we see your development schedule? As in: What you will do next. Something like Frizzank did

[Guest]

Well, the next thing to do would be to iron out all the imperfections around Kerbin. There are still plenty. Once that's figured out, the mod would probably expand to other planets.

[AndreyATGB]

Well, the next thing to do would be to iron out all the imperfections around Kerbin. There are still plenty. Once that's figured out, the mod would probably expand to other planets.

Most important issue right now is rocket instability on pad. Kerbin can look amazing but it's hard to even launch right now.

[ferram4]

I'm currently working on a plugin to try and fix the instability on the pad and tighten up rockets as a whole; ultimate goal is that a purely vertical stack of diameter greater than 1.25m shouldn't need struts between the parts to not flex all over the place / engines shouldn't dance around under high thrust. Should be ready for a first release soon. Currently fixes parts sliding around but doesn't fix angular movement, which is really the big problem.

[NathanKell]

For rescaling...yeah. Batch would be good, but scaling is inconsistent. Some part makers use 1:1, some use 80%, some use 64%. And often while crewed craft are scaled down, uncrewed probes are left 1:1.

I mean, heck, for Squad, the two capsules are roughly 64% (the Mk1 is more like 66%), but the probes are all over.

Probably the best bet would be to write an offline script that would process a list of partnames and rescale values, and output corrected part CFGs. I mean, you could do it ingame too, I guess.

On another note: Just to see if everything's working right, Dragon01, do you want to try manually rescaling and matching data for the Soyuz? I'll handle Gemini. We can then see if they perform the way they should given correct stats. Or if the MFS stuff is confusing just get the sizes and dry weights right and I'll do engine/tank configs for you.

(Those are the two LV/spacecraft configs that I know are complete simulations--are their other complete models with many real-world stats available to check?

[NathanKell]

ferram, great! Another thing you might try at least as a temporary thing is to write your own launch clamp module that keeps parts rigid until it's disengaged, as a lot of the stability problem is the first few milliseconds on the pad. I mean, in flight things dance around some, but the first few ms are killer. And LaunchClamp makes things worse rather than better at the moment, must be something weird in the joint code.

[asmi]

ferram, great! Another thing you might try at least as a temporary thing is to write your own launch clamp module that keeps parts rigid until it's disengaged, as a lot of the stability problem is the first few milliseconds on the pad. I mean, in flight things dance around some, but the first few ms are killer. And LaunchClamp makes things worse rather than better at the moment, must be something weird in the joint code.

The purpose of clamps is exactly to allow simulation to settle down, if you disable physics before start, it will starts settling down during launch, and you don't want this to happen

I'm off today to finalize alpha version of ECLSS mod - BobCat is waiting, and I just can't let that man wait

Edited October 26, 2013 by asmi

[Guest]

(Those are the two LV/spacecraft configs that I know are complete simulations--are their other complete models with many real-world stats available to check?

Well, Proton is pretty well documented, but it's not a complete vehicle, meaning you need a payload. Also, if you use BobCat's fairings, you might have screwed-up aerodynamics, since they're much bulkier than the real deal. Ariane 5 from ESA pack with ATV would be another good choice. Space Shuttle (the CSS) is a heavy WIP, but we've got tons of tech data on it. I think I'll do the Soyuz TMA first, then we'll see. Oh, I don't like the Module Manager, so the configs will be drop-in replacements for the time being. With the amount of data being changed (i.e. pretty much everything), I think this makes sense.

[ferram4]

@NathanKell: If you want a gauge of how things are going, consider this rocket:

It has two stages and just enough dV to put that tiny little Apollo-CSM-like vehicle in orbit.

It has two struts on it, total. Two. No unreasonable wobble in flight at all; engines don't slide around or anything funky like that. No other boosts to joint strength. The only reason there are any at all is because the 2nd stage engine to decoupler to fuel tank joint chain isn't stiff enough. I figure I'll add a some code to make sure that the parts connected to stack decouplers are strutted together, play with some numbers, and call it good.

As for the Launch Clamp issues... those are the nature of the beast; adding too many constraints can make things much, much worse because the clamp joints fight each other.

[Guest]

OK, could somebody tell me what units KSP uses? I'm assuming kN for thrust, T for mass and kW for electric charge. But I could be wrong.

[ferram4]

Thrust is in kN. Mass is in metric ton / tonne. Electric charge is kind of arbitrary I think... try kJ or kW-h; keep in mind it has to be units of energy, not power.

[ANWRocketMan]

Looking good @ferram!

[asmi]

It has two struts on it, total. Two. No unreasonable wobble in flight at all; engines don't slide around or anything funky like that. No other boosts to joint strength. The only reason there are any at all is because the 2nd stage engine to decoupler to fuel tank joint chain isn't stiff enough. I figure I'll add a some code to make sure that the parts connected to stack decouplers are strutted together, play with some numbers, and call it good.

Wow! "I want it! I want it so bad it's driving me mad!" © The Beetles

[NathanKell]

Wow! "I want it! I want it so bad it's driving me mad!" © The Beetles

QFmaximalT.

Dragon01, EC is, if you go by the RTG, kW-s. If you go by the lights, much less than a kW-s. I would suggest using maybe deka-watt-second? So 1Kw = 100EC/s.

Does anyone have a rough estimate on weight vs. charge for batteries over the last 50+ years? I know it's advanced by leaps and bounds, but I don't know numbers.

So far this graph is all I found: http://www.batterypoweronline.com/images/PDFs_articles_whitepaper_appros/ReddyBatteryFuture.pdf

[Guest]

Well, I'm gonna take kW-s and adjust the lights. I'm trying to get Soyuz to work now. That thing's gonna be really hard to fly, it's main engine is terribly weak and carries very little fuel.

Also, MFT could use some way of making bipropellant RCS. It's a rather common solution, yet KSP doesn't allow that.

Edited October 26, 2013 by Guest

[Legwan]

One suggestion regarding atmosphere height - would it be possible to increase the value even higher without scalling density at lower altitudes?

Based on the neat source I've found PDF (+ Implemententation):

"In all but the heaviest satellites (those with a mass to area ratio well in excess of 100 kilogram per square metre), the actual lifetime from an altitude of less than 180 km is only a few hours."

In fact if one plays with the calculator even 180-200km results with rapid decay and the lowest orbit for the satellite was 215km (Sputnik 1 via Wikipedia).

Of course it's more a matter of self control here if we want real life behavior (like not parking at 150km orbit for > 1 revolution). Nonetheless further increasing the atmosphere height is tempting

BTW I've done some more research on orbital decay and it seems like a nice idea for a plugin - in short: all satellites in LKO need dV to maintain the orbit (amount depending on height), scan through all flights, deduct dV, if no left play with orbital parameters next time vessel becomes active, Ioncross does something familiar just with life support. I'm not a programmer so feel free to steal the idea.

Finally I'd like to thank developers of this mod - You've fulfilled my last wish towards this game.

Felling like it should be a separate mode unlockable at some stage in career more (kind research on space telescope/planet detection + placing big junk in orbit)

Edited October 27, 2013 by Legwan
grammar bits

[NathanKell]

Yeah, I was about to post that we should just do kW-s, otherwise the battery size is insane.

Woking on Gemini right now.

I'll see about RCS; the problem is the game seems to only allow one type of resource for RCS, and even changing that is hard (you have to set the resourceName *and* call SetResource() on the RCS module). I'll see if I can make my own module later maybe.

Maybe I could make a hack plugin with a fake single resource that, every time it's used, replenishes it from the appropriate bipropellant resources?

[Guest]

Heh. I just realized you ended up with a much easier task than me. Gemini is a simple 2-stage rocket with spacecraft on top, Soyuz has an intricate LES system (that I'm also going to replicate), 3 stages and a rather complicated spacecraft. Right now, using MFT fuel values made the thing somewhat overweight. That probably means those engines run fuel-rich, but I haven't been able to find a proper ratio for anything but the Soyuz Service Module engines (which seems to run oxidizer-rich for some reason).

Edited October 27, 2013 by Guest

[MedievalNerd]

@NathanKell: If you want a gauge of how things are going, consider this rocket:

Can we try it out too? The wobble is killing me. Literally.

[asmi]

One suggestion regarding atmosphere height - would it be possible to increase the value even higher without scalling density at lower altitudes?

Based on the neat source I've found PDF (+ Implemententation):

"In all but the heaviest satellites (those with a mass to area ratio well in excess of 100 kilogram per square metre), the actual lifetime from an altitude of less than 180 km is only a few hours."

In fact if one plays with the calculator even 180-200km results with rapid decay and the lowest orbit for the satellite was 215km (Sputnik 1 via Wikipedia).

Of course it's more a matter of self control here if we want real life behavior (like not parking at 150km orbit for > 1 revolution). Nonetheless further increasing the atmosphere height is tempting

BTW I've done some more research on orbital decay and it seems like a nice idea for a plugin - in short: all satellites in LKO need dV to maintain the orbit (amount depending on height), scan through all flights, deduct dV, if no left play with orbital parameters next time vessel becomes active, Ioncross does something familiar just with life support. I'm not a programmer so feel free to steal the idea.

Finally I'd like to thank developers of this mod - You've fulfilled my last wish towards this game.

Felling like it should be a separate mode unlockable at some stage in career more (kind research on space telescope/planet detection + placing big junk in orbit)

I was trying to do something like that, but faced odd behaviour of KSP's Orbit class - but that was back in 0.19-ish times. Maybe I'll try to do that again once I get a second. Such mod would definetly be very useful as you won't need to worry about final stages flying around just above the edge of atmosphere.

[asmi]

Heh. I just realized you ended up with a much easier task than me. Gemini is a simple 2-stage rocket with spacecraft on top, Soyuz has an intricate LES system (that I'm also going to replicate), 3 stages and a rather complicated spacecraft. Right now, using MFT fuel values made the thing somewhat overweight. That probably means those engines run fuel-rich, but I haven't been able to find a proper ratio for anything but the Soyuz Service Module engines (which seems to run oxidizer-rich for some reason).

Why don't you just take BobCat's Soyuz-U/FG and change parameters to real ones from Wiki or some other source? The beauty of realism is that you don't need made-up number - use real ones

Edited October 27, 2013 by asmi

[NathanKell]

Astronautix has the fuel ratios for Soyuz U and Soyuz FG engines.

Yeah, Titan II GLV and Gemini is easier, but it's still wicked hard getting the spacecraft weights and Isps exactly right. Still wouldn't trade you though.

asmi, I've been thinking about that too, but I think I'll leave that to you.

I'm already maintaining like five mods, although with help. :]

[asmi]

I'm already maintaining like five mods, although with help. :]

Fine, but can you please please please add option to disable all heat production for engines in MFSC? I really hate this stupid system!