[1.7-1.9] SMURFF: Simple Module adjUstments for Real-ish Fuel-mass Fractions 1.9.1 (02019 Nov 12)

[NathanKell]

Yeah, didn't have time.

[komodo]

Hey now. Just cause Val is, ahem, MIA, doesn't detract any from the plethora of goodies the ole naught point five brought us. I never said I wasn't having fun  I also think Trajectories, while not explicitly lying to me, mislead my reentry corridor a bit... It's fun having the reentry tiger by the tail, honestly.    

I have this feeling both of you know the tipping point well, where you start to garner as much fun poking at your gamedata folder as you do playing the game, I'm right on the edge as well, having a blast (awesomeness modifier 0.5) personally. Thanks to both for your creations!

[Kerbas_ad_astra]

Alright, one more bugfix release -- now, SpaceY engines get buffed and RAPIERs and the like don't.

• 2015 12 07 (1.1.3): More selection adjustments.
  • Changed engine mass buffs to apply to engines which only have oxidizer-burning modes (so that SpaceY's engines get buffed -- they use a MultiModeEngine module to simulate the all-nozzle vs. center-out/center-only modes of operation, so excluding all multi-mode engines wasn't fair.)

On 12/3/2015, 10:07:54, NathanKell said:

Yeah, didn't have time.

We never do. 

On 12/4/2015, 2:11:57, komodo said:

Hey now. Just cause Val is, ahem, MIA, doesn't detract any from the plethora of goodies the ole naught point five brought us. I never said I wasn't having fun  I also think Trajectories, while not explicitly lying to me, mislead my reentry corridor a bit... It's fun having the reentry tiger by the tail, honestly.    

I have this feeling both of you know the tipping point well, where you start to garner as much fun poking at your gamedata folder as you do playing the game, I'm right on the edge as well, having a blast (awesomeness modifier 0.5) personally. Thanks to both for your creations!

You're welcome!  I do know that feeling -- moving to 1.0.5 has simplified my mod load some, but at times it definitely felt like I was spending more time managing my 1.0.4 installs than playing them.

[KerbonautInTraining]

I'll have to do more testing but it appears the ISP boost SRB's get causes KW solids to explode before burning out. Without the patch the boosters are designed to come really close to overheating (for whatever reason) and the increase in burn time just manages to tip them over the edge.

Of course, the fix is just a matter of going into the cfg's and taking a 0 out of the "heatproduction" value, so this isn't anything game breaking.

 

Did I mention how awesome this mod is? Because it's pretty awesome.

[Kerbas_ad_astra]

I'm sure the Kerbals prefer to think of it as "automatic hands-free decoupling". 

I'm glad you like SMURFF!

[lork]

So, this mod will work fine for real solar system? will i get payload fractions similar to the ones we get in real life? or should i use 64k?

[KerbonautInTraining]

21 minutes ago, lork said:

So, this mod will work fine for real solar system? will i get payload fractions similar to the ones we get in real life? or should i use 64k?

It's balanced for RSS or Kscale10.

If I understand the OP correctly, using this in RSS will give you mass fractions that are similar to using stock parts in 64k. 

[lork]

33 minutes ago, KerbonautInTraining said:

It's balanced for RSS or Kscale10.

If I understand the OP correctly, using this in RSS will give you mass fractions that are similar to using stock parts in 64k. 

But 6.4x with the stock engines post 1.0 nerf have worse payload fraction than real life right? at least it looked like it when i tried.

[Kerbas_ad_astra]

Here's a "Kerbal X Heavy" that gets 7.5 km/s of delta-V and a payload mass fraction of about 3% (that is, the rocket is 33x bigger than its payload -- in line with the Falcon 9 and such) with stock parts:

 

It's also important to bear in mind that real rockets can get pretty poor payload fractions -- 50, 75, even 100 times more rocket than payload (especially rockets that use solid or monopropellant stages -- e.g. the Athena II and M-V -- and going to exotic low orbits, like sun-synchronous).  Liquid bipropellant rockets should be more in the 25-50x range.

Other design notes on the Kerbal X Heavy: each stage has bout the same delta-V, with the lion's share coming from the higher-efficiency upper stage (which is extended -- a pair of X200-16's / a X200-32, vs the single X200-16 of the original).

[lork]

I have one last question, does this mod works with the latest update of nearfuture propulsion?

[Kerbas_ad_astra]

From a brief check of the new tank mass fractions, it works in the sense that nothing gets sent into negative masses or anything.  That said, there have been changes to the various tank mass fractions.

The xenon tanks have a more stock-like mass fraction, so they still get treated fairly.

The argon tanks are about 30% heavier than they used to be, which is now heavier than is predicted to be achievable.  I suspect that this was done for game balance reasons -- the argon tank mass fraction now matches the xenon tank mass fraction (tank mass = 78% fuel mass = 56% overall mass) -- but in reality, because the argon atom is so much lighter than the xenon atom, it just doesn't pack as densely at similar temperatures and pressures as xenon.  (The reason to consider argon vs. xenon is that it's more abundant and thus cheaper by orders of magnitude, which may offset the need to use more of it depending on the mission.  Not a big deal in the stock solar system, but in RSS...)  I'll include a fix for this at the next release.

Liquid hydrogen tanks have been removed (and weren't being patched by recent versions of SMURFF anyway), and in Nertea's dev thread, there's some discussion going on about rebalancing them, including differentiating short-term lifter tanks and long-term orbital storage tanks via boiloff and mass ratios; we'll see how that goes.

I have no idea about lithium -- presumably it can be stored as a solid quite efficiently, but it needs to be vaporized before being fed to the engines, which is hard enough on Earth and more so in microgravity, which will push up tank mass considerably.  I'll leave those alone until I can get a better idea about lithium engine technology.

[komodo]

It took a minute to realize you were talking of NF Propulsion, then it clicked in  

[NathanKell]

@Kerbas_ad_astra @komodo I just realized I should mention. I don't think ROMini does this, either. Need to rescale any override drag cubes when the parts rescale.

Happily I wrote some code in RO to do this, which you're both free to use ( @komodo you should do it in a script, since that avoids having to pre-specify the number of cubes ). https://github.com/KSP-RO/RealismOverhaul/blob/master/GameData/RealismOverhaul/DragCubeRescaler.cfg

Note relevant to present-SMURFF since you don't rescale, though, right?

[Kerbas_ad_astra]

Nope, and I don't plan to start.  I might be able to get away with rescaling just command pods (filtering by looking for ModuleCommand and a comma in bulkheadProfiles - that is, two different sizes), since I'm not aware of any (stock or mod) which use MODEL nodes or custom drag cubes, but then some of the old radial sizes don't map to proper new sizes (1 and 2 get mapped to 1.5 and 3, which is OK, but 0.5 and 3 get stuck in limbo).  Compatibility is also a concern -- growing the pods is one thing, but I don't know enough about KSP's inner workings to say how everything attached to them will react.

This leaves mass as the other lever to fix ballistic coefficients.  Heat shields are way heavy (1300 kg for 2.5-meter stock shield vs. 850 kg for the 3.9-meter Apollo shield and 272 kg for the 1.9-meter Mercury shield), so there's room to improve there.  Command pods are not as consistent as fuel tanks and engines, which makes me hesitant to tinker there, but a 50% cut across the board will bring capsule ballistic coefficients in between the Apollo/Orion/Gemini values (450-500 kg/m^2) and Soyuz (800 kg/m^2) -- stock parts lead to values of 1100+ kg/m^2, and a 50% cut to pods and shields will reduce that by not-quite-half (since parachutes, docking ports, etc. aren't being patched).  Not all crew parts are balanced equally, but cabins are ~0.4-0.6 tons per Kerbal and command pods are ~0.6-1.3 tons per Kerbal.  The Mercury, Gemini, and Apollo capsules were ~1-1.8 tons per human, and the 747 (minus engines) is about 0.4 tons per passenger, but humans are about twice as large as Kerbals.  Not all components of a life-supporting craft can scale quadratically or cubically with size, so I'm okay with fixing the mass buff to 50%.  (After the deductions for internal monoprop tanks.)

This is all subject to lots more testing on my part.  When I played with RSS in 1.0.4, I had several (single-digit number) successful recoveries from LEO and the Moon and one failure on a lunar-return trajectory (failed to decelerate to a safe chute-deploy speed, if I recall correctly) -- not a whole lot of data (no suborbital flights at all), and I'm not a typical player.  I'm planning to make a release tomorrow with just the argon rebalance for NFT, and then we'll see when I get time to stand up a 1.0.5 RSS install and start lobbing pods around.

[Kerbas_ad_astra]

3 hours ago, Joco223 said:

Is there a mod that limits number of ignitions on engines but its without real fuels? For KSP 1.0.5

There's a recompile of Engine Ignitor for 1.0.2 (look in the back of the thread), but Honeyfox hasn't been around since August, DennyTX didn't really adopt the mod, and who knows how 1.1 will change things.

Since RealFuels is being actively developed and has all of EI's features, it would probably be easier to make a patch that sets up engines with ModuleEnginesRF (with limited ignitions, minimum throttle, and such) but leaves them burning LiquidFuel and Oxidizer.

[Joco223]

1 hour ago, Kerbas_ad_astra said:

There's a recompile of Engine Ignitor for 1.0.2 (look in the back of the thread), but Honeyfox hasn't been around since August, DennyTX didn't really adopt the mod, and who knows how 1.1 will change things.

Since RealFuels is being actively developed and has all of EI's features, it would probably be easier to make a patch that sets up engines with ModuleEnginesRF (with limited ignitions, minimum throttle, and such) but leaves them burning LiquidFuel and Oxidizer.

So just a stripped down version of RF? I'll look into it and see what can i do

-EDIT-

I looked into it and i have no idea how that patch could be made. I think im going to try and install real fuel and just make a cfg for the engines it doesnt support

Edited January 8, 2016 by Joco223

[Kerbas_ad_astra]

Version 1.1.4 is out!

• 2016 01 08 (1.1.4): Argon adaptation
  • New ArgonGas patch for the tank efficiency changes made in Near Future Propulsion 0.6.0.
  • I've made some slight changes in the backend for handling Cryogenic Engines. In my testing, I haven't noticed any changes in spacecraft mass (either in the VAB or in flight), but please be careful!

Edited January 8, 2016 by Kerbas_ad_astra

[Joco223]

Is there a version of SMURFF that is a bit harder than this one? I play with KSP SKY and for stock i need to build too big rockets for larger missions but with SMURFF is just way too easy. Is there any mod to fix this?

[Kerbas_ad_astra]

Not at this time.  You could tinker with the constants in SMURFF.cfg so that parts aren't improved as much, but that isn't very user-friendly at all.

You're the second person to ask about intermediate buffs, and after some more testing, I think there's room for adjustment.  With SMURFF, rockets need to be 12-15x as massive as their payload to get to 7 km/s, while stock rockets need to be ~25x as big -- still better than real rockets, but that's a sizeable gap in between.  (Earlier, testing for 7.5 km/s for 64k, I found the gap to be narrower, 15-20x for SMURFF and 30x for stock, but maybe I've gotten better at designing rockets since then. So @komodo, I think I was wrong earlier -- you'll get the opportunity for an intermediate buff.)

Rather than release a spectrum of separate configs, what I'll do is reorganize them so that there's a couple of "master control" levers that govern buffs applied to fuel tanks, engines, and (maybe) crew cabins.  I'll set them to 1 for current adjustment levels, and people interested in playing on intermediate solar systems can set them to 0.5 or so.  Three cheers for the power of Module Manager!

Edit: and this will make testing the pod balance easier for me as well -- I can just set "podlever" to 0 or 1 depending on what I'm testing, and then there will be no need for commenting and uncommenting that code block.

Edited January 9, 2016 by Kerbas_ad_astra

[Joco223]

1 hour ago, Kerbas_ad_astra said:

Not at this time.  You could tinker with the constants in SMURFF.cfg so that parts aren't improved as much, but that isn't very user-friendly at all.

You're the second person to ask about intermediate buffs, and after some more testing, I think there's room for adjustment.  With SMURFF, rockets need to be 12-15x as massive as their payload to get to 7 km/s, while stock rockets need to be ~25x as big -- still better than real rockets, but that's a sizeable gap in between.  (Earlier, testing for 7.5 km/s for 64k, I found the gap to be narrower, 15-20x for SMURFF and 30x for stock, but maybe I've gotten better at designing rockets since then. So @komodo, I think I was wrong earlier -- you'll get the opportunity for an intermediate buff.)

Rather than release a spectrum of separate configs, what I'll do is reorganize them so that there's a couple of "master control" levers that govern buffs applied to fuel tanks, engines, and (maybe) crew cabins.  I'll set them to 1 for current adjustment levels, and people interested in playing on intermediate solar systems can set them to 0.5 or so.  Three cheers for the power of Module Manager!

Edit: and this will make testing the pod balance easier for me as well -- I can just set "podlever" to 0 or 1 depending on what I'm testing, and then there will be no need for commenting and uncommenting that code block.

Nice! I'll look into the .cfg and see what can i do. Right now you only need about 6.5km/s of DeltaV to an orbit and SMURFF makes it a bit to easy.

[Kerbas_ad_astra]

31 minutes ago, Joco223 said:

Nice! I'll look into the .cfg and see what can i do. Right now you only need about 6.5km/s of DeltaV to an orbit and SMURFF makes it a bit to easy.

No, wait!  It's too late for me, but you can still save yourself -- if you gaze into the config, the config will gaze into you...

Ia, ia, Cthulhu ftaCFGn!

...

I've actually just worked it out. It wasn't too hard, since most of the math is just repeated for various fuel types once I've worked it out the first time -- I just had to work on shoehorning the math into the limitations of Module Manager (break things down one step at a time, chase out the typos...).  Check it out on GitHub if you dare.

[Joco223]

5 minutes ago, Kerbas_ad_astra said:

No, wait!  It's too late for me, but you can still save yourself -- if you gaze into the config, the config will gaze into you...

Ia, ia, Cthulhu ftaCFGn!

...

I've actually just worked it out. It wasn't too hard, since most of the math is just repeated for various fuel types once I've worked it out the first time -- I just had to work on shoehorning the math into the limitations of Module Manager (break things down one step at a time, chase out the typos...).  Check it out on GitHub if you dare.

I just changed the @maxThrust and @mass and it worked without problems

[Kerbas_ad_astra]

What, of just the engine patches?  Was that what made it too easy to make your rockets work?

[Joco223]

18 minutes ago, Kerbas_ad_astra said:

What, of just the engine patches?  Was that what made it too easy to make your rockets work?

I just edited those too thing in the SMURFF.cfg and lowered the power of engines to a bit stronger than stock and increased the mass to 0.5 from .375. I thiink this is the perfect balance for my taste because i play KSP on a laptop that is not that stong so i don't want to play KSP at 2-3fps

-EDIT-

Now with edited .cfg my unmanned duna lander has about 15.9km/s of Delta V on the launch pad and wheighs about 306t and 79 parts(I use procedural parts)

Edited January 9, 2016 by Joco223

[Kerbas_ad_astra]

I've done some suborbital testing (boosting to a Mercury-like 190 km, lobbed out over the ocean) with the Mk1, the Mk3-9 and PPD-1 (3.75m) from Near Future Spacecraft, the 1.875m Soy-Juice descent module from HGR, and the 4K pod from Lithobrake Exploration Technologies (2.5m, even heavier than the Mk1-2 at 5 tons) -- plus the appropriately-sized heat shield, parachute, and some RCS thrusters.  With stock pod and heat shield masses, pods generally slow down to a safe speed (~260 m/s) at around 2 km above sea level. This makes for a harrowing reentry over water, requiring Jeb-like reflexes and attention to the speedometer to open the parachute right when the speed is low enough, and will likely make overland landings infeasible.  The 4K couldn't even manage that -- it plunged into the water at over 300 m/s.  On the other hand, with pod and heat shield masses halved, pods of various sizes were slow enough between 4 and 6 km (except the 4K, but even that one made it at 2500 m when it outright crashed before).

(This is what drogues are for, but the player doesn't get them until Advanced Landing at level 6 of the tech tree, so they really shouldn't be assumed to be available, especially for early suborbital flights.)

Next up is orbital reentry testing, both from low orbit and high-energy return (e.g. lunar).  Based on the suborbital testing and comparisons I made above, I'm determined to cut the masses; I just need to verify that the heat shields still do their jobs when they're put to the test.

Also coming down the pipeline: the patches are going to be FOR[zSMURFF], since SpaceY's Cryogenic Engines patches were running FOR[SpaceY-Expanded], overwriting the thrust buff when they deleted and recreated the engine modules (SP comes after SM).