Realism Overhaul

[AbeS]

Yes, pressing '.' and then ',' repetitively causes stuff to nudge to the East!

[MedievalNerd]

Yes, pressing '.' and then ',' repetitively causes stuff to nudge to the East!

Strangely enough, even stock KSP does this as well (the weird re positioning), but not enough as to make you move. But I've been able to screw up my launch clamps by doing so.

[NathanKell]

Oh! Of course! You said east.

I wonder if it's to do with rotation? And the rescale has a higher linear velocity, despite lower angular velocity.

[AbeS]

So back to the realism overhaul, how hard would it be to scale all the rockets? Would we (you guys actually) just need to scale everything from 0.64 to 1.00? Or scale stuff based on real life parts? Or both?

For example Stock parts would go from 0.64 to 1.00 and FASA parts would go from their actual size to the real ones (like Nathan did).

EDIT: Also the masses would vary and the engines stats too, that makes it more complicated I think?

Edited October 30, 2013 by AbeS

[NathanKell]

I wouldn't touch the existing tank and engine sizes; there's still a need for .625, 1.25, and 2.5m rocket cores. Addons give us 3.75m and 5m. All we need now are the sizes above that.

Probe parts--MedievalNerd was working on a pack of smaller probe parts, for early probes. (Explorer 1? 13kg.) That should work well for this mod.

Squad's crewed capsules, however, should be scaled by 1/.64, yes. Though the masses can stay as they are. If you add the recovery equipment and heatshield mass in each case, they come out about right.

But that leads us to the question of what our standard sizes should be. Should we stick with the existing 1.25m per step, and have the crewed parts be odd man out? Should we go to steps of 1m, to better fit US capsule and rocket diameter? (Mk1 rescaled by 1/.64 would have a 1m top and 2m bottom, and the Mk1-2 would have a 2m top and 4m bottom, and Titan II has a 3m core). Of course S-IVB (and the Saturn IB) had 6.6m diameter cores, so that doesn't fit so well. Back to fitting for Saturn V, with a 10m core.

I don't know Russian cores off the top of my head, I'm afraid.

[AbeS]

I am no expert but by reading a bit about the R-7, Proton, Zenit and Energia families, the diameters are almost the same as the ones you mentioned:

R-7 has 2.95 m

Proton 4.1 m

Zenit 3.9 m

Energia 3.9 m (strap-on boosters), 7.75 m (core stage)

[NathanKell]

Sounds like it might be worth going to steps of 1m then. Or maybe steps of 2m above 4m.

How about: steps of 1m (size 0 = 0.5m) until 4m

Then steps of 2m until 10m

That means we only need 6m, 8m, and 10m diameter tanks/engines/decouplers. That can be done with rescales.

What we really need are more realistic engine models.

[AbeS]

I think those diameters sound good! And a dumb question, what do you mean with the more realistic engine models?

[NathanKell]

We have vacuum engines with tiny, wrongly-curved nozzles, lower-stage engines with nozzles that...I just can't even, and engines that are all nozzle and no turbopump. Squad's (and, following suit, modders') engines' nozzles usually bear no relation to the role of the engine or it's performance. Consider for example Apollo's Service Propulsion System. It's nozzle is gigantic in KSP terms--2.5m diameter or so, ~3m length. But the throat is miniscule--as small as the throat on an LV-T45 at most, if not rather smaller. And it delivers ~70kN thrust. Vacuum-rated nozzles should be very large, very light, and have a very high area ratio (exit:throat). Sea-level-optimized nozzles should have low area ratios (which almost all of Squad's nozzles do, fine) but they should also be rather longer. Only the LV-T30 and T45 even approach the right aspect ratio.

This is _not_ my area of expertise. Just what I've picked up. IIRC regex, rhoark, asmi, and others know rather more.

[AbeS]

I get it! I thought you meant the fact that thrust and isp are independent or something like that, but thinking about it that also should be related to the nozzle like in real life!

[Starwaster]

I get it! I thought you meant the fact that thrust and isp are independent or something like that, but thinking about it that also should be related to the nozzle like in real life!

I suppose, but sooner or later suspension of disbelief has to come into play as well.

[Camacha]

Squad's crewed capsules, however, should be scaled by 1/.64, yes. Though the masses can stay as they are. If you add the recovery equipment and heatshield mass in each case, they come out about right.

But that leads us to the question of what our standard sizes should be. Should we stick with the existing 1.25m per step, and have the crewed parts be odd man out? Should we go to steps of 1m, to better fit US capsule and rocket diameter?

.

I still do not understand why you would go through all that trouble to end up with not proportionally scaled craft, especially if those are not proportional amongst each other. First we were playing with scale models of planets, with the planets fixed we will be flying with scale models of rockets.

[SFJackBauer]

Although it is good to reuse existing models from a practical standpoint, in the long-term, isn't it better to develop an independent pack, with real-life modular parts possessing the correct size, mass, and fuel amount?

For example, building a Delta IV could be broken down into assemblying:

- RS-68

- 5m x 25m tank with 40t of LH2

- 5m x 10m tank with 160t of LOX

- 5m interstage with decoupler

- RL-10

- 5m x 6m tank with 22t of LOX

- 5m x 2m tank with 5t of LH2

- 5m payload and fairing adapter

Those parts could be mixed and reused to create other realistic rocket models. The shuttle external tank for example, its about the same length of the Delta IV first stage, only exchanging the round cylindrical LH2 tank for a conical one, and a bottom round seal. Also the Ares or SLS could be built from it since they had/will have cryogenic cores. The solid boosters could be built from individual segments.

The same philosophy could be carried through everything else. What I wonder is what is more demanding - keeping track of all the released rocketry packs, tweaking their mass/fuel values as Modular Fuels does now, and also resizing them as it is being proposed, or develop pre-built modular parts from scratch?

[p3asant]

Although it is good to reuse existing models from a practical standpoint, in the long-term, isn't it better to develop an independent pack, with real-life modular parts possessing the correct size, mass, and fuel amount?

For example, building a Delta IV could be broken down into assemblying:

- RS-68

- 5m x 25m tank with 40t of LH2

- 5m x 10m tank with 160t of LOX

- 5m interstage with decoupler

- RL-10

- 5m x 6m tank with 22t of LOX

- 5m x 2m tank with 5t of LH2

- 5m payload and fairing adapter

Those parts could be mixed and reused to create other realistic rocket models. The shuttle external tank for example, its about the same length of the Delta IV first stage, only exchanging the round cylindrical LH2 tank for a conical one, and a bottom round seal. Also the Ares or SLS could be built from it since they had/will have cryogenic cores. The solid boosters could be built from individual segments.

The same philosophy could be carried through everything else. What I wonder is what is more demanding - keeping track of all the released rocketry packs, tweaking their mass/fuel values as Modular Fuels does now, and also resizing them as it is being proposed, or develop pre-built modular parts from scratch?

I don't see any real need for editing/adding bigger tanks. Using stretchytanks is pretty easy solution and adapters can be re-scaled ( e.g i have the novapunch 3.75m-5m rescaled for 5-6.5 and 6.5-11.25 or so).

Only problem is the lack of big (and small) realistic engines.

Also for upper stage engines adding couple versions of the KwRocketry service engine in different sizes would probably work because it looks like a real vacuum engine.

[SFJackBauer]

I don't see any real need for editing/adding bigger tanks. Using stretchytanks is pretty easy solution and adapters can be re-scaled ( e.g i have the novapunch 3.75m-5m rescaled for 5-6.5 and 6.5-11.25 or so).

Only problem is the lack of big (and small) realistic engines.

Also for upper stage engines adding couple versions of the KwRocketry service engine in different sizes would probably work because it looks like a real vacuum engine.

The service engine looks like a real Apollo SM vacuum engine.

See that's the problem. With resizing you end up with a bunch of things of different sizes that look the same, losing identity. When in real-life they will probably look completely different. For example the KW Wildcat XR makes for a good J-2/J-2X model - a high-thrust, high ISP vacuum engine. You can't just downsize it and pretend it is its smaller counterpart, the RL-10 - well you can, but that defeats the purpose of having different models (I'd say the KW equivalent for the RL-10 would be the Vesta VR9D for the dual version and Wildcat V for the single).

The same about tanks - the NP 5m tanks are kerolox, Saturn-V style. You see KW actually included two variants of their largest tanks - a white and a orange one - because the orange large first stage is always correlated to a cryogenic stage like in Delta IV or the shuttle, due to the insulation. I know you can't judge what is inside the tank from the appearance only, but it counts.

And lastly, we could give things their real names and their real stats directly into the part configs. This way you could quickly see - ah I want to use an F-1 engine, then I need a kerosene/lox tank - without the plethora of configurations and overrides present in MFS, with different tech levels, different modes for some engines, and different fuel for tanks. IRL you wouldn't take a Delta IV first stage and fill it with kerosene - you would design a completely new tank.

[pina_coladas]

Just change the tank texture to something orange on the fly when using LH2! Stretchy tanks allows that kind of texture swapping already. It's too bad that the current version of stretchy tanks is so finnicky to use with modular fuels though... Is a fork possible?

[Starwaster]

Just change the tank texture to something orange on the fly when using LH2! Stretchy tanks allows that kind of texture swapping already. It's too bad that the current version of stretchy tanks is so finnicky to use with modular fuels though... Is a fork possible?

What's finicky about it?

You know there's a patch for it right? Originally mine but Nathan took it and incorporated a metric **** ton of other features and you should totally go get it.

Between us we fixed a bunch of stuff.

[pina_coladas]

I'm using Nathan's patch. My experience has been that it does make stretchy tanks work with MFS, but the MFS dry mass and fuel density values are sort of tacked-on and constantly being overwritten by the default stretchy tanks values. So every time I re-size a tank its values are recalculated with the "stock" ratios. That's not too terrible, but the MFS values also seem to get overwritten whenever I grab and move a tank, and whenever I load a craft in the VAB. So every time I revert to the VAB or load the craft in the VAB I have to go through each tank in the MFS menu to get the values right and then try not to touch anything before I hit launch.

Am I an idiot who installed something wrong? I assumed the lack of support from the stretchy tanks author made it difficult to add MFS compatibility and it was this way for everyone. It really takes the joy and wonder out of making Saturn V replicas unfortunately.

[NathanKell]

pina_coladas, try the latest patch in my signature, and MFS v3 full (both just released). I believe it is finally working correctly now. It used to only half work (sometimes giving the symptoms you describe), that's why I recently fixed it.

I'm currently trying to get the colliders to play nice with ferram's KJR.

[NathanKell]

I still do not understand why you would go through all that trouble to end up with not proportionally scaled craft, especially if those are not proportional amongst each other. First we were playing with scale models of planets, with the planets fixed we will be flying with scale models of rockets.

Um...since feram released KJR, we're plannng on doing full-scale rockets.

(Although even if we _were_ playing with scale rockets, that wouldn't matter as long as the ballistic coefficients matched, as ferram and I were discussing above; the point of rescaling planets is that it changes the deltaV [and therefore DRASTICALLY changes the mass ratio required], but playing with 64% scale rockets just means they suffer a few hundred m/s more drag if you play with scaled masses, or 50 m/s less drag if you play with real masses. Unless you fix the ballistic coefficients, in which case it's no different at all than 1:1 scale. Velocities and forces don't scale with the rocket size, and a 100m rocket vs a 64m rocket, well, hard to tell the difference vs. a 13000m diameter planet)

SFJackBauer, since there already are released and playable Delta II, Delta IV, and Atlas V mods available, if you want to give them realistic size, mass, and other stats that would be awesome.

I do not necessarily support using real engines and tanks generally, however, as it gives us much more flexibility to have just a few different engines per type per size, and let MFSC rescale their tech levels. Same with tanks; with StretchyTanks we can clone any existing tank size, and with an update I'll soon add to MFSC we can use the same tech levels system to change their dry mass fraction to be tech-appropriate.

The real issue is--as I've found with MFSC, whose new spreadsheet supports over a hundred engines and solid motors, that there _still_ is far from enough variety, and I've had to make do horribly. For instance, there are like _five_ engines that have decent vacuum nozzles. And we have to pretend they're open-cycle, closed-cycle, _and_ pressure-fed. We just don't have the models.

That said, there _are_ a number of models that are clones of real tanks/engines/etc., and it's worth having fully-real versions to use. I did that with Gemini, for instance, and it wouldn't be hard to do for the 1:1 F-1 engine, or the 1-1 S-IC somebody released, or the NK-33 (called RD-33) in the KOSMOS pack (if cBBp et al let us).

Finally, regarding the orange foam. That was a comparatively recent invention, a new kind of insulation for cryo tanks. You'll note that the S-II and S-IVB on Saturn V (and the S-IVB on Saturn IB and the S-IV on Saturn I) all did not have it. And that some rockets that only have cryo LOX still have that foam (it's better cryo insulation, and LOX itself is cryo, though less so than LH2!).

I do definitely support its existence as a tank paintscheme choice for ST and others, though!

[asmi]

That said, there _are_ a number of models that are clones of real tanks/engines/etc., and it's worth having fully-real versions to use. I did that with Gemini, for instance, and it wouldn't be hard to do for the 1:1 F-1 engine, or the 1-1 S-IC somebody released, or the NK-33 (called RD-33) in the KOSMOS pack (if cBBp et al let us).

I'm gonna try to convince BobCat to make RD-170 engine, because I just love it, and he is the best man for the job

Finally, regarding the orange foam. That was a comparatively recent invention, a new kind of insulation for cryo tanks. You'll note that the S-II and S-IVB on Saturn V (and the S-IVB on Saturn IB and the S-IV on Saturn I) all did not have it. And that some rockets that only have cryo LOX still have that foam (it's better cryo insulation, and LOX itself is cryo, though less so than LH2!).

I do definitely support its existence as a tank paintscheme choice for ST and others, though!

Shuttle's ET needed insulation because it was flying in the airstream (and as such was subjected to airloads and heating), while inline HydroLOx stages are in the aerodynamic shadow of the PLF, and as such are not heated by airflow (remember that in supersonic and hypersonic flight regime airstream behind the shockwave is almost nonexistent). Look at Centaur for example - although it uses HydroLOx, there is no insulation. DIV has insulation since its' CCB is supposed to be common with DIVH, which again has its' boosters flying in the airstream.

Edited October 31, 2013 by asmi

[NathanKell]

Ah, thanks for that!

Heh, and thanks for poking BobCat too.

I gather the orange on the Atlas's Centaur is just paint? The orange on the CCB is presumably for the same CCB reason you cite, unless again it's just paint because LOX doesn't need the foam?

[Camacha]

and a 100m rocket vs a 64m rocket, well, hard to tell the difference vs. a 13000m diameter planet)

I kind of disagree there - as explained before- since a planet that is ~156% of what it visually should be at a certain height will show. That much difference is quite pronounced. However, as you wonderful lads intend to do full scale rockets, this will not be any issue at all.

[Camacha sits back down quite happily]

Edited November 1, 2013 by Camacha

[Guest]

I'm gonna try to convince BobCat to make RD-170 engine, because I just love it, and he is the best man for the job

CBBP made such an engine once, along with RD-180 and RD-191. I think I might still have them around, but they're very old and might not work properly. They still look great, though.

[philonius]

Not sure if this has been discussed in this thread before, but I just rewatched BBC's 'Space Race' (which I highly recommend, awesome miniseries) , and one of the things that strikes me was just how experimental these rocket designs were, in other words, how often an unforeseen design flaw / manufacturing error caused the rocket to explode.

Has there been any thought about incorporating failure probabilities in the realism overhaul? Fuel line ruptures, stuck fairings, 'Houston, we've had a problem here' ? Perhaps a per-part failure chance, which can be influenced by techlevel and experience with flying the part. First time trying a new rocket? Expect debris.

Perhaps a mod could be made that uses subassemblies for this purpose: When you design a booster, save it as subassembly. Then the mod tracks how often the subassembly has flown, with failure chance decreasing with flight rate. A new rocket's failure chance would be a function of tech level and how new the components are. A rocket comprising of an proven design with a modified upper stage would be more reliable than a new rocket.

If this can't be done automatically, it might be done manually. Upon launch, a dialog window could ask the player to identify the rocket's stages, and the failure chance is then set. Perhaps the bravery/stupidity of the crew (if any) might modify this setting.

Not sure if this would significantly add to the fun in the end, but it would introduce some choices to the gameplay. Do I chance the lives of my brave kerbals on an unproven rocket, or do I do some test flights first? Obviously, this will be a more pressing choice when the money system goes live. Also, it provides a use for launch escape systems.

Edited November 1, 2013 by philonius