[WIP][1.8.x] SSTULabs - Low Part Count Solutions (Orbiters, Landers, Lifters) - Dev Thread [11-18-18]

[blowfish]

Yeah, that's what I was saying, but are you going to fix that issue with the ICPS?

The fairing is part of the service module. I don't think it has anything to do with the ICPS.

[davidy12]

Yeah I was referring to the fairing between the ICPS and the SM.

[blowfish]

Yeah I was referring to the fairing between the ICPS and the SM.

It's part of the SM.

[RedParadize]

Yeah I was referring to the fairing between the ICPS and the SM.

It's part of the SM.

Blowfish is right, it is in the SM. If you don't want to wait, you can try to fix it yourself. You will see, its not hard at all.

In my version, I downscaled the Orion and its servise module to 0.8 and had to adjust fairing, it took me 5 min to do.

[A1Ch1]

Okay, so for those who were having problems with the radial decouplers...

Kerbal Joint Reinforcement doesn't recognize SSTUCustomRadialDecoupler as something to reinforce, it looks for ModuleAnchoredDecoupler. Adding that module fixed the wobbly booster problem when I tried it (using RSS/RO).

[Shadowmage]

Kerbal Joint Reinforcement doesn't recognize SSTUCustomRadialDecoupler as something to reinforce, it looks for ModuleAnchoredDecoupler. Adding that module fixed the wobbly booster problem when I tried it (using RSS/RO).

Actually, according to KJR documentation, that specific stiffening type is disabled...

"

These types are currently not used, but removing the a in front of them will cause KJR to make use of them again; their lack should not affect stiffening appreciably but does help reduce overhead and strange stiffening situations

string adecouplerStiffeningExtensionType5 ModuleDecouple --Decoupler stiffening will look for parts beyond this part type to add to stiffening

string adecouplerStiffeningExtensionType6 ModuleAnchoredDecoupler --Decoupler stiffening will look for parts beyond this part type to add to stiffening

string adecouplerStiffeningExtensionType7 ProceduralFairingBase --Decoupler stiffening will look for parts beyond this part type to add to stiffening

---

"

So, that fact that adding the module does -anything- for you is strange in itself. It shouldn't.

You also shouldn't be adding the 'ModuleAnchoredDecoupler' module, as the SSTUCustomRadialDecoupler -is- a ModuleAnchoredDecoupler (it is a directly derived class); as such, it will already be recognized anywhere that something is looking for the ModuleAnchoredDecoupler (any Unity calls to GetComponent<ModuleAnchoredDecoupler> will return SSTUCustomRaidialDecoupler as well). And.. I posted pics yesterday of KJR working just fine with the existing module, so I know that is not the problem.

[A1Ch1]

Well that's odd then. I was working on the RO config and the boosters were very wobbly, so I added an extra decoupler module thinking KJR was looking for that and the joint became rigid. Removing the module resulted in springy SRMs again

https://github.com/A1Ch1/RealismOverhaul/commit/c34ad30e84edc774f471b3b243cbbee3f4ab945c

[blowfish]

Well that's odd then. I was working on the RO config and the boosters were very wobbly, so I added an extra decoupler module thinking KJR was looking for that and the joint became rigid. Removing the module resulted in springy SRMs again

https://github.com/A1Ch1/RealismOverhaul/commit/c34ad30e84edc774f471b3b243cbbee3f4ab945c

The proper way to do it would probably be to add it to KJR's data, in this file. Pretty simple to add and create a PR.

[A1Ch1]

The proper way to do it would probably be to add it to KJR's data, in this file. Pretty simple to add and create a PR.

It turned out not to be the issue since KJR ignores ModuleAnchoredDecoupler, as Shadowmage pointed out.

[blowfish]

It turned out not to be the issue since KJR ignores ModuleAnchoredDecoupler, as Shadowmage pointed out.

Ah, you're right. Interesting.

[Shadowmage]

Just going to leave this here for a bit... not quite done, but I'm really liking the results so far, so I thought I would share

I did notice yesterday during some unrelated testing that if I decoupled the upper-stage of my test craft while still on the launch-pad (so the core+srbs were 'debris'), that KJR would stop stiffening the SRB joints almost immediately, and they would revert to stock behavior. Though I could not find any way to make that happen on the functional craft.

So there might be something else at play related to craft configuration, KRJ code, or even mod loading/initialization sequence. Basically I'll need someone to provide a minimal test case that can replicate the problem before I can begin looking into it, as I have been unable to reproduce the issue in a live craft on my install. By minimal I mean basically a Stock + SSTU + KJR setup, with a test craft file that I can try loading.

[blowfish]

Wow ... amazing job on that F-1 You could probably spare a few more polys on that injection manifold though ... a few more triangles on the render mesh isn't going to have any measurable performance impact.

I suspect that KJR does not apply stiffening to debris craft. FAR makes similar simplifications in the name of performance, so I'd be willing to bet ferram4 did something like that in KJR. Maybe I'll ask the next time I see him on IRC.

[Shadowmage]

Hmm.. yes I think you are right... more poly's were in order for the manifold. Looks much better

[VenomousRequiem]

Oh lord I'm so excited... I won't need any other lifters packs whenever this is finished.

By the way, I'm sorry I can't be more help!

[CobaltWolf]

Oh lord I'm so excited... I won't need any other lifters packs whenever this is finished.

By the way, I'm sorry I can't be more help!

0_0 Traitor

[riocrokite]

these engines are amazing, awesome stuff \o/

edit:

maybe I've missed it but are engine shrouds separate parts? If so I wonder how they behave (wobbliness-wise) when they are stuffed between super powerful engines below and super heavy stuff above (taking into account their relatively low mass).

Edited November 4, 2015 by riocrokite

[Shadowmage]

Oh lord I'm so excited... I won't need any other lifters packs whenever this is finished.

By the way, I'm sorry I can't be more help!

Should start having enough stuff to be quite usable on its own here very shortly. Will be adding a 5m and 6.25m rescaled stock fairing for this next release... so at least you can make fairings for the larger tank sizes. Still just the stock fairing, but I'll get the custom parts w/integrated decouplers setup here before too long.

0_0 Traitor

Hehe, to be fair, she is just being a 'little bit' of a traitor, and popping in to say hi from time to time

Nice work on your parts BTW. You've certainly managed to do a bit more 'stock-alike' textures than I have aimed for, and your parts look like they would work quite well with stock

these engines are amazing, awesome stuff \o/

edit:

maybe I've missed it but are engine shrouds separate parts? If so I wonder how they behave (wobbliness-wise) when they are stuffed between super powerful engines below and super heavy stuff above (taking into account their relatively low mass).

The shrouds (i.e. the mounts with integrated engine fairings, e.g. the SLS engine mount) are all integrated into the engine cluster part. So, single part, no joints to worry about

And they work very well so far. Tested the F1 cluster/Saturn-V mount last night, and it had no problems regarding stability (needs some better effects though; not nearly impressive enough, and balance/stats are still WIP).

Regaring engine balance....

I'm trying to come up with the initial set of functions/algorithsm/derivative for deriving KSP stats for the engines from their real world stats, at least for the basis of their stats (can always manually adjust the final stats). Seems like there are two/three things to balance: ISP, Thrust, and Mass (actually, raw TWR; mass will derive from a specific aimed TWR range).

ISP - Obviously I cannot use the real-world ISP values for many engines, as they fall outside of the range of stock ISP values (mostly higher). What I would propose would be to basically rescale the ISPs based on a custom-derived curve with endpoints set by the stock ISP range. For example, real-world vacuum ISP ranges from ~310 (f1, hypergolic fuels), to ~465 (rl-10b, lh2). KSP liquid fueled ISPs range from ~ 250 (puff mono engine) to 350 (poodle). Basically I would 'normalize' the KSP ISPs for the engine given the existing ISP range; e.g. the F1 is one of the lowest ISP real-world engines, and as such would be at the very bottom of the KSP ISP scale, @ 250 (give or take). Whereas the SSME/RS-25 is near the top of the range (455 vac), so it would be mapped to the top of the KSP-ISP range (345-350). Would do the same for the the sea-level ISP. This seems like a fairly comprehensive and decently-balanced method for 'balancing' the engines ISP vs. stock. Open to suggestions/improvments/tweaks though.

Now, the vacuum thrust and mass are actually the tricker/touchier subject (SLT will derive from vac thrust and ISP delta). What would you feel is an acceptable scalar value to apply for the thrust from the real world values? It seems that stock is aiming for a 1/3'ish ratio (based on screenshots of their RS-25, with ~700kn of thrust; about 1/3 the value of the real-world ~2200). However, I also have no idea what scale they have done their engine in. Using a 1/3 value would also result in some fairly wimpy engines. Stranger yet is their KR-2L engine with its 2000 thrust, and incredible ISP. Some 'strange' balance stuff going on in stock, to be sure.

Also, when I was looking over my 'real world engine stats' sheet, I realized that I had the SC-B-ENG clusters with really close to the real-world thrust for the RS-25 (I had 2000, vs the ~2200 of the real one) (was the only way to get it to lift the kinds of loads it needed to). So... I may leave the existing parts as-is, but the new RS-25 engine/clusters will have much lower thrust, using whatever method/function we decide upon. The lack of LH2 as a fuel further complicates this balancing; in order to be able to lift a near-proper-scaled KSP-SLS (using LFO), these engine clusters need gobs of thrust, due to the density of the fuel used (and they need gobs of fuel due to poor ISP).

Hmm... quite stuck on the thrust bit at the moment. Perhaps I'll scale the vac thrust as a deriviative of the raw isp-scaled ISP delta (e.g. if new isp==75% of old ISP, new thrust ==75% of old thrust?). Though, this seems a bit of a naive/heavy-handed/blunt approach. Yep, still stuck. Most of the problem is that I don't want to 'nerf' the SC-B lifting capacity too much with the new engine clusters. Though, I suppose it was a bit absurd lifting 300t with that design, when the real world one will struggle for ~150. So, it might be inevitable that it will have a reduced upper-range capacity; though this might even make it a more viable launcher for the CSM stack as well (currently it is OP/overkill to use just for the CSM). Some of the other new engines might also offer new opportunities to gain back some of this capacity; such as using the F1b liquid-boosters.

Thoughts? Ideas? Suggestions?

To even further complicate things, after I decide upon the thrust of the engine, I need to decide what TWR range that I should aim for. Real world TWRs go upwards of 150. KSP stock TWRs' max out at 27 (for the mammoth engine cluster). To come anywhere near close to this poor of a TWR ratio, every single one of these engines will end up weighing quite a bit more than the real-world engine does even though the engine is scaled down/visually smaller. Yes, those Kerbals must be building out of some sort of super-dense steel-unobtanium alloy. For example; the real SSME only weights ~2.5t, and puts out 2200kn of thrust (~88 TWR). The most powerful stock engine weighs 15t and only puts out 4000kn of thrust (~26.66 TWR); given the same amount of thrust (~2000), a KSP engine would still weigh more than 3 times the real article (7.5t for 2000kn, 26.66 TWR) .

And... one further layer of complication on the balancing -- I need to differentiate many engines according to where they will be on the tech tree (or place them on the tree based on their stats, one way or the other). So, will need a few engines that are pretty much obsoleted by mid-game, and a few mid-game engines that get upgraded/obsoleted by end-game. Ideally I would find a good niche for each of these engines where they are not fully obsoleted (or use some sort of usage-based incremental improvements as was suggested previously).

Edit:

Pretty much finished F1 texture:

I'm certainly getting faster (though perhaps not any better) at doing these textures. Got that one done entirely yesterday; including the complete UV unwrap and layout. Granted, it did take most of the day (likely 8-12 hours for UV/layout/texturing), but that is tons faster than I've done in the past (2-3 days to UV/layout, another 2-3 to do the texturing). Have learned a few new tricks in the past week or two that is making things... so much faster and less painful - like how to auto-generate a good UV-seam-mask from the UV layout, and apply that to noise/etc textures, pretty much eliminating visible seams, all while allowing for even more crazy (or organic) unwrapping to be used which itself speeds up the unwrapping process.

Edited November 4, 2015 by Shadowmage

[davidy12]

First of all great to hear on the engine fairings/mounts.

Second of all, VenomousRequiem is a GIRL!?

VR: Just to let you know, I'm not criticizing you or saying only guys play KSP, just a little surprised.

PS: Congrats, your the second 3d modeler I know (first being my cousin's GF).

- - - Updated - - -

Also, on the subject of Stockalike. I'd say this is the limit for Stockalike parts (in terms of textures, IVA, etc).

[VenomousRequiem]

Second of all, VenomousRequiem is a GIRL!?

Heh... silly nerds.

As in I'm the 2nd 3D modeler you know of that's a girl? Or just... in general?

Either way that's hardly an achievement.

Edited November 4, 2015 by VenomousRequiem

[davidy12]

I know that there are more 3d modelers that are girls. But the only one who I work with are my cousin's GF. So your the second in my life who I've met.

[VenomousRequiem]

I know that there are more 3d modelers that are girls. But the only one who I work with are my cousin's GF. So your the second in my life who I've met.

We've hardly met, eh? Maybe this is a conversation for a less public place, feel free to send me a request on steam. ;b

[davidy12]

Yeah, sorry. Anyways, one more thing.

What's next after the F1?

[VenomousRequiem]

Yeah, sorry. Anyways, one more thing.

What's next after the F1?

Hopefully the RS-68?

[riocrokite]

the texturing looks amazing Shadow, I'll have to ask you for a quick tips'n'tricks tutorial one day

as for ISP/thrust/mass stuff - personally I liked a kind of simple approach of KW rocketry mod. So the higher thrust, the more mass and then lower isp. Most of engines were quite balanced. Also you might consider adding delay to thrust change; I also liked this thing in KW rocketry - the biggest engines had helluva delay so they couldn't be used efficiently for vac operations, also the delay meant that they used a bit of fuel before getting to 100% thrust and lift off thus further reducing their effective isp.

I wouldn't take KR-2L, or the biggest engine as isp point of reference; for me their isp is weird and they are too op; so I would expect that squad might nerf them sooner or later. You might take a look at awesome but not updated anymore rebalance of engines by stupid_chris:

http://forum.kerbalspaceprogram.com/threads/75272-0-24-x-Stock-ReBalance-v1-4-11-09-14

I used this mod extensively when I used to play KSP (hehe)

I think that he basically downgraded isp of largest engines and upgraded smaller engines so they all followed the rule - bigger engine - higher thrust - less isp. In this way every engine had its purpose.

Another (maybe future), more complicated approach would be to divide engines into RP-1/LOX and LH2/LOX (and CH4/LOX?). Then introduce new fuel tanks for that engines (LH2 being much bigger/heavier for the same amount mass of fuel etc), or add another fuel config (LH2/LH2+Ox)for your existing tanks. You can see what Nertea did in his LH2/LOX cryogenic engines mod; from what I remember at least for vacuum engines, the higher isps were offset by higher tank mass so in the end vac engines were maybe 5-7% better dV-wise than their counterparts. I think that for SL engines LH2 and LF engines have similar performance so I think personally I wouldn't bother using them since it meant higher, more unstable rockets (with bigger 1st stage) with the same performance.

[JoseEduardo]

Oh lord I'm so excited... I won't need any other lifters packs whenever this is finished.

By the way, I'm sorry I can't be more help!

[2]

KW will only remain for some structural bits, which might not last because Shadowmage said he probably would look into that in the future, and FASA probably just for Apollo CM/LEM and Titan if Shadowmage doesn't decide to make the UA120X SRBs and LR87 and 91 engines

Heh... silly nerds.

As in I'm the 2nd 3D modeler you know of that's a girl? Or just... in general?

Either way that's hardly an achievement.

I wouldn't say silly nerds, Kerbal isn't a game that draws much attention from women, how many women you know that likes rockets and flight mechanics?

I have been in the flight sim community since 2002, and since then I have seen very few women interested in it, with the same thing happening for real flight crews, with very few women interested in being pilots (and I have been trying to convince one because I see potential in her for it) instead of being flight attendant (or being one first and then going after the cockpit)

Pretty much finished F1 texture:

http://i.imgur.com/M1hwkSe.png

I'm certainly getting faster (though perhaps not any better) at doing these textures. Got that one done entirely yesterday; including the complete UV unwrap and layout. Granted, it did take most of the day (likely 8-12 hours for UV/layout/texturing), but that is tons faster than I've done in the past (2-3 days to UV/layout, another 2-3 to do the texturing). Have learned a few new tricks in the past week or two that is making things... so much faster and less painful - like how to auto-generate a good UV-seam-mask from the UV layout, and apply that to noise/etc textures, pretty much eliminating visible seams, all while allowing for even more crazy (or organic) unwrapping to be used which itself speeds up the unwrapping process.

- - - Updated - - -

the texturing looks amazing Shadow, I'll have to ask you for a quick tips'n'tricks tutorial one day

as for ISP/thrust/mass stuff - personally I liked a kind of simple approach of KW rocketry mod. So the higher thrust, the more mass and then lower isp. Most of engines were quite balanced. Also you might consider adding delay to thrust change; I also liked this thing in KW rocketry - the biggest engines had helluva delay so they couldn't be used efficiently for vac operations, also the delay meant that they used a bit of fuel before getting to 100% thrust and lift off thus further reducing their effective isp.

I wouldn't take KR-2L, or the biggest engine as isp point of reference; for me their isp is weird and they are too op; so I would expect that squad might nerf them sooner or later. You might take a look at awesome but not updated anymore rebalance of engines by stupid_chris:

http://forum.kerbalspaceprogram.com/threads/75272-0-24-x-Stock-ReBalance-v1-4-11-09-14

I used this mod extensively when I used to play KSP (hehe)

I think that he basically downgraded isp of largest engines and upgraded smaller engines so they all followed the rule - bigger engine - higher thrust - less isp. In this way every engine had its purpose.

Another (maybe future), more complicated approach would be to divide engines into RP-1/LOX and LH2/LOX (and CH4/LOX?). Then introduce new fuel tanks for that engines (LH2 being much bigger/heavier for the same amount mass of fuel etc), or add another fuel config (LH2/LH2+Ox)for your existing tanks. You can see what Nertea did in his LH2/LOX cryogenic engines mod; from what I remember at least for vacuum engines, the higher isps were offset by higher tank mass so in the end vac engines were maybe 5-7% better dV-wise than their counterparts. I think that for SL engines LH2 and LF engines have similar performance so I think personally I wouldn't bother using them since it meant higher, more unstable rockets (with bigger 1st stage) with the same performance.

I agree the delay would be a nice feature, but the optional config files for not having them, like KW does, would be good too

about using real fuels, what about a MM patch for using RealFuels instead? Shadowmage already made his plugin to communicate with RO/RF, so that is a thing that already works, maybe a MM patch for a stockalike RF config could be offered for those willing to use non-stock fuels but also not interested in having RO?