[0.90WIP] Procedural Parts - Parts the way you want 'em 0.9.21, Dec 19

[Eggman360]

Ah it's nice to see that the force empty code is in, I'll update my mod packs to V 0.1.2 and add force empty functionality to Kethane and certian parts, my only concern is will it still show you the resource amount in the VAB/SPH or will it just say "Kethane Tank" and no visible volumes?

If that's the case then it might be necessary to have a tag in the right click GUI (in kthe procedural kethane tank for example saying "Can Hold xx units of kethane" otherwise how will people know how large to make them so they're big enough?

[OtherBarry]

Ah it's nice to see that the force empty code is in, I'll update my mod packs to V 0.1.2 and add force empty functionality to Kethane and certian parts, my only concern is will it still show you the resource amount in the VAB/SPH or will it just say "Kethane Tank" and no visible volumes?

From memory, there's no tweakable fuel amount bar, but everything else is the same. It didn't feel like anything was missing or that it would be hard to work out.

[StainX]

Nice Mod!

I uninstalled all non-procedural tanks, even the stock ones... no problems so far... thank you for the work you put into this!

[OtherBarry]

Nice Mod!

I uninstalled all non-procedural tanks, even the stock ones... no problems so far... thank you for the work you put into this!

While it's fantastic that you like the mod so much, you may want to be wary of deleting stock parts, as many mods use they're textures, as they assume everyone has them. I've run into that problem a few times after going on a deletion spree. Particularly with the fantastic Landertron mod.

[Rioliki]

Nice Mod!

I uninstalled all non-procedural tanks, even the stock ones... no problems so far... thank you for the work you put into this!

Instead of deleting stock fuel tanks, you`ve better just make them invisible by changing configs. I`m not sure enough, but this is achieved by changing

subcategory = -1

[TK421d]

Thanks for the bug report. Looks like it's attaching to the node on the other end.

I'll look into it, shouldn't be too hard to fix I imagine.

Edit:

Yep, was easy fixed

Was due to a bug in KSP. I was working around for when a ST attaches to something else, but not when the ST was the parent.

i noticed a similar oddity, add 2 procedural tanks to an object, grab the root one and set it aside (happens more often with copy-grab). now re-attach at the new lowest point. there will be about a 1/2 meter gap but the parts act connected.

[K3-Chris]

Does the RCS and regular tank really need to be separate parts? Also I got modularfueltanks installed and for some reason these tanks don't seem to support that? A bit annoying having to use 2 tanks to have LF+Oxi and MP.

[OtherBarry]

Does the RCS and regular tank really need to be separate parts?

They don't need to be separate parts, but personally I find it better to have two separate tanks. You could quite easily make them one part by copying the RESOURCE part of the TankContentSwitcher module from the RCS tank cfg file into the same module on the regular tank cfg underneath the other resources.

Also I got modularfueltanks installed and for some reason these tanks don't seem to support that?

RealFuels works perfectly, so it shouldn't be hard to add MFT compatibility. I'll look into it tomorrow, actually today, its almost 2am here...

In the interim, you can edit the cfg files yourself to include whatever you like. The cfg files are pretty well documented, but if you need help, the spreadsheets in the source folder of the mod, as well as Eggman's mod pack may be of use. I'll be able to explain this better in a few hours when I wake up.

[RaccoonTOF]

@K3|Chris: The reason they are separate parts is because the tankage mass for RCS tanks is a different weight than normal fuel tanks per volume. This is especially true for the extra tank types you get with Modular Fuel Tanks or Real Fuels. There are some issues with Modular Fuel Tanks if you are not running the correct version of MM, make sure you are using the latest version in the mod packs you have installed, rather than the most recent dev release version of it. Also make sure you ARE running the most recent PP DLL, which I dunno if the first post has been updated to include the one just posted a few pages back or not yet.

@swamp: SRB's max allowable thrust to diameter still appears to be off by nearly an order of magnitude for replicating real-world SRBs. Very obvious example that fails: a 3.75m diameter surface-optimized SRB has a max thrust of 2975 kN. The Space Shuttle 5-section SRBs (3.71m diameter) have a real world liftoff thrust of between 14,012kN and 15,617kN depending on time period. And the proposed Advanced Rocket Booster SRB design has a projected 20,015 kN in the same diameter. Are you sure that you aren't calculating the max allowable values in tons-force instead of kN in the math for the code? 15,617kN is 1,592 ton-force, which would fall right in the range being set for the SRBs in game

[OtherBarry]

@K3|Chris: The reason they are separate parts is because the tankage mass for RCS tanks is a different weight than normal fuel tanks per volume...

Actually, the dry density of a fuel tank is based upon which resource is in it. Not that you can really tell in game as the only tank with switchable fuel types, all of the resources have the same dry density.

@swamp: SRB's max allowable thrust to diameter still appears to be off by nearly an order of magnitude for replicating real-world SRBs...

This is what the RealismOverhaul config file is for. If you don't want to do the stepDiameter changes as well, have a look at the conversation I had with WhiteOwl a few pages back. It'd probably also help to use RealFuels.

[RaccoonTOF]

Yes, I'm using the entire RO suggested modlist (including RealFuels), and about 2 dozen others besides (active texture management is my dearest friend ). The numbers above were with 0.9.3 of PP installed, with the config changed from "notused" already. I have similar "unreal" issues with the secondary standard kerbin install that I also have though, though I didn't check exact numbers there. I know that they are of course lower due to the significantly smaller scale, but even trying to model them "performance-wise" rather than "realistic numbers for thrust- and weight-wise" they still fall very short (so do the stock SRBs in the game of course, but I would hope that the ones in PP are at least capability-wise useful in the same sorts of situations one would use SRBs in real life - and right now they are not, they are almost purely for looks.)

As for the dry density of tanks - it is dependent on TWO things: the tank class, AND the actual tanks installed. The tank(part) type determines the basemass per tank(part) volume, and the resource within the tank(module) determines the tankage mass for the tank(module) actually chosen. The "default" tank type is 0.000625 tons per total liters of tank volume, while RCS tanks have a base mass of 0.0016 tons per total liters of tank volume. This is without ANY resources added to the tanks, it is the base mass for the empty part itself. (In RealFuels these numbers are 0.000011 tons per liter total volume for the default type, and RCS is 0.0004 ton per liter volume - stock tanks really have WAY too much of their mass in the structural part of the tanks compared to real life tank weights).

EDIT: Oh, and with neither Modular Fuel Tanks NOR RealFuels installed, just using PP, they are different as well - dryDensity for Liquid tanks is .1089 t/kL and for RCS tanks is 0.22 t/kL. He already had said that he had MFT installed though, so those were the first numbers I checked

Edited April 24, 2014 by RaccoonTOF

[OtherBarry]

Sorry, I was half asleep when I wrote that. You're right, in realfuels and MFT mass is partially dependent on thentype of the tank. In stock PP, RCS and liquid fuel/ox have different densities, but the density is based on the lines in the tank content switcher module, not on other modules in the tank, so by adding the RCS resource lines from the RCS tank to the liquid tank, you would still have the 0.22 density of RCS when you select it.

If your using the stepDiameter cfg, you should change the "*= 1.25" lines to "*= 2.5", as this should give you proper thrust limits. Should be just over the max thrust of a Titan iv SRB wi a 3m diameter. Although I haven't tested it, so let me know if that's wrong.

Also, PP is designed to be as close to stock as possible, so the SRB's won't be any more powerful. Unless you edit the Cfgs yourself.

Edited April 24, 2014 by OtherBarry

[swamp_ig]

There's a fixed version of the config here: https://github.com/Swamp-Ig/ProceduralParts/blob/master/ModuleManager/Rescale1m.cfg.notused

[RaccoonTOF]

Ah, excellent. I hadn't opened up the config myself, just renamed it - I thought the 2.5 multiplier was already included in the 0.9.3 version. The above should fix the issue nicely, about to go check it out. (Looking over the cfg more closely this time, I also changed the step size to 0.05 instead of 0.1, so you can still get the quarter-meter values at least).

EDIT: It's still too low, but it is a LOT better. Current max thrust for a 3.7m SRB is now 7.24MN in-game, which is indeed enough to replicate the Titan IV SRB, but still less than half that needed to replicate the space shuttle SRBs, and about 1/3 that needed for the Advanced Booster (Solid) design proposed by ATK for SLS. If I understand the math correctly now, this means I should be able to get proper values by putting ~4.6 into that multiplier, so I'll go test that now

Edited April 25, 2014 by RaccoonTOF

[swamp_ig]

You can get quarter meters by using the slider between the arrows.

I know it's not super intuitive that there's a slider there, but I had to live within the confines of what was possible

[RaccoonTOF]

Oh, I use the sliders already too - but KSP seems to not like leaving parts left exactly to what the slider gets set to. When you go away and then back to the part, the value seems to be changed to some seemingly arbitrary value that was just "close" to your slider setting. So I like using the .05 for the exact quarter-meter settings, and the sliders just for fine visual tweaking instead.

Also, it appears that the proper multiplier for thrustScaleFactor to be able to replicate all current/previous SRBs designed, as well as the proposed new Advanced Booster SRB option, is *7. This gives a 3.7m booster a max thrust of 20.3MN, which covers the Adv. SRB 20.01MN proposed value with a little room for leeway if the actual motor performs a bit better than ATK expects (as they tend to have a history of doing - one of the few companies in the space industry that pretty consistently under-projects performance numbers ). Unfortunately, it does NOT match for burn duration when built to proper length - a 53m surface SRB at 15.1MN thrust has a max fuel capacity to burn for 1m17s. The Shuttle SRBs of the same size and thrust burned for 2m12.5s. However, total delta-V supplied appears to be pretty close to accurate, since the KSP SRBs supply full thrust for the entire burn duration, rather than tapering off at the end of the burn.

I am testing it on a Block 0 SLS mockup that I'm making, with real-life weight and size values on the core stage, tweaked main engine settings to duplicate RL, and a dummy weight payload - the SRBs were the last piece missing for accurate numbers, and now I can get to within 2-3% of accurate total dV and surface TWR values for the Core + Booster assembly combined with the projected payload capacities - so good enough for me (even though in real life, a 3% loss in dV capacity can mean the difference between a proper orbital trajectory, and a tumbling re-entry and burning up of the payload, as happened with a recent Chinese launch that had the orbital burn cut off a little bit sooner than planned )

Edited April 25, 2014 by RaccoonTOF

[Galane]

How about offset cone tanks or nosecones where one side is vertical? That would be for making nicer looking side boosters or for doing crazy things like adapting four Rockomax diameter tanks to a stock quad coupler.

[swamp_ig]

Hmm...

The lengths and diameters should snap to the nearest 0.001 m . Thrust will round to the nearest 1 kN. There's no config variables controlling those, it's hard-coded.

If they're snapping to some larger increment I'd like to know so I can debug it. Values close to the end of the slider can't be picked, I'm aware of that (like 1.003 or 1.002).

Can you give me a minimal test case / steps to reproduce?

Edit: Have replicated and fixed the issue.

- - - Updated - - -

How about offset cone tanks or nosecones where one side is vertical? That would be for making nicer looking side boosters or for doing crazy things like adapting four Rockomax diameter tanks to a stock quad coupler.

Been requested a couple of times already. Not currently high priority I'm sorry.

- - - Updated - - -

That's annoying.... since when did the forum merge posts?

Edited April 25, 2014 by swamp_ig

[Solestis]

Sometimes when I copy a procedural Tank part in the VAB, it starts to lag terribly and the Error Console gets spammed with errors.

[Error]: *ST* unable to find child node from parent: proceduralTankLiquid (UnityEngine.Transform)

The only way it seems to get rid of this error spam is to get rid of all the procedural parts on my ship

I searched for this error but couldnt find anything, anyone else has this problem or knows a fix?

[K3-Chris]

Could we get a concave shape for smooth cones? They all seem to be straight or convex, or possibly some way to set up custom curves? Or is that massive amounts of work?

Would also be nice if the procedural nose cones allowed surface attachment.

Edited April 25, 2014 by K3|Chris

[RaccoonTOF]

First page answers the first part of that Chris:

Features That Are Not Planned

Shapes with 'holes' in them and concave shapes - including toroids.

As for surface attachment, that should be a pretty easy change I would think. Change the NoseCone.cfg file from

attachRules = 1,0,1,0,0

to

attachRules = 1,0,1,1,0

If you want to use surface attach as partial clipping (to somewhat replicate "flat side" or "offset" nosecones, for example) change the last 0 to a 1 as well (which allows the part to still be placed even when it is in collision with another part).

[K3-Chris]

Why is a concave cone a no-no? I can understand toroids, that's a very different shape, but if you can pick from convex and straight versions of cones why not a concave?

[swamp_ig]

Could we get a concave shape for smooth cones? They all seem to be straight or convex, or possibly some way to set up custom curves? Or is that massive amounts of work?

Concave shapes won't work very well. This is a unity issue - colliders for parts must be convex.

I've removed the ability to create custom curves essentially for this reason - it's too hard to get the parameters to work otherwise.

Just to expand on that a bit - It's not just the convex issue, if you can freely edit curves you'll be able to create all sorts of wierd stuff like self-intersecting and overlapping curves, stuff that goes off to infinity... who knows what. Given the confines of what unity will allow you to do, plus the limits I've placed on parts, the curves you're actually allowed are pretty limited and more or less collapse down to the set that are provided.

Would also be nice if the procedural nose cones allowed surface attachment.

Yes this is true, will enable for next release.

Edited April 25, 2014 by swamp_ig

[K3-Chris]

Ah that explains it, sorry didn't know about that particular issue.

[mikebell]

With 0.9.3 everything is looking good :-)

I've got a couple of small issues though:

- Tech limits on tank size are ignored if you go out of VAB and back into it after adding the part

- If you have multiple SRBs on the same stage, then all the thrusts end up the same, even if you specify different thrusts in VAB (e.g. if you want a sold fuel "main engine" at the bottom of your stack + boosters on the side).

Also, is there an easy way to up the max diameter to 2m at lower tech levels when using 1m scale? As it is you get stuck with a 2m pod on top of a 1.5m stack. To work around this I just added:

TECHLIMIT {

TechRequired = generalRocketry

diameterMax = 2.0

volumeMax = 3.5

}

to the RealFuels tank files and made a similar change to SRBs.