Jump to content

[0.23.5] Realism Overhaul: ROv5.2 + Modlist for RSS 6/30/14


NathanKell

Recommended Posts

Was the realism overhaul ever set up to have compatible capsule, fuel tank, engine and coupler sizes? I may not have the right versions of everything, but last I tried, there didn't seem to be a way to build a rocket where everything fit together correctly. (2.5M engines, 2 or 3 M couplers, shrouds that didn't work etc).

I know its a work in progress, just wondering if that particular issue was fixed.

Link to comment
Share on other sites

you don't need engines that are the same size of your tanks! For example: 4m pod with a 4m tank attached, then a conical (rounded) tank that goes from 4m to 3m, then attach a 3m engine. Then use a procedural interstage adapter and maybe add bigger tanks below! Tons of possibilities!

there are 0.5, 1, 2, 3, 4m decouplers btw

Link to comment
Share on other sites

NeonEagle: 1, many engines are not throttleable (per their RL selves). 2, some parts are engine clusters, i.e. single RS-25D, quintuple RS-25D.

Streetwind: There is no single "official" recommended suite of mods, but IIRC the default for the bundler is pretty solid. The main choice involved is, do you want full-on real engines only, or do you want RftSEngines, where the engines have realistic performance but don't 100% model real engines and thus allow more generic rockets. Also, if you're using RPL as your tech tree, you want RftSEngines.

frisch: the bundler is the solution because we *can't* have a single big zip due to mod licenses. And I certainly am not taking you as complaining, no worries! :)

As AbeS says, capsules should all have appropriately-sized decouplers, both resizes of stock decouplers and ones from Deadly Reentry. Though, yeah, I've basically stopped ever using non-procedural interstages or decouplers; they work for everything, and are super-useful *and* super-pretty. A bit tricky to get the hang of, but once you do... :)

(Also, you don't need to match engine size to tank size. You don't even need a conical tank; often a 2m engine will look fine on a 3m tank, for instance.)

Link to comment
Share on other sites

Yeah, the bundler came with RftS and RPL pre-selected, so that's what I went with... not convinced I like the style of RPL (it kind of ruins immersion), but I'll give it a chance for more than the first two nodes.

I left most of the core mods intact and only added some things like Kerbal Alarm Clock, Engine Ingitor, Docking Port Alignment Indicator and so on. Threw out MechJeb and Procedural Fairings.

RftS has actually given me a rather hilarious problem that I never thought I would be having - I am struggling to find an engine that is weak enough for the first two missions. With just a stayputnik, a 0.5m x 1.8m stretchy tank, an engine, a communotron and some chutes, I end up with a rocket that weighs less than 0.7t full and less than 0.2t empty, but the smallest bottom-mounted engines (discounting the orbital maneuvering systems with their anemic 1.5 kN) are giving me around 28-30 kN of thrust even when choosing the worst possible fuel. That's a TWR of over 4 on a liftoff stage, rapidly increasing to over twice that before even leaving the lower atmosphere... and I cannot throttle them down (not even with the thrust limiter), and I can only ignite them once. As a result, my rockets are traveling upwards faster than they do while falling back down, giving me the choice between shutting the engines off early and thus not reaching the target altitude at all, or burning themselves up before getting high enough. Deadly reentry? Pah, reentry is a gentle caress compared to launch! :P

Didn't have very much time to experiment yesterday evening, but I'm thinking about rigging something up with two of the 12 kN radial engines (I wish I could have one single inline 12 kN engine!) on a slightly shorter tank so they cut out early, and then another small tank on top with two radial OMS engines to ignite afterwards. I'll still be well inside the atmosphere, but I just need them to assist me in going up in a straight line a little longer.

Edited by Streetwind
Link to comment
Share on other sites

Still it's good that you have brought the issue. Remind that the "Stock SAS" option may not be better in all situations, for example once in orbit it is a RCS guzzler since it over-corrects so much. You have to switch it off then.

Wait a moment... does dtobi's gimbal plugin provide either: gimbal range tweakable during flight or roll channel coefficient sort-of-thing (so that it will, for example, turn 10 deg for pitch/yaw but only turn 2 deg for roll channel to avoid exceeded shake)?

Link to comment
Share on other sites

Wait a moment... does dtobi's gimbal plugin provide either: gimbal range tweakable during flight or roll channel coefficient sort-of-thing (so that it will, for example, turn 10 deg for pitch/yaw but only turn 2 deg for roll channel to avoid exceeded shake)?

IIRC, kind of. It can set a specific max range for each axis that the engine will gimbal on, so certain engines from SSE (for instance) will have a large pitch vector, but a mediocre roll and yaw control. You can change the mode between precise movement and smooth movement as a tweakable, along with gimbal range, but you can't tweak the gimbal ranges in-game. Currently however, I believe that all of the engines in the SFJB Real Engines have Roll, Pitch and Yaw set to the same values as eachother for a single engine.

So basically:

Gimbal Range: Not tweakable, but much more controllable and more easily configurable.

Gimbal Speed: Tweakable.

Gimbal Movement: Tweakeable.

Roll Control: It has it, and it's tweakable.

Link to comment
Share on other sites

Wait a moment... does dtobi's gimbal plugin provide either: gimbal range tweakable during flight or roll channel coefficient sort-of-thing (so that it will, for example, turn 10 deg for pitch/yaw but only turn 2 deg for roll channel to avoid exceeded shake)?

No to both questions. But that seems a loaded question - what exactly are you suggesting? Because given the same gimbal configurations (I test with 4 different plugins) MechJeb is outperformed by the stock SAS (in the ascent phase, at least).

Link to comment
Share on other sites

No to both questions. But that seems a loaded question - what exactly are you suggesting? Because given the same gimbal configurations (I test with 4 different plugins) MechJeb is outperformed by the stock SAS (in the ascent phase, at least).

Because according to my experience of using my own plugin, if the gimbal range / roll channel coefficient are not properly adjusted (in most cases: too large), it will perform quite badly with either MJ or stock SAS... I think it's necessary to allow some tweaking in these values, perhaps even during flight.

Link to comment
Share on other sites

Because according to my experience of using my own plugin, if the gimbal range / roll channel coefficient are not properly adjusted (in most cases: too large), it will perform quite badly with either MJ or stock SAS... I think it's necessary to allow some tweaking in these values, perhaps even during flight.

Ah OK. But isn't the issue the decoupling between the controller and the control surface? In other words, the difference between the commanded deflection and the actual deflection? Because from what I've tested, the higher the response speed is, the better for MJ/SAS to handle it. As you can see in the video I posted, once the initial roll-wobbling, its stable as a rock. But I still don't understand why MJ commands so much roll during the first 60 seconds or so and can't stabilize it - sounds more like a bug since it does not happen in the other axes.

Fiddling manually with those values in flight, specially when you have an engine cluster, is not something I would expect when using an autopilot. I'd rather fly with SAS only.

Link to comment
Share on other sites

Ah OK. But isn't the issue the decoupling between the controller and the control surface? In other words, the difference between the commanded deflection and the actual deflection? Because from what I've tested, the higher the response speed is, the better for MJ/SAS to handle it. As you can see in the video I posted, once the initial roll-wobbling, its stable as a rock. But I still don't understand why MJ commands so much roll during the first 60 seconds or so and can't stabilize it - sounds more like a bug since it does not happen in the other axes.

Fiddling manually with those values in flight, specially when you have an engine cluster, is not something I would expect when using an autopilot. I'd rather fly with SAS only.

That's a good point... if the gimbal/control surface have response time, the lag might result to oscillation. but i wonder if it's possible to write a generic autopilot system that is aware of these and can generate proper control command accordingly.

As for exceeded roll wobbling, I guess it's of multiple factors: first, MJ AP might automatically refine its control coefficients with time, second is perhaps the angular drag factor increases after your rocket gains enough speed, third is, the destination attitude of MJ AP contains roll angle information and it might have a big initial error before launch, but after the "initial roll-wobbling", the error got close enough to zero and the roll channel command will be much smaller as well.

Link to comment
Share on other sites

lireaper: the output log is a file. Upload it to dropbox.

Note that we want output_log.txt, *not* ksp.log

jandcando: delete KER. Delete toolbar and install it fresh. Try it. If it works, then install KER. If it breaks, it's pretty clearly a KER issue.

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
×
×
  • Create New...