[1.12.x] Alcubierre Warp Drive (Stand-alone)

[capi3101]

Alright, excellent. That oughta get me started...

RSS still good?

[capi3101]

Hey @JadeOfMaar, do you happen to have any data on Niven from the Galileo's planet pack? I was starting the process of putting together the velocities chart and while the other bodies in the Ciro/Grannus system all had nice pictures on the wiki with the information I needed, Niven's page didn't have one. I'm specifically looking for its radius and mass.

Haven't had time to delve into the others yet.

 

Recording of the YouTube video is in post at this point. Aside from me rambling on at the beginning of video about Gigantors and botching the final nadir warp over Tylo, it's serviceable as a tutorial; I'm wondering if I need to re-do it. I also realized a little while ago that for certain bodies (such as Moho), instead of an east/west course what might be required is a degree of great circle navigation (to keep a velocity vector's direction in the same plane as the target body), a topic of which I have little working knowledge. Sticking to east/west is certainly an easier proposition but will likely increase the probability that corrections will be required at the target.

[JadeOfMaar]

@capi3101 It's interesting that you mention this. I just serviced the GPP wiki, adding pages for Nero's new moons and updating other existing images. I just fixed Niven. Thanks for noticing.

6 minutes ago, capi3101 said:

an east/west course

Prograde?

Anyway, I probably cannot help you with your tutorial... I'll definitely see about understanding and enjoying it.  In the meantime (and because I like life support mods) I know enough that I can use Karborundum, or engineer my own OP sub-light propulsion system for dealing with orbital issues after a warp.

[capi3101]

Alright, thanks for the quick reply/fix.

I might have to see what an initial prograde heading does next time I'm playing around with the mod (it'd have to be along the horizon, of course). Got two Moho/Kerbin trips in my near future, looks like...


Sometimes I really dislike the KSP forum software...case in point, I have the warp data for the Galileo's Planet Pack ready, but I can't just paste it from my Excel spreadsheet into a forum table...I mean I can do one cell at a time, but who wants to do that, right?

Out of time for today at this point - I'll try to get it up tomorrow.

Edited October 16, 2017 by capi3101

[JadeOfMaar]

@capi3101 Just make it into a Google Sheet and link it.  

[capi3101]

On October 16, 2017 at 6:02 PM, JadeOfMaar said:

@capi3101 Just make it into a Google Sheet and link it.  

I suppose something like that would be easiest...

 

Coming across a bit of a problem - I've gotten out of the nice mods that have their own wikis and now I'm having to trudge through .cfg files looking for the data I need. I noticed that both New Horizons and Extrasolar don't have any specific data on planetary mass, but they both have values for geeASL. I'm assuming this value is the surface gravity in units of standard gravities. For example, Heba from Extrasolar lists 0.886 as its geeASL, which I assume is 0.886 * 9.803 = 8.685458 m/s². Assuming then that the radius value given (5067800) is in meters, I can derive its mass:

g = (GM/r²) (in units of standard gravities)
therefore (gr²)/G = M = ((5067800)² * 0.886)/6.67*10^-11 = 3.4115*10²³ kg.

From which I'd get my values - 2.5 Failsafe = 5,067,800 m at 1,498.3 m/s; 0.625 Failsafe = 7,601,000 m at 1,340.2 m/s.

What I need to know is whether or not that assumption is correct. If it is, everything's peachy. If not...

(Occurs to me, after all this math is done of course, that I could've just gone straight for the gravitational parameter - GM = gr² - not bothering with mass at all...)

EDIT: realized on my way home from work - six hours ago - that the units match up only if you're using g in terms of m/s, not standard gravities. Means the velocities I calculated for Heba earlier were off - the altitudes are still right. 4,691.3 and 4196.0 should be the correct values. 

Edited October 18, 2017 by capi3101
Realized something

[capi3101]

Boy, @JadeOfMaar, you weren't kidding around about KSS, were you? Good news is that the figures are ready for the others; I'm hoping to have the whole thing released to y'all later today.

[JadeOfMaar]

@capi3101 Nope! Not kidding  Congrats in advance, man.

[capi3101]

Hmm...the KSS wiki data is incomplete; there are 24 worlds that don't have sufficient information available for me to make the calculations for the jump to light sp.....sorry, was having a bit of a Star Wars moment there.

Probably going to have to download the mod and crack the .cfg files open. Might take longer than I'd hoped.

[StarCrusher96]

2 hours ago, capi3101 said:

Hmm...the KSS wiki data is incomplete; there are 24 worlds that don't have sufficient information available for me to make the calculations for the jump to light sp.....sorry, was having a bit of a Star Wars moment there.

Probably going to have to download the mod and crack the .cfg files open. Might take longer than I'd hoped.

I can provide the data for those 24 worlds. If you still need it.

[capi3101]

Just now, StarCrusher96 said:

I can provide the data for those 24 worlds. If you still need it.

Yes, please. I was going to say "Good timing on the release of the Kelnis system"...

I did download the mod and go through the .cfg files and I was able to find most of the data I was lacking. At this point, most of the data I'm still missing is for the Kerolon/Kerilim system. Specific worlds are Uk, Jes, Rix, Phal, Iave, Nisc, Nisc's Submoon, and Velar. Radii and geeASL are the figures I need. I'd also ask for a geeASL value for The All, but a) if I was reading the lore correctly, it wouldn't have one in the conventional sense, and b) lack of information there might make it all the more interesting to get to. 


YouTube video is finally up but I've hit another snag - apparently the eggheads at Google have removed the ability to add annotations to videos. This is a problem because the video has some points in it that I wanted to clarify through text, and there were other spots where I was really wanting to make fun of myself. Has anybody made a YouTube video since March that might be able to give me some guidance on what I might want to do instead?

Links will be up when I'm ready to share them with y'all. Soon. Seriously, soon...

[capi3101]

Alright, y'all. I'm not 100% happy with how these turned out, but they're ready for general use. My original post on the topic of navigation with this mod is here, and my full orbital velocities guide is now available for download here.

https://www.youtube.com/watch?v=sW5Yi2aNeZY

 

My apologies for the quality of the video - my rig isn't the best in the world. It's serviceable, though.

Bear in mind that all of this is for linear velocity mode. While the orbital velocities guide data is applicable for both modes, the techniques I use in the video assume linear velocity mode. Still haven't tried to grok angular momentum mode just yet.

Edited October 20, 2017 by capi3101
trying to get the YouTube video to embed...

[StarCrusher96]

On 20-10-2017 at 6:29 PM, capi3101 said:

Alright, y'all. I'm not 100% happy with how these turned out, but they're ready for general use. My original post on the topic of navigation with this mod is here, and my full orbital velocities guide is now available for download here.

https://www.youtube.com/watch?v=sW5Yi2aNeZY

 

My apologies for the quality of the video - my rig isn't the best in the world. It's serviceable, though.

Bear in mind that all of this is for linear velocity mode. While the orbital velocities guide data is applicable for both modes, the techniques I use in the video assume linear velocity mode. Still haven't tried to grok angular momentum mode just yet.

An endless list of KSS data

[capi3101]

47 minutes ago, StarCrusher96 said:

An endless list of KSS data

Naw, it's not endless. It just happens to take up the better part of four pages (i.e. about 25% of the document)... 
I was happy to do it, too. And ifwhen y'all add more systems, I'll be happy to do those too.

Saw your post about the Galactic Neighborhood mod this morning...I'll have to begin following the development of that one, I think.

 

Been trying to gain the knowledge I need to do delta-V maps too...and I've been considering what additions might need to be made to delta-V maps if you're using the Alcubierre mod (i.e. calculating a data point for low-orbit to failsafe orbit).

Edited October 23, 2017 by capi3101

[capi3101]

Going to save a link here for the DoubleDouble mod; looks like another mod like KSS that adds a bunch of binary systems. Haven't got time to crunch the numbers for it today.

On another note, does anybody know who all is/was responsible for the development of the community delta-v maps for the Kerbol system? I've been trying to grok the number crunching that goes on with those for years now; I figure it's about time I went to the source. @Kowgan, @CuriousMetaphor and @JellyCubes are the names I've seen on those maps to date - are there any more? I'd wager the developers of the same mods for which I crunched the Alcubierre figures would appreciate having delta-v maps as well and I'm certainly willing to do the work.

[helaeon]

On 10/20/2017 at 10:29 AM, capi3101 said:

Bear in mind that all of this is for linear velocity mode. While the orbital velocities guide data is applicable for both modes, the techniques I use in the video assume linear velocity mode. Still haven't tried to grok angular momentum mode just yet.

If you have questions about how angular momentum mode works feel free to PM me or ask here (tag me though). I wrote that mode and came up with the math. I wrote it mostly for myself as linear gives you free energy and I thought some kind of orbital energy conservation was more plausible as to what a warp drive would do while moving around a gravity well. RoverDude was cool enough to let me include it in the official version. 
In many ways it operates opposite of the linear mode. As you climb the gravity well (get further from the planet) your orbital velocity goes down. It's designed that it is necessary to use your thrusters more, but also allows repeated slingshots to gain orbital energy. You'll notice in linear velocity mode that doesn't work. The thing that got me to write it was the energy you have at Kerbin shouldn't allow you to escape Kerbol when you go out to Jool, you should still be solidly captured by Kerbol.
All that said, I have NO idea how it behaves with multiple star systems because all the calculations are relative to previous and current SOI, so between star systems... No idea what that reference point is.

[capi3101]

56 minutes ago, helaeon said:

All that said, I have NO idea how it behaves with multiple star systems because all the calculations are relative to previous and current SOI, so between star systems... No idea what that reference point is.

Most of the star systems I covered in the guide are replacements for the Kerbol system; I know KSS uses The All as the center of its 'universe'. Couldn't say about DoubleDouble or maybe Extrasolar.

I'll have to give the orbital momentum model another go; still haven't really had time to try it. I do know the one time I attempted to use the same maneuvering techniques for it as for linear velocity mode was a spectacular failure; wound up at Dres going about 33%c...

[helaeon]

@capi3101 Going from Kerbin to Dres you should end up going quite fast relative to Dres, but if you look at your orbit around Kerbol it should be a skinny ellipse. I think I remember needing 20 or so successive jumps to capture and did need to use thrusters. .33c though, that's excessive. I haven't ran a warp mission since early 1.2 though so it's possible that something snuck in and it's misbehaving.

Here's generally what's being done: When you engage the warp drive your orbital velocity and direction is stored as a vector, as is the orbit.h vector (your orbital angular momentum). Lots of math is done to figure out at any given current position what your velocity must be in order to make orbit.h always remain the same, provided the direction of the velocity vector does not change. This includes hand-offs between SOIs.
So if you do the usual maneuver where you drop near the star, change your orbital direction and fall a bit to gain more velocity then warp to your target... in angular momentum mode that only helps to get your velocity vector pointed as you want you gain no energy so it doesn't help you have a better relative velocity to your target planet. I do seem to remember there being some compounding floating point errors that can happen when getting very very close to Kerbol, or when your orbital ellipse becomes very very narrow. I did test it against OPM at the beginning and it worked very very well. I took quite a few trips out to Plock.

What I've found is that you warp to your target and need to use nearly the same delta-V as if you did a regular transfer at that same launch time (with some exceptions). Other than slingshotting I imagine there are other clever things one might do with the Oberth effect to gain a lot of orbital energy without spending as much delta-v as if you warped straight to your target.

I might build a warp ship this weekend and see if everything is still working as I expect.

[helaeon]

Just built a ship and there is definitely something wrong with the angular momentum mode. I'll see if I can figure out what.
It seems to not be doing the calculations (as there is no change in orbital velocity), but it's trying and chugging along in the attempt especially in map mode.

[CaribouGone]

13 hours ago, helaeon said:

Just built a ship and there is definitely something wrong with the angular momentum mode.

Good, i'm not crazy!

i was playing around with this recently (still in 1.2.2) and got very bizarre results. it seemed like crossing soi's was part of the problem.

i would suddenly gain Millions m/s of velocity. my orbit in map mode was literally a straight line out of the solar system, regardless of where in system i was. (the same thing happened at Moho, Eve, Dres, and Jool.)

I'll try again in 1.3 and try to provide some slightly more quantitative results.

[capi3101]

56 minutes ago, CaribouGone said:

i would suddenly gain Millions m/s of velocity. my orbit in map mode was literally a straight line out of the solar system, regardless of where in system i was. (the same thing happened at Moho, Eve, Dres, and Jool.)

This jives with what happened to me during my Dres attempt. If something's not working the way it should be, it would explain a great many things.

[jd284]

On 3/1/2017 at 7:51 AM, helaeon said:

@jd284 very good.
Something to note: I wrote the angular momentum conservation mode. The reason it's there is so you don't get free energy (or at least not as much as conserving velocity). As you climb out of the gravity well you have to give up your kinetic energy for potential energy and vice versa. You describe this effect pretty well. I do agree that the velocity mode is much simpler and easier to understand.

The conservation modes were meant more as a gameplay option than to solve a delta-v problem. Also, that multiple slingshots work with angular momentum mode, they don't with velocity. That's what got me to write it in the first place. Also didn't like that when I went from Kerbin to Eeloo suddenly I had enough energy to escape the solar system. That just seemed... wrong.

Very late reply, but I've been thinking about this on and off for a while.

First, I don't think either velocity or linear or angular momentum conservation is any more realistic than the others. I find it just as wrong when you move closer to the sun and gain velocity out of nowhere just to conserve angular momentum.

The only reasonably realistic way to fix this would be to conserve both energy and momentum. That's possible, of course, but only following the (non-warp) trajectories. That means the warp drive would just boost you along your trajectory at warp speed, but not be able to change the trajectory.

I'm not sure if it's worth implementing that mode though. It would be kind of interesting and quite realistics, but probably pretty useless for gameplay, since effectively it does nothing the time-warp button doesn't do already, more or less. (Except that it happens in less game time of course.)

So you'd still need to do all the usual delta-v changes using regular drives, plus you've got a super-heavy warp drive to include in those delta-v calculations. I doubt that'll be a very popular option...

[helaeon]

@jd284 back when I made the alternate mode (3 years ago! wow), I wanted to do something similar to what you were saying. However there are some limitations on the kind of math you can do with vectors; or being a chemist rather than a physicist or mathmetician I am not well versed enough in vector math to know how to do it.
So the thought was that orbital speed was a function of orbital energy and your position and travel direction. That you give yourself energy somewhere in your orbit and that allows you to climb higher in the gravity well on the opposite side, much like throwing a ball upward. So, rather than preserving velocity I'd preserve the total energy of the system. There wasn't a simple way to figure out which direction your travel vector should be so I  just left it the same. Interestingly when it  AM mode is working as intended you do need to "pay for" your jump with the usual delta-v but at the end, and with the added benefit you can use repeated sling-shot passes (can keep warping back just before periapsis) or use Oberth and always burn just on either side of periapsis then come back. Also it allowed on hyperbolic orbits to do repeated slingshots, which is something that should actually happen. I used those follow on effects as further evidence that even just from a gameplay standpoint the AM mode was a good take on how a warp drive might behave.

@CaribouGone Looking back at the code and doing some tests last night there is certainly something going on with the SOI hand-off. I think it has to do with the changes to Krakensbane and the new FloatingOrigin. Krakensbane was always a major issue in changing the velocity of your craft outside of normal game mechanics, which actually led to using it to re-position the ship in game. I have a couple of ideas on how to feed the changes to velocity into the game and try and get around FloatingOrigin and Krakensbane. Also, wondering if anyone knows of a way to suspend or over-ride G-Forces as that too is a major issue now with kerbals being able to pass out. Any thoughts as to what might be going on are welcome.
 

I'm thinking about while mucking around in the code again I might finally add that Elite Dangerous like limit on top speed based on your distance from a gravity center and the mass of that object. Is that still wanted? I think I'd have it apply to both energy conservation modes.

 

ETA an hour later: I think I have it fixed such that it behaves. I'm still getting an error in one of the modules I'll try to knock down even though it doesn't seem to affect anything. I'm thinking I do want to add an Elite Dangerous type variable speed. Throttle is just too fiddly inside a 1:10 system, if you target Jool and warp from Kerbin at 5c you have about 3 frames between SOI hand-off and slamming into the planet. Current thought is calculation based on inverse square, and that your maximum warp will be very low as you near the fail-safe.

Edited October 29, 2017 by helaeon
Further Testing

[RoverDude]

3 hours ago, helaeon said:

I'm thinking about while mucking around in the code again I might finally add that Elite Dangerous like limit on top speed based on your distance from a gravity center and the mass of that object. Is that still wanted? I think I'd have it apply to both energy conservation modes.

I like this idea

[capi3101]

3 hours ago, helaeon said:

I'm thinking about while mucking around in the code again I might finally add that Elite Dangerous like limit on top speed based on your distance from a gravity center and the mass of that object. Is that still wanted? I think I'd have it apply to both energy conservation modes.

This might not be a bad idea - I recall reading a few pages back some folks having issue with using the mod for interstellar travel because of the top speed of the engine being something like 16c as it stands right now (or thereabouts - my observations put the top speed somewhere around 11.5c). Travel time to the closest star was something like half a year, with no time warp possible. After reading that, I thought about suggesting another engine for the mod with a higher top speed, but making top speed dependent on distance to a gravitationally significant body would render that completely unnecessary.