I was looking for this thread and @Garaman led me to it from a Discord post. This part of Hummel's original post was the most edifying: Okay, okay, I can roll with that (as well as pitching and yawing . . .). . . There do seem to be some anomalies in the system, but not really a big deal . . . probably a result of the fact that the same underlying algorithms run all three types of vehicles in both type of milieu (atmosphere and space) which is understandable. For example, if restrict our consideration initially to our "left and right" thrusters (radial-in and radial-out in other terminology systems): one would assume the the "optimum" configuration for any given construct / thruster configuration would be four left and four right with each thruster mirrored to its corresponding sister thruster on the opposite side of the ship, and with each of the four on a side positioned at the maximum practical distance from both the longitudinal and median planes (not sure if those terms from anatomy have alternative analogues in aerodynamics so I'll just use what I know). If we think of a construct as a rectangle, then what I just described would tend to result in the four "turning" thrusters being positioned at each of the "corners" of the box. With four of them, you would get something like the maximum value from having equal distribution of force potential in all directions. This same logic would presumably hold true for all the planes of thrust (up, down, left, right, forward and back): having four thrusters of equal size positioned in symmetrical positions "at the corners" of each thurst plane would presumably optimize the torque as well as thrust of the configuration. However, I'm not so sure that this is how it plays out, and the quickest and easiest thing I can cite to support this supposition is the Ayamara: https://steamcommunity.com/sharedfiles/filedetails/?id=2431001110 The ship which has come to define the "Whirly Bird" ship-to-ship combat meta (and a fine piece of workmanship). A friend and I have been tinkering on the Ash Reforged Eden freedom server trying to discern what makes the Ayamara "tick," meaning: how is it that it achieves such exceptional maneuverability. I suspect plenty of other users have already figured this out, given the plethora of Ayamara clones now on the workshop, so I do not presume to claim "discovery" of @Artemis-Rouge secret formula. With that said however, we have made a few anecdotal observations--which may or may not be accurate, and this is the primary reason for my posting in this thread in the first place: to compare notes with other users about how this "new" (yes 2 years old is "new" to me . . . I still remember when the recipe for "plastic" in the game used "grain"! . . . I have blueprints that are probably as old as some of you . . . mind you, they use flint sparkers for propulsion and have thatch for the window screens . . . but I digress . . .) system works, and in particular that part of it referenced in the quote with which I started this post 1. It appears that there is an "optimum" distance above the horizontal (aka transverse or axial) plane at which optimum "value" in terms of pitch and yaw is added by forward and backward thrusters (retrograde and anti-retrograde). At a distance less than this, the effect on maneunver is reduced and likewise at a distance greater than this, manuver is reduced. These are anecdotal observations based on very limited "sampling" on a Freedom server so it may not be true. For a time, we thought that the effect if the thrusters were above the midline was greater than if they were below the midline but additional tests seem to have eliminated that idea. 2. When it comes to "up" and "down" force, it appears that "up" is significantly less important than "down." For example, the Ayamara has 3 Driver thrusters facing "down" and only two "Large" thrusters ("Advanced in the Reforged Eden mod I invariably play with) for up. Moreover, the Down are distributed with one at the extreme front of the ship, and two at the extreme back. As I said before, I suspect that the fact that the algorithms running this are used in both gravity wells and in space is accounting for this, and resulting in down being more "valuable" to maneuver even when gravity is not pulling a construct "down." Not a big deal, but perhaps worth noting and discussing (so we can exploit it for our our interests! Mwahhahahh! ) 3. When it comes to Yaw, any configuration other than two thrusters of a size equivalent to the largest used in propulsion/anti-propulsion placed on the extreme aft of the construct and perhaps slightly above the mass centerline seems to be a waste of resources. You get (apparently) way more bang for your buck by using one left and one right placed on the extreme rear end of the thing than by placing one on each corner. 4. The "thruster package" on each of the four corners model one might come up with based on how maneuvering thrusters have often been configured for actual spacecraft would seem to be a waste.
For most new players this will seem over-complicated. They can just stick with RCS blocks and focus on just learning how to build and what device does what. Lots of people have very bad 3d space orientation and awareness and will have no clue what to do to fix bad pitch/ yaw/ roll just with thrusters.
Is this a bug? Thrust limit is set on lower percentage than it actually reaches. The vessel fly totally horizontally, on a 0.86g planet.
@dichebach The commonly held notion is that to get maximum yaw/pitch/roll for a given amount of thrusters, you need to have them all at the furthest possible extremes away from the center of mass of the ship. However, this is a simplification that assumes that ALL of the mass of the ship is actually located at the center of mass. If you make a big Plus-Sign shaped ship where all the ship systems are on the extreme edges, the thrusters are there too, and little more than a cockpit is in the center, then having the thrusters at the extreme edges will actually not give you very good yaw/pitch/roll, because all of the thrusters are right where the bulk of the mass is. The game seems to take this fact in to account, because I tested that exact style of ship, with my thinking being that I could roll super fast, dodge every shot easily, and place the fixed guns and cockpit at the center so that the spinning barely affects my aim at all, and I discovered that this does not actually work. In other words, the effectiveness of a given thruster's position for determining how much yaw/pitch/roll it will contribute is fairly ship-specific, and the assumption that having them far away from the center is only true of most if the mass is at the center. As for the Ayamara having a lot more down facing thrusters than up facing thrusters, that's just because while doing the "dodge roll" maneuver, you need to be rolling while also either going up or down, so you only need high up force OR high down force, not both, and only one of those directions is useful when you have a lot of cargo and want to get off a high-g planet
Absolutely right. But even if mass is not all in the center of the ship, the "best rotation speed/ force" will still be achieved - for any design - when thrusters are as far away as possible from the center of mass. SO values may not be very high, but will still be better than with thrusters placed closer to center of mass. As mentioned in post #2 : Thrusters generating Torque The calculation itself depends on a lot of factors ( inertia tensor of the ship, the positions and orientations of all the thrusters of the ship compared its the center of mass,..etc) BUT information about what amount of torque is contributed by RCS and/or Thrusters is available in the STATISTIC tab of the control panel. And as explained on wikipedia (I'm not a physicist) : For an extended rigid body, the moment of inertia is just the sum of all the small pieces of mass multiplied by the square of their distances from the axis in rotation. For an extended body of a regular shape and uniform density, this summation sometimes produces a simple expression that depends on the dimensions, shape and total mass of the object. So my understanding is that "distribution of mass" along an axis is probably sampled in a few points, then averaged to generate the "simple expression" used in the angular acceleration formula. The "from the center of mass" part means that this is the 0,0,0 point of all axis, from where each mass sample point is measured, and then "multiplied by the square of their distances from the axis in rotation". And since I'm not a physicist but I can still figure out some things by experimenting, I can understand that forces applied in one axis can produce effects on the other axis, as stated in post #2. Simple experiment with holding a bicycle wheel (not rotating) and trying to rotate it perpendicular to its normal rotation axis is much easier than when the wheel is turning on its main axis. SO if the game physics reflect that, then angular acceleration will be affected on all axis by any force applied to any axis, to some degree. .
That makes a lot of sense. Interesting stuff! Looking forward to a new Golden Age with Empyrion. I recently passed the 3500 hours played (hours listed on Steam anyway, I suspect it is more because of either an older HD that died or the fact I used to play in offline mode a lot) and I'm really looking forward to 1.7. Despite my grumbling, the game is fantastic and deserves all the success it has achieved x10! Looking forward to another 3500 hours in Empyrion and what all the creative folks in the community come up with.
