Pitch Roll and Acceleration for (nearly) Everyship.

[Mephane]

Its great info which for me, shows why i feel the 'conda is one of the more forgiving ships .... its not the quickest to turn .... but its the most consistent and throttle doesn't matter quite so much as say an FDL which is highly throttle sensitive

More like all ships are highly throttle sensitive, except for some reason not the Anaconda. The question now is whether this anomaly is an intentional part of the Anaconda's *begins mystic incantation* ~flight model, flight model, flight model~ or a genuine bug.

 

[Sayne]

maybe he was reverse thrusting? Clippers are fast in reverse. But if a dbs is beating a clipper then the ai needs more tweaking

It was an NPC not a human, I don't PVP outside of my RoadRunner Cobra.

A human in a DBS absolutely should be able to beat a bigger ship, because it should be more maneuverable. I doubt I could beat a human in a Clipper, but I also don't PVP and lack the bloodlust and aggression to go against an equally bloodthirsty and aggressive foe.

 

[Archgeek]

Regarding the oddly-high standing pitch rate of the Anaconda and other long ships -- I'm starting to think the flight models were set using full-on rigid-body simulation of the ships' visual designs with thrust vectors where the manuvering thrusters are visible. With the Clipper and 'Conda, the control ports used in pitching are a long way from the centre, resulting in a great stonking pair of lever-arms and a surprising amount of torque. Conversely, the same model would explain the 'Conda's terrible roll performance -- it's a comparitvely narrow ship, so the control ports struggle to apply any real torque to the beast. Ships with just plain bad performance are kind of squat, boxy things that can't get as much torque to overcome their mass. By the same token, the imperial pods are likely what give the Clipper such good yaw control, providing a big fat lever arm around the z-axis.

 

[despiser]

It was an NPC not a human, I don't PVP outside of my RoadRunner Cobra.

A human in a DBS absolutely should be able to beat a bigger ship, because it should be more maneuverable. I doubt I could beat a human in a Clipper, but I also don't PVP and lack the bloodlust and aggression to go against an equally bloodthirsty and aggressive foe.

Currently yes, I can beat NPC Pythons rather easily in my Clipper. But I still think it sucks to have such disparity between human and NPC skills. It really puts a damper on reproducing accurate physics

 

[Krieger]

Regarding the oddly-high standing pitch rate of the Anaconda and other long ships -- I'm starting to think the flight models were set using full-on rigid-body simulation of the ships' visual designs with thrust vectors where the manuvering thrusters are visible. With the Clipper and 'Conda, the control ports used in pitching are a long way from the centre, resulting in a great stonking pair of lever-arms and a surprising amount of torque. Conversely, the same model would explain the 'Conda's terrible roll performance -- it's a comparitvely narrow ship, so the control ports struggle to apply any real torque to the beast. Ships with just plain bad performance are kind of squat, boxy things that can't get as much torque to overcome their mass. By the same token, the imperial pods are likely what give the Clipper such good yaw control, providing a big fat lever arm around the z-axis.

Indeed torque increases linearly with the length of a body (thrusters assumed mounted at the extremes). But the moment of inertia of a body increases with the square of the length of the body. Angular acceleration equals torque divided by the moment of inertia so in principle angular acceleration of a body becomes lower as its length increases (if all other dimensions and thruster strength are kept equal). Of course this natural tendency can be countered by mounting stronger thrusters or by concentrating the mass of the ship closely around its centre of mass, but in practice I'd expect larger ships to have lower angular (and linear) acceleration than smaller ships.

 

[Cliffson]

'Try' to get close if you can. I am always trying to get as close to larger ships as I can when fighting. They will often end up presenting there tops or undersides for extended periods as they try maneuver. If you can get this right it is often possible to sit just 'above' and off the back of them and murder them.

Yup this works like a charm, was fighting a dangerous Clipper today while undermining in my DBS, at about 700m he was just face tanking, closed to 2-300m and could laterally thrust round him faster than he was turning.

Anyone want to do the maths and work out the maximum safe distance for each ship for a 9s pitch speed based on the lateral acceleration?

 

[Yank31]

Indeed torque increases linearly with the length of a body (thrusters assumed mounted at the extremes). But the moment of inertia of a body increases with the square of the length of the body. Angular acceleration equals torque divided by the moment of inertia so in principle angular acceleration of a body becomes lower as its length increases (if all other dimensions and thruster strength are kept equal). Of course this natural tendency can be countered by mounting stronger thrusters or by concentrating the mass of the ship closely around its centre of mass, but in practice I'd expect larger ships to have lower angular (and linear) acceleration than smaller ships.

Just out of curiosity*, would this all still work in space, in the void?

I mean, *if* there's no friction... does the dimension, and even weigh, matter?

I'm having conflicting feelings about this :s


Edit :

*I meant, in the real world. Because as far as I know, ED flight model (FA ON) is closer to an atmospheric model than a space one, is it not?

 

[Cliffson]

Just out of curiosity*, would this all still work in space, in the void?

I mean, *if* there's no friction... does the dimension, and even weigh, matter?

I'm having conflicting feelings about this :s


Edit :

*I meant, in the real world. Because as far as I know, ED flight model (FA ON) is closer to an atmospheric model than a space one, is it not?

Yup, what we are dealing with here is momentum (linear and angular) and this holds everywhere as it's *complicated* but it is a property of the body in question and not it's surroundings.

In simple terms, equations of motion like F=ma (the rate of change of momentum) only hold in a vacuum on earth they'll be F=ma+terms for friction.

 

[Poseidal]

Of course this natural tendency can be countered by mounting stronger thrusters or by concentrating the mass of the ship closely around its centre of mass, but in practice I'd expect larger ships to have lower angular (and linear) acceleration than smaller ships.

Both of these are present in the Clipper, and the latter is true for it as well. Thus you see the design rationale for it and how and why it moves in the way it does.

 

[Krieger]

Yup, what we are dealing with here is momentum (linear and angular) and this holds everywhere as it's *complicated* but it is a property of the body in question and not it's surroundings.

In simple terms, equations of motion like F=ma (the rate of change of momentum) only hold in a vacuum on earth they'll be F=ma+terms for friction.

Exactly. But this tendency also holds for atmospheric flight models. Torque from aerodynamic control surfaces (or thrusters) increases linearly with aircraft length (or another dimension) while aerodynamic damping (of angular rates) increases quadratically with length. The result is the same: the larger the dimensions of an aircraft the slower it rotates.
.
We don't know how FD modelled the behaviour of ships with FA ON exactly, but it feels quite similar to an aerodynamic model with some typical aerodynamic phenomena left out, like stalling and roll-yaw coupling.

 

[Cliffson]

Updated for 1.5 Beta ships.

Could one of you lovely mods change the title to include Update for 1.5 please?

Note when comparing Viper 4 to Viper 3, the 4 is much easier to get to a good mass/optimum ratio and so has comparable maneuverability to the Mk3 in practice.

 

[CMDR Talcuris]

Awesomely useful table, only thing it's missing are the ship specific values for the speed modifier used by the number of engine pips allocated, like the Type 9 having around 0.3x max speed at 0 pips and the vulture having .9x
Would be immensely interesting to see those values for the new ships!
edit:all the values for the old ships can be found at http://elite-dangerous.wikia.com/wiki/Thrusters

 

[Achilles7]

This is not maximum figures, optimum mass is a misnomer, 50% of optimum mass is the best performance. Cobra should be roughly 400*1.16 so around 464 theoretical maximum at 50% optimum mass A grade thrusters.

I've updated OP now with this information.

464 is the fastest my Cobra managed, so according to my experience it is bang on!

 

[ahiredgun]

I m only posting to save topic for future reference.

 

[rucky]

Very informative! Thanx!

 

[jgm]

This is very useful information. Any chance of adding yaw to it?

 

[Cliffson]

Awesomely useful table, only thing it's missing are the ship specific values for the speed modifier used by the number of engine pips allocated, like the Type 9 having around 0.3x max speed at 0 pips and the vulture having .9x
Would be immensely interesting to see those values for the new ships!
edit:all the values for the old ships can be found at http://elite-dangerous.wikia.com/wiki/Thrusters

Good idea, I even meant to test that I'll try that next time I go on the Beta.

I'll planning to also add a "practical A grade performance" table based on building ships in Coriolis and using the multipliers previously calculated just for information.

This is very useful information. Any chance of adding yaw to it?

To be honest when I did the original testing I started measuring Yaw but then decided it seemed a bit pointless as in every case it was quicker to roll then pitch. Yaw for every ship seemed good enough for what it is useful for, fine tuning aim/direction.

I've not sure I've got the will to go through every ship and check Yaw as it'll take hours, but I'll keep a mind out to measure it as I naturally move between ships. Of course, feel free to measure it yourself and I'll happily add it.

 

[Cliffson]

Updated for 0 Pip ratio.

 

[MrCoffee76]

Good work, however I don't see Yaw rate. Yaw rate is very relevant.

 

[Heavenly-Hammer]

Great work. Yea...Cutter could use a little buff to the pitch speed. Maybe something more like 25.1 | 16.6 | 30.4.

Still the slowest pitch speed in the game, but not 60% slower than the Corvette and 50% slower than the Anaconda. That does seem a bit unreasonable.