16571 Posts in 2728 Topics - by 4924 Members - Latest Member: durizap

Author Topic: Thruster some testing 0.11.7767  (Read 330 times)

Pob

  • Full Member
  • ***
  • Posts: 112
    • View Profile
Thruster some testing 0.11.7767
« on: March 20, 2017, 06:45:37 PM »
I've just done a clean install to do some testing in the beta branch 0.11.7767

I founded a ship, a stick size 2x2x10 made of Iron Stone, this gives it some mass
mass = 8.16kt
now I stick a 2x2x2 cube of standard thrusters on the front and see what difference Material makes to the thruster.
IronTitaniumNaoniteTriniumXaniumOgoniteAvorion
Mass0.200.120.130.080.110.180.14
Break Thrust4.84.84.84.94.84.84.8
Yaw/Pitch0.580.620.610.640.620.590.61
Energy Required36.0034.3632.4030.4128.5927.0025.66

Let's start with break thrust, all pretty much the same, only difference is Trinium and it's minor
Yaw/Pitch scales directly with Mass. (I'll come back to this)
Energy require is a direct scale with material

Looking at this we can say a couple of things
1- Material has no effect on thruster force, this is shown
2- Position/Distance from the centre of mass has an effect of the Yaw/Pitch remembering to take into account the mass of the thruster, this is shown as break thrust remains constant while Yaw/Pitch changes with mass.
The Yaw/Pitch changes as the mass of the thruster shifts the centre of mass, shown by adding mass to the other end of the stick to equal out the total mass causes the Yaw/Pitch numbers to flatten out.
3- The best material for Thrusters vs manoeuvrability is Trinium, because it has the lowest mass, Xanium is a very close 2nd, it maybe 1st, for the reason of lower power draw which means less energy to support them that in turn means less total mass. Avorion is 3rd, while titanium is lighter the lower power draw puts Avorion is solid 3rd

For everyone building ships I say this, Mass and Position over material when building thrusters
« Last Edit: March 20, 2017, 06:48:20 PM by Pob »

Ohm is Futile

  • Full Member
  • ***
  • Posts: 177
    • View Profile
Re: Thruster some testing 0.11.7767
« Reply #1 on: March 20, 2017, 08:16:45 PM »
*brake thrust  :P

Also, you wrote yaw/pitch but roll is influenced as well, although I do understand that people don't tend to care as much...

As for your conclusions, Trinium is infinitely better than Xanion. On some ships, thrusters might account for about 3% of energy required. Multiply that percentage by the actual difference in energy usage between Xanion and Trinium. That's 3% of 6% or 0,18%.

Now, consider the mass difference. 0.08 vs 0.11 is actually almost 14%. Then compare that to the rough percentage of the mass of your ship thrusters might make up. Let's go with a fairly conservative estimate of about 10%. 14% times 10% is 1,4% or roughly 770% more significant than the energy drain.

Then again, that's all very subjective, at least in terms of energy consumption. Of course, thrusters represent a large share of the energy drain on small ships without shields and hyperspace cores and that kinda stuff... but chances are you don't have much choice in terms of materials for such a small ship and neither will you care about min-maxing something that small.

EDIT: might be interesting to compare those numbers with actual generator numbers. How much mass per material do you require to generate a certain amount of energy?
« Last Edit: March 20, 2017, 08:20:24 PM by Ohm is Futile »

SinBad

  • Newbie
  • *
  • Posts: 22
    • View Profile
Re: Thruster some testing 0.11.7767
« Reply #2 on: March 20, 2017, 09:56:59 PM »
good experiment, but its a little flawed as each test article has a different mass.
run the numbers again, but this time remove mass differences from the equation by adding/removing mass from the stone stick after thruster placement so that each test article (stick+thruster) has the same mass. your centre of mass will move as well, but thinning/thickening the stone stick instead of shortening/lengthening it would help to combat this. otherwise using a very long stick would make the changes in COM position much smaller.

Ohm is Futile

  • Full Member
  • ***
  • Posts: 177
    • View Profile
Re: Thruster some testing 0.11.7767
« Reply #3 on: March 20, 2017, 10:05:36 PM »
Yeah, it's not perfect, but the truth is that we already know that material don't provide more or less thrust. They are all equal in that regard. So yeah, moving the CoM could influence the data, but if you used a symmetrical design, then mass would be equally distributed, thus leaving the CoM untouched... but we're not gonna learn anything new by doing that. As I said, mass and thrust are already known variables.


Anyways, to me this thread is much more interesting in terms of energy consumption.

Pob

  • Full Member
  • ***
  • Posts: 112
    • View Profile
Re: Thruster some testing 0.11.7767
« Reply #4 on: March 20, 2017, 11:47:49 PM »
Actually I didn't know if thrust varied by material I had assumed it did.

As to roll, No, all tests had 0.0 roll, this is because the thruster is aligned to the CoM so has an effective multiplier of 0
This is part of what I'm getting at, Position vs CoM * Thruster size vs Total Mass is the important factor in rotational manoeuvrability
If I had Stuck the thruster on the side of the stick in the middle then it would have Roll and weak Yaw but 0 Pitch.

To add the correct extra mass to counter the material difference you'd need to add it to the front of the thruster not the sides to keep the CoM the same distance from the thruster and then all the values bar energy would be the same.

So keeping this in mind we can come up with the following rules for Directional thrusters.
X = left/right = yaw, Y = up/down = Pitch, Z = forward/back = roll
Z facing thruster = break thrust, Y axis offset = Pitch, X axis offset = Yaw, Z axis offset = no change
Y facing Thruster = vertical strafe, X axis offset = Roll, Z axix offset = pitch, Y axis offset = no change
X facing thruster = horizontal strafe, Z axis offset = yaw, Y axis offset = roll, X axis offset = no change

So to get the most out of your thrusters you want them as far as possible from the CoM in a direction other than the facing direction.

So with our stick ship, long on the Z axis, sticking a front/back facing directional thruster on the front will only give break thrust and 0 rotational movement in any direction.
stick a up/down facing directional thruster on the front and you'll get vertical strafe (numbers unlisted for that) and pitch rotational movement, but no roll or yaw.

Ohm is Futile

  • Full Member
  • ***
  • Posts: 177
    • View Profile
Re: Thruster some testing 0.11.7767
« Reply #5 on: March 20, 2017, 11:53:59 PM »
So to get the most out of your thrusters you want them as far as possible from the CoM in a direction other than the facing direction.
So to get the most out of your thrusters you want them as far as possible from the CoM in a direction other than the facing direction.

So to get the most out of your thrusters you want them as far as possible from the CoM and not pointing directly at the CoM.

Pob

  • Full Member
  • ***
  • Posts: 112
    • View Profile
Re: Thruster some testing 0.11.7767
« Reply #6 on: March 21, 2017, 01:53:43 AM »
So to get the most out of your thrusters you want them as far as possible from the CoM in a direction other than the facing direction.
So to get the most out of your thrusters you want them as far as possible from the CoM in a direction other than the facing direction.

So to get the most out of your thrusters you want them as far as possible from the CoM and not pointing directly at the CoM.
better way of putting it.

did another quick test.
stick again made of 10x2x2 iron stone, but out of 5 separate 2x2x2 blocks, reason will become clear.
Stick a 2x2x2 iron direction thruster on the end facing left/right, Yaw = 1.15
Move the thruster in one, so now it's a 8x2x2 then 2x2x2 thruster then another 2x2x2 block, this does move the CoM slightly, Yaw = 0.51
Move the thruster in again, now a 6x2x2 then 2x2x2 thruster then 4x2x2 block, Yaw = 0.15

So I put forward that the upshot of all this is that the reign of the cube ship is over, now the best design is the dumbbell, a long stick with rings of thrusters at each end.
or maybe a ship inside a cube frame made of thrusters.
« Last Edit: March 21, 2017, 01:55:31 AM by Pob »

SinBad

  • Newbie
  • *
  • Posts: 22
    • View Profile
Re: Thruster some testing 0.11.7767
« Reply #7 on: March 21, 2017, 08:56:28 AM »
I don't see how that makes cubes less attractive. A dumbell that covers all 3 axis IS a cube. The only thing you have shown is that the closer to com the thruster is, the less effective it is. Thats a known effect. Its because the farther from com a thruster is, the longer of lever arm its force has to induce rotation. Distance from com=force multiplied by lever length.
In my cubes, I put as much of the high density stuff in the centre as possible, and the last layer before the armour is thrusters, as far from com as possible. The armour is then a thin single plate per face, as armour hp is 100% dependant on block volume, so a single plate can be only 0.5 thick if its 30x30. I've not had any problems with manouverability yet.

Pob

  • Full Member
  • ***
  • Posts: 112
    • View Profile
Re: Thruster some testing 0.11.7767
« Reply #8 on: March 21, 2017, 12:43:26 PM »
Ahh but here is where a dumbbell has the advantage, thunk same total volume/mass then a dumbbell will be much longer on one axis and Distance from com=force multiplied by lever length, so if the ship is twice as long in one axis then you're doubling the leverage of the thrusters.

AstroOwl

  • Jr. Member
  • **
  • Posts: 93
    • View Profile
Re: Thruster some testing 0.11.7767
« Reply #9 on: March 21, 2017, 01:50:37 PM »
Ahh but here is where a dumbbell has the advantage, thunk same total volume/mass then a dumbbell will be much longer on one axis and Distance from com=force multiplied by lever length, so if the ship is twice as long in one axis then you're doubling the leverage of the thrusters.

Except that the moment of inertia along two axis which are perpendicular to that would be increased. In other words, lever arm would be bigger, yes, but ship shape would make it harder to turn in that directions. And that effect would me more significant than increasing the lever arm.

If you make a "cube" ship of size XbyXbyX (X*X*X=X^3 volume), moment of inertia is (m/12)*(X^2+X^2)=(m*X^2)/6 along each direction.

If you make a ship of same total volume and mass, but with length of 4X and other sides of 0.5X (check: volume = 4X*0.5X*0.5X=X^3), its moment of inertia for Pitch and Yaw would be:
I = (m/12)*((4X)^2+(0.5X)^2)=(m/6)*(8X^2+0.125X^2)=8.125*(m*X^2)/6.
That is 8.125 times more than before.

But the lever arm would be just 4 times bigger (thruster is just 4 times farther from COM).

Since rot. speed ~ lever arm*thruster volume/moment of inertia, it becomes 2 times lower for pitch and yaw than before.

list of moments of inertia, i used the cuboid one
what is moment of inertia (tldr: it's like mass, but for rotational movement around axis)

SinBad

  • Newbie
  • *
  • Posts: 22
    • View Profile
Re: Thruster some testing 0.11.7767
« Reply #10 on: March 21, 2017, 08:39:16 PM »
and thats just one reason why cubes make more sense. i smell a new thread...

SageThe13th

  • Sr. Member
  • ****
  • Posts: 284
    • View Profile
Re: Thruster some testing 0.11.7767
« Reply #11 on: March 21, 2017, 09:35:08 PM »
and thats just one reason why cubes make more sense. i smell a new thread...

We had a long talk about cube ships here:

http://www.avorion.net/forum/index.php/topic,2332.0.html

The cube meta or any single shape meta is going to largely be a PvP "problem".  But, with the RNG I'm pretty sure that PvP balance is screwed anyway.

SinBad

  • Newbie
  • *
  • Posts: 22
    • View Profile
Re: Thruster some testing 0.11.7767
« Reply #12 on: March 21, 2017, 10:16:05 PM »
i didnt see that one, thanks. ill copy my thread post there and remove the new thread. tidy forums are nice forums  ;)