Ship design rule of thumb?

[Iranon]

Please stop repeating the same falsehood (that's trivially easy to check) over and over.

Double internal components for twice the capability and twice the component weight -> you need more armour to cover the whole thing at the same thickness but LESS THAN TWICE AS MUCH.
So for the scaled-up design instead of two of the originals, you need LESS ARMOUR.

[Mor (OP)]

Should TWICE mean anything to me? @DIT_grue suggested that all the formulas given are linear, i showed at least one that it isn't. If you want something more specific you are going to need the rest (Cite, please). Also you might want to consider that engine components are limited by size 50.

[Iranon]

Yes it should. The scaled-up ship has twice as many components under armour, but less than twice the armour to achieve the same protection.
If you build two smaller ships, your armour thickness doesn't magically double, so having the same protection requires less armour on the larger ship. Double the size -> twice the capability and thicker armour, or twice the capability, same armour and weight savings.

This shouldn't need examples because it's so straightforward, but here two tiny but armoured fighters:

Code: [Select]

Gnat class Fighter    167 tons     2 Crew     37.4 BP      TCS 3.34  TH 32  EM 0
9580 km/s     Armour 4-2     Shields 0-0     Sensors 1/1/0/0     Damage Control Rating 0     PPV 1
Maint Life 0 Years     MSP 0    AFR 33%    IFR 0.5%    1YR 2    5YR 26    Max Repair 16 MSP
Intended Deployment Time: 0.3 months    Spare Berths 0   

32 EP Magneto-plasma Drive (1)    Power 32    Fuel Use 336.02%    Signature 32    Exp 20%
Fuel Capacity 5 000 Litres    Range 1.6 billion km   (46 hours at full power)

Gauss Cannon R3-17 (1x3)    Range 16 000km     TS: 9580 km/s     Accuracy Modifier 17%     RM 3    ROF 5        1 0 0 0 0 0 0 0 0 0
Fire Control S00.1 8-2000 (FTR) (1)    Max Range: 16 000 km   TS: 8000 km/s     37 0 0 0 0 0 0 0 0 0

This design is classed as a Fighter for production, combat and maintenance purposes


Doubling all internal components gives me

Code: [Select]

Gnat - Copy class Fighter    329 tons     4 Crew     73.8 BP      TCS 6.58  TH 64  EM 0
9726 km/s     Armour 5-4     Shields 0-0     Sensors 1/1/0/0     Damage Control Rating 0     PPV 2
Maint Life 0 Years     MSP 0    AFR 65%    IFR 0.9%    1YR 3    5YR 51    Max Repair 16 MSP
Intended Deployment Time: 0.3 months    Spare Berths 1   

32 EP Magneto-plasma Drive (2)    Power 32    Fuel Use 336.02%    Signature 32    Exp 20%
Fuel Capacity 10 000 Litres    Range 1.6 billion km   (46 hours at full power)

Gauss Cannon R3-17 (2x3)    Range 16 000km     TS: 9726 km/s     Accuracy Modifier 17%     RM 3    ROF 5        1 0 0 0 0 0 0 0 0 0
Fire Control S00.1 8-2000 (FTR) (2)    Max Range: 16 000 km   TS: 8000 km/s     37 0 0 0 0 0 0 0 0 0

This design is classed as a Fighter for production, combat and maintenance purposes
Exact doubling of all internal components... but a total of 2.1 armour for 5 layers instead of 1.2 for 4.
If we replace two of the smaller fighters with one larger one, we use the same internal components but less armour for a cheaper, faster, longer-ranged, better-armoured package.
The advantages are more pronounced the larger ship can adjust internal components (in ths case, eliminating the second fire control).

While there are reasons for smaller vessels (sensor footprint, shipyard investment, tactical flexibility, maintenance concerns), armour efficiency isn't one of them; quite the contrary.

[0111narwhalz]

Now, I don't know the math, but it would seem that doubling everything shouldn't change any ratios (armor aside).  Think of strapping two identical rockets together: The acceleration doesn't change; you have twice the thrust and twice the mass.  The armor doesn't change; you've just strapped them together.
Now, let's actually make a singular ship with twice the systems (and thrust).  Acceleration will change, because certain systems (such as the bridge) have not been duplicated.  Also, assuming the armor is based on surface area (and ships are similar in shape), armor should obey the square-cube law.  This means that a larger ship should have less surface area than a pair of smaller ones.  And, because we assume the same armor thickness, surface area is equivalent to mass. 
So, larger ships are more efficient than conglomerations of smaller ships because:
   They don't need redundant systems, such as bridges.
   They don't need as much armor to cover the same volume

Again, I don't have the math, but I'm sure Dev-Man has thought through this just as much.

[Mor (OP)]

@Iranon I am not saying you are wrong, only that that I know two things
1. On the lower end of ship size, there should be an increase in performance (tied to engine size\fuel consumption)
2. That armor requirements grow exponentially (surface of a sphere is power 2) which means that with a sufficiently large x  it will exceed linear growth.

You make the assumption that double the size components will maintain operational parameters (compared to using 2 ships of a kind) will be 1:2 or 1:1. While i am saying that I without the formulas that govern everything I can't  go beyond the general principle above. Its great that you want to experiment but i'd suggest using more extreme case than 167ton fighter.

EDIT:

Also, assuming the armor is based on surface area (and ships are similar in shape), armor should obey the square-cube law.  This means that a larger ship should have less surface area than a pair of smaller ones.

Good argument, since their radius is effect by tonnage(volume), then it should be something like 1 to 1/n sqrt 3 (unless I am having a brain fart)

as for the missile example its not necessarily true because e have crew, fuel, maintenance, power morale etc consideration and I can't say for sure that they all scale and non are derived. 

[Charlie Beeler]

Yes the model for the required armor is based the surface of a sphere.  But volume will always grow faster than the surface.  To complicate things a row of armor is actually 1/4 of the result of calculating the surface area. 

Example 1:  5k/ton ship (100 hull spaces) has a radius of 3 for a surface of 104.  The resulting columns are 26. 

Example 2: 25k/ton ship (500 hull spaces) has a radius of 5 for a surface of 304.  The resulting columns are 76.
(these are not pure results because rounding)

Note that total 'volume' increased by a factor of 5 the surface area increased by a little less than a factor of 3.

This is the relevant code for Standard Aurora

    Radius = ((ClassSize * 0.75) / PI) ^ (1 / 3)
    SurfaceArea = (Radius ^ 2) * 4 * PI
    ReqArmStrength = ArmourThickness * (SurfaceArea / 4)


As far as engine performance in Aurora is concerned it is a straight linear calculate for speed.  (hull spaces / total engine power) * 1000.  All factors being equal for power production the same percentage of hull devoted to engines will produce the same maximum speed regardless of ship size.  Note that for purposes of speed fuel consumption is not relevant in Aurora.

Fuel efficiency/consumption is also straight forward.  The baseline is 1liter per EPH.  There are only 3 modifiers: 1) Fuel Consumption tech and 2) Engine Size 3) MIN/MAX Engine Power Modifier.

Fuel Consumption Tech is the most basic modifier.  Liter Per Power Hour (LPH) * tech.

Engine Size is also basic.  Engine Size equates to reduction percentage.  i.e. 10hs engine reduces fuel consumption by 10%.

MIN/MAX Engine Power Modifier is not as obvious.  it is Consumption Modifier = Engine Power Modifier ^ 2.5.


All of this information is available in various posts within the Mechanics section of the forum.  Some of it does take some digging.  The engine performance formulae are found in the v5.4/v6 changes. 

[Iranon]

@Iranon I am not saying you are wrong, only that that I know two things
1. On the lower end of ship size, there should be an increase in performance (tied to engine size\fuel consumption)
2. That armor requirements grow exponentially (surface of a sphere is power 2) which means that with a sufficiently large x  it will exceed linear growth.

You make the assumption that double the size components will maintain operational parameters (compared to using 2 ships of a kind) will be 1:2 or 1:1. While i am saying that I without the formulas that govern everything I can't  go beyond the general principle above. Its great that you want to experiment but i'd suggest using more extreme case than 167ton fighter.

EDIT:Good argument, since their radius is effect by tonnage(volume), then it should be something like 1 to 1/n sqrt 3 (unless I am having a brain fart)

as for the missile example its not necessarily true because e have crew, fuel, maintenance, power morale etc consideration and I can't say for sure that they all scale and non are derived.

You wanted an example, I gave you one. Small fighters don't change anything, something simple with a decent armour percentage to make the things in question noticable seemed appropriate. Of the things you know: I'm not even considering engine efficiency - very obvious whether it applies or not, in my design practice it usually doesn't (most engines I design are size 1 or size 50).
We don't need to be given formulas when things obviously behave perfectly linearly.

"That armor requirements grow exponentially (surface of a sphere is power 2) which means that with a sufficiently large x  it will exceed linear growth."

This is where the real problem lies. The quadratic growth you describe is relative to the radius, not the mass. The increase in mass relative to the radius is cubic, so the increase in surface relative to mass is (which is what we're looking at when considering armour weight at a given thickness relative to ship size) is logarithmic.

You seem to assume that the ship with doubled components would be twice as long. And possibly confuse the implications of the square-cube-law in Aurora or in real life.
IRL, you can't just double something in all dimensions for 4x the surface area and 8x the volume and weight... structural strength increases at a lesser rate than load, so things will fail eventually (there are other considerations depending on what you are looking at, but this seems the most applicable one). In Aurora, you can... and save tons of armour.

[Mor (OP)]

Yes the model for the required armor is based the surface of a sphere.  But volume will always grow faster than the surface.  To complicate things a row of armor is actually 1/4 of the result of calculating the surface area. 

Yes,@0111narwhalz already reminded of that. n small craft with similar volume(power 3) to a big one will have a 1:n sqrt 3. So when you calculate surface (power 2) the relation of  n small crafts with r compared to big craft with r*nsqrt3 will be the formula I already given above (that is if I din't messed it up in my head). Which support that idea regardless of the coefficient. Anyway thanks for digging up the formulas for the engines, I can make use of them! If someone can dig up the rest for derived stats we can even put this to rest with a pretty graphic.

[83athom]

Its great that you want to experiment but i'd suggest using more extreme case than 167ton fighter.

Then here I come. FOr this I will double every component and let Aurora auto-set crew spaces.

Code: [Select]

test class Cruiser    48 250 tons     1408 Crew     70215 BP      TCS 965  TH 5000  EM 45000
5181 km/s     Armour 10-118     Shields 1500-300     Sensors 150/150/0/0     Damage Control Rating 50     PPV 468.4
Maint Life 3.85 Years     MSP 46476    AFR 372%    IFR 5.2%    1YR 4953    5YR 74293    Max Repair 7875 MSP
Intended Deployment Time: 12 months    Flight Crew Berths 0   
Hangar Deck Capacity 2000 tons     Magazine 1880   

5000 EP Photonic Drive (1)    Power 5000    Fuel Use 5%    Signature 5000    Exp 10%
Fuel Capacity 1 000 000 Litres    Range 74.6 billion km   (166 days at full power)
Omega R300/360 Shields (100)   Total Fuel Cost  1 500 Litres per hour  (36 000 per day)

Triple 40cm C6.25 FGR Laser Turret (5x3)    Range 1 400 000km     TS: 25000 km/s     Power 126-19     RM 12    ROF 35        42 42 42 42 42 42 42 42 42 42
Twin 15cm C6.25 FGR Laser Turret (10x2)    Range 720 000km     TS: 100000 km/s     Power 12-12     RM 12    ROF 5        6 6 6 6 6 6 6 6 6 6
CIWS-1000 (5x16)    Range 1000 km     TS: 100000 km/s     ROF 5       Base 50% To Hit
Fire Control S02 175-100000 H10 (1)    Max Range: 350 000 km   TS: 100000 km/s     97 94 91 89 86 83 80 77 74 71
Fire Control S02 700-25000 H10 (1)    Max Range: 1 400 000 km   TS: 25000 km/s     99 99 98 97 96 96 95 94 94 93
Vacuum Energy Power Plant Technology PO-40 (6)     Total Power Output 240    Armour 0    Exp 5%

Size 4 Missile Launcher (50% Reduction) (100)    Missile Size 4    Rate of Fire 50
Missile Fire Control FC405-R1 (10%) (1)     Range 405.0m km    Resolution 1
S4 ASM A (470)  Speed: 180 000 km/s   End: 20.3m    Range: 219.8m km   WH: 49    Size: 4    TH: 600/360/180

Active Search Sensor MR270-R1 (10%) (1)     GPS 360     Range 270.0m km    MCR 29.4m km    Resolution 1
Active Search Sensor MR2700-R100 (10%) (1)     GPS 36000     Range 2 700.0m km    Resolution 100
Thermal Sensor TH2-150 (10%) (1)     Sensitivity 150     Detect Sig Strength 1000:  150m km
EM Detection Sensor EM2-150 (10%) (1)     Sensitivity 150     Detect Sig Strength 1000:  150m km

Missile to hit chances are vs targets moving at 3000 km/s, 5000 km/s and 10,000 km/s

This design is classed as a Military Vessel for maintenance purposes

Code: [Select]

test 2 class Cruiser    95 900 tons     2816 Crew     139939.5 BP      TCS 1918  TH 10000  EM 90000
5213 km/s     Armour 10-186     Shields 3000-300     Sensors 150/150/0/0     Damage Control Rating 100     PPV 936.8
Maint Life 4.05 Years     MSP 93201    AFR 735%    IFR 10.2%    1YR 9089    5YR 136339    Max Repair 7875 MSP
Intended Deployment Time: 12 months    Flight Crew Berths 1   
Hangar Deck Capacity 4000 tons     Magazine 3760   

5000 EP Photonic Drive (2)    Power 5000    Fuel Use 5%    Signature 5000    Exp 10%
Fuel Capacity 2 000 000 Litres    Range 75.1 billion km   (166 days at full power)
Omega R300/360 Shields (200)   Total Fuel Cost  3 000 Litres per hour  (72 000 per day)

Triple 40cm C6.25 FGR Laser Turret (10x3)    Range 1 400 000km     TS: 25000 km/s     Power 126-19     RM 12    ROF 35        42 42 42 42 42 42 42 42 42 42
Twin 15cm C6.25 FGR Laser Turret (20x2)    Range 720 000km     TS: 100000 km/s     Power 12-12     RM 12    ROF 5        6 6 6 6 6 6 6 6 6 6
CIWS-1000 (10x16)    Range 1000 km     TS: 100000 km/s     ROF 5       Base 50% To Hit
Fire Control S02 700-25000 H10 (2)    Max Range: 1 400 000 km   TS: 25000 km/s     99 99 98 97 96 96 95 94 94 93
Fire Control S02 175-100000 H10 (2)    Max Range: 350 000 km   TS: 100000 km/s     97 94 91 89 86 83 80 77 74 71
Vacuum Energy Power Plant Technology PO-40 (12)     Total Power Output 480    Armour 0    Exp 5%

Size 4 Missile Launcher (50% Reduction) (200)    Missile Size 4    Rate of Fire 50
Missile Fire Control FC405-R1 (10%) (2)     Range 405.0m km    Resolution 1
S4 ASM A (940)  Speed: 180 000 km/s   End: 20.3m    Range: 219.8m km   WH: 49    Size: 4    TH: 600/360/180

Active Search Sensor MR2700-R100 (10%) (2)     GPS 36000     Range 2 700.0m km    Resolution 100
Active Search Sensor MR270-R1 (10%) (2)     GPS 360     Range 270.0m km    MCR 29.4m km    Resolution 1
Thermal Sensor TH2-150 (10%) (2)     Sensitivity 150     Detect Sig Strength 1000:  150m km
EM Detection Sensor EM2-150 (10%) (2)     Sensitivity 150     Detect Sig Strength 1000:  150m km

Missile to hit chances are vs targets moving at 3000 km/s, 5000 km/s and 10,000 km/s

This design is classed as a Military Vessel for maintenance purposesLooking at these two, you will see not much is different. What you can't see from this however is that there were more efficiencies when I doubled the components. The first required 26.2 pieces of  armor while the second only 41.5. The second one had less than double crew spaces for the same deployment length. I actually needed less than double the reactors to power the weapons, but I kept those doubled anyway (was most likely due to the sheer power output than any efficiencies though). And finally, build cost (and amount of resources) did not double, they increased yes but didn't double. Also, the second design is actually likely to build faster than the first due to the shipyard's largeness efficiency.

[Charlie Beeler]

Yes,@0111narwhalz already reminded of that. n small craft with similar volume(power 3) to a big one will have a 1:n sqrt 3. So when you calculate surface (power 2) the relation of  n small crafts with r compared to big craft with r*nsqrt3 will be the formula I already given above (that is if I din't messed it up in my head). Which support that idea regardless of the coefficient. Anyway thanks for digging up the formulas for the engines, I can make use of them! If someone can dig up the rest for derived stats we can even put this to rest with a pretty graphic.

The formulae that I posted are what Steve is using in Aurora currently, not theory or assumption.  I've either gathered them from Steve's change posts or direct correspondence with him.  They are presented here because it is better to discuss what the game code is actually doing vs what someone assumes it might do. 

[Mor (OP)]

I agree 100%, its much easier to identify all the extreme cases with all formulas in hand, but the engine stats alone does't provide a complete picture e.g. you require additional crew.

[sloanjh]

2. That armor requirements grow exponentially (surface of a sphere is power 2) which means that with a sufficiently large x  it will exceed linear growth.

Two things:

First, a mathematics nit to pick:  Armor requirements grow exponentially as a function of size " (added implied function dependency) is a mathematically incorrect statement.  Growing exponentially means that the variable that is changing is in the exponent, e.g. e^x.  Total armor mass to get the same depth of armor grows like radius^2 = volume^2/3 = mass^2/3 (since volume is proportional to mass in Aurora).  This is polynomial growth, which means the exponent is a constant as a function of the variable.  Linear growth is mass^1.  Since 1 is bigger than 2/3, this means that the percentage of the ship's mass required to get the same depth of armor goes like mass^(-1/3), which decreases as mass grows (a ship that is 8 times as big in volume/mass (radius twice as big) only needs 4 times as much mass to get the same depth of armor (1/2 as much as linear growth), so it can devote more internal mass to mission systems.  Or it can apply the same bonus mass to extra armor, resulting in 8 times as much armor (same percentage) but more rows deep.

Second, the reason I'm making a big deal of this (besides lobbying for correct usage of the concept of exponential growth) is that this (bigger ships are more capable than an equal mass of smaller ships) has been a primary design goal by Steve since Day 1 of Aurora.  The reason goes back to swarm fleets in StarFire.  In StarFire, a common tactic was to design hordes of small ships (corvettes) that would overwhelm an equal mass/cost of large ships.  Steve wanted to design the game mechanics to counteract this tendency, so he consciously set things up so that large ship's armor scaled slower than linearly, while pretty much every thing else scaled linearly.  The only thing that I can recollect where small ships had an advantage has already been mentioned upthread; that very small ships didn't require a bridge so you got a mass savings there.  For everything else, I'm pretty sure that if you glue two identical ships together (including armor mass), you end up with one ship that goes at the same speed and has twice as much as everything.  You can then get savings around the margins - only need one bridge, plus with the new engine rules you can replace e.g. 2x25 engines with 1x50.

So the "complete" picture is that the game mechanics are explicitly set up 1) to mostly be linear and 2) to make large ships (in the limit of large ships) more efficient than small ships - this can be seen in Charlie's numbers.  The major effect of the size-50 limit is that this efficiency growth is less pronounced (because you can no longer make bigger, more efficient engines), which takes you back to the old days where engine efficiency didn't depend on size.  (Note that I wrote a ton of doctrine posts in those days talking about how this effect of everything being linear heavily favored single-role ships for things like survey and jump engines.)

John

[TheDeadlyShoe]

You shouldn't say that's the only advantage, when detection rules are a thing.

[Mor (OP)]

this means that the percentage of the ship's mass required to get the same depth of armor goes like mass^(-1/3), which decreases as mass grows

yes. You are the third(?) person to point it out after I already corrected it.

this (bigger ships are more capable than an equal mass of smaller ships) has been a primary design goal by Steve since Day 1 of Aurora.  The reason goes back to swarm fleets in StarFire.  In StarFire, a common tactic was to design hordes of small ships (corvettes) that would overwhelm an equal mass/cost of large ships.  Steve wanted to design the game mechanics to counteract this tendency, so he consciously set things up so that large ship's armor scaled slower than linearly, while pretty much every thing else scaled linearly.

Offering meaningful yet balanced choices has been part of game design even before the AI took over the job of GM. The issue of quality vs quantity is something that that every strategy game had to address. A basic example in Aurora would be the choice between more crew or higher grade crew.

This is usually addressed by giving each play-style a specific advantage\disadvantage, or adding diminishing returns in more sandboxy games. Besides keeping things interesting its much easier to keep the AI competitive this way.

Anyway since that post, I have reviewed other mechanics and I agree with you. That other than the connivance/performance factor of handling huge fleets, most things scale linearly, except armor that scaled slower than linearly. Which is why made this suggestion , that boils down to refit cost scaling up higher than linearly.

[83athom]

Offering meaningful yet balanced choices has been part of game design before the AI took over the job of GM. The issue of quality vs quantity is something that that every strategy game had to address. A basic example in Aurora would be the choice between more crew or higher grade crew.

I go back to my example of SC FAF, by the time one team has a small army (a dozen or so) of bots and defenses up (possibly T2), the other usually has a tsunami of hundreds of T1 bots that roll right over the first team and can somehow build 10 every second or so.

This is usually addressed by giving each play-style a specific advantage\disadvantage, or adding diminishing returns in more sandbox games. Besides keeping things interesting its much easier to keep the AI competitive this way.

I quite like this idea. Maybe at the game start (or races screen) you can set your own races affinity that give slight buffs/nerfs to various things like research (specific branches), crew efficiency/automation (ships would need +/- a % of total requirements), building, etc. And NPRs would have it randomly generated for them.