Stock tank balance

[CarnageINC]

I've come up with a part volume based stockalike balance.  Sadly I asked this question then got caught up in my calculations and forgot to come back and check on what was going on.  I've designed a stockalike fuel balancing system that will work for any mod (I believe) so long as the volume of the part is known.  Its not a perfect catch all but it aligns very well with most stock parts when I use the formula on them. 

There are stock parts that the system can not account for, such as the "Oscar" and Round 8 tanks (which are over powered) and tanks like NCS adapter (which is greatly under powered).  Mk2 and Mk3 tank performance is also underwhelming but considering those are tanks for vehicles that perform "high stress" maneuvers, those numbers can be overlooked.  The system does take in account of 'normal' rocket tanks and 'high performance' tanks, in addition to monopropellant, Xenon and Liquid Fuel only.

 

 

Edited May 14, 2017 by CarnageINC

[Alshain]

The round 8 is certainly not overpowered.  They actually fixed that one just a few patches ago because it was so underpowered.

[linuxgurugamer]

I think you need to say what "overpowered" means.  In this case, it probably holds more fuel than it's volume would allow.  It's cute, but really a very small tank

Edited May 14, 2017 by linuxgurugamer

[Yemo]

From my experience, I strongly recommend a fixed dry mass to capacity ratio for the following reasons:

1. Gameplay/Design

The "lego principle" of being able to substitute 1 triple tank with 3 normal tanks might not be realistic, but is a concession to the design limitations when using fixed block sizes. In reality a tank would be custom made to a specific size. In KSP there are instances when the right tank size is not available. Eg If there is a normal shuttle tank and a triple shuttle tank, but not a double tank. With fixed ratios, you have a linear increase of capacity and mass for length 1, 2, 3, 4 (multiples of normal length, as 1 = 1 normal, 2 = 2 normal tanks, 3 = 1 triple tank, 4 = 1 triple tank + 1 normal tank) etc. With non-fixed ratios it would go like 1 - 2 linear, 3 drastically better fuel to mass ratios, 4 worse fuel to mass ratio than length 3, 5 (one triple, 2 normal tanks) even worse fuel to mass ratio than length 3. This would introduce efficiency mali simply based on the available block sizes and their combinations, which would imho be detrimental to design choices.

2. Mod compatibility 1st degree

Mods like TweakScale and Procedural Parts are linear. Thus if someone has eg engines for 1.875m diameter from another mod, but does not like the provided 1.875m tanks. The person has TweakScale installed to make up for such part catalogue limitations. If a 1.25m tank is upscaled to 1.875m, it would have drastically worse stats than a downscaled 2.5m tank (eg because the player likes the orange tank color from 2.5m). Same goes for round/special shape tanks when designing landers and so on.

3. Mod compatibility 2nd degree

Mods from point 2 might be used or configured for other mods, eg rescales, tech trees etc., compounding the issues from 2.

4. Other reasons

While real life rocket tanks are mostly designed around a unidirectional force vector, which drastically reduces the need for structural load considerations, many players in KSP use the tanks for other applications as well. For example space planes, or as structural parts on space stations, rovers, etc. There is currently no way to distinguish between those applications in terms of structural loads.

 

As to the actual values for a fixed ratio approach, I recommend the following:

For liquid fuel/oxygen tanks (as are nearly all stock tanks)
mass = 1 ton, liquid fuel = 720, oxygen = 880

For monoprop (no stock ratio, all over the place)
mass = 1 ton, monoprop = 2000

Based on tests with monoprop vs LOX rockets, there is a range of ratios suitable to preserve desired efficiency differences for design/gameplay considerations. The actual value within this range is arbitrary.

However between all those alternatives, the 1 ton = 2000 monoprop capacity value stands out from a design perspective, as it results in the same wet mass as a LOX tank with the same dry mass. For 95% of the players, this is probably irrelevant. But when combined with fuel switching mods/procedural parts, it seriously improves the design/redesign process for the 5% of players who are into that. As a tank content/part switch does not alter the center of mass for already designed spaceplanes and landers (where a CoM change might affect the placement of lifting surfaces on spaceplanes and RCS thrusters on landers). Different fuel options and ratios can be tested in the design stage without having to recheck/redo the parts which are placed around the existing center of mass. And when the value within the desirable range is arbitrary, why not make life easier for those 5% when it does not cost anything.

[CarnageINC]

5 hours ago, linuxgurugamer said:

I think you need to say what "overpowered" means.  In this case, it probably holds more fuel than it's volume would allow.  It's cute, but really a very small tank

linuxgurugamer is quite right.  Looking at the actual volume of the parts then seeing what Squad put in them makes some parts have a fuel density that is more than double what 'stock' cylinder tank holds.  I can see why Squad put high values on smaller tanks, this is a game after all.

@Yemo the fuel balance and soon to be fuel and dry mass balance will all be based for stockalike balance so mod compatibility should not be an issue.  When I look at the masses, I'll see how stock values overall stack up and work from there.  I'm curious to see if they are similar to your recommendations, hopefully they will be similar since it seems like you put in a lot of thought and testing.

[SilverWolf]

If it's volume and capacity you're looking for, I already did this, in 3Ds Max.