Jump to content

Stock tank balance


Recommended Posts

This is a question is a shot in the dark, I would be surprised if anything comes from it.

I have a question for the devs or anyone who's figured out how Squad came up with the fuel numbers for the stock tanks.  Has anyone created a formula or found a way to get 'stockalike' tank balance using volume?

I'm helping @linuxgurugamer out on a rebalancing project and we are trying to balance around the stock tanks.  I've calculated the volumes of most stock tanks plus the mod we are working on.  I can see some linear efficiency in stock tanks as they grow in size but its not consistent with any other tank sizes.  Some stock tanks are way under powered, a few are way over powered.

A suggestion to the devs, you should look at your part volumes versus what your putting into them.  The fuel densities are all over the place. 

 

Link to comment
Share on other sites

As far as I know, Squad balanced the tanks by their mass fractions. LFO tanks are something like 89% fuel, Monoprop and Xenon somewhat less (I think Xenon is near 56%). In some cases the tank capacity does not match the volume for aesthetic purposes (probably the radial tanks, or maybe the toroidal), but the mass should be right. 

They pick a size, pick a capacity, then calculate the mass from there. Within limits, the square-cube law should give larger tanks better mass fractions, but I don't think that much detail is needed outside of realism overhaul. 

Link to comment
Share on other sites

Everything outside of the LFO tanks is a mess and a perfect example of squads neglect of the games art attests and game balance, but generally the consensus among non-realism users is that all tanks should use LFO's 9:1 fuel mass ratio because the average user wouldn't notice the difference in fuel mass fractions and thier impact on dv and would assume anything that doesn't come up how they expect must be a bug and not realistic fuel tank design.

As for fuel volume ratios since squad has no consistency outside of most LFO tanks you are basically free to enforce your own standards if consistency is your aim. For example for the monoprop tanks in MoarMk1 I did +50% more units for a given tank volume (so if a Ft-400 hold 400 units of lfo then a monoprop tank of the same dimensions would hold 600 units.)

Edited by passinglurker
Link to comment
Share on other sites

16 minutes ago, Capt. Hunt said:

I always figured that the different types of tanks just had different amounts of internal structure for support.  They aren't just aluminum cans after all.

It would make sense but its not what ksp does. Cubed square law would dictate that even with structural supports the bigger you go the better mass ratio you get but lfo tanks are generally a flat 9:1 across the board (aside from oddballs like the mk2 tanks) it would make sense to be consistent about this because it's more intuitive to the player to just compare the engine stats alone. Not that it would be bad if squad took the engineering approach to this and reflected mass ratios accurately it would be a good learning experience for many players, but unfortunately when you see differing mass ratios in the game it's not because of some grand design, but because the part was thrown in as a place holder to be refined later and never was. So mass fractions become less educational and more "how can I best cheese the part catalog for internet street cred?"

As a result monoprop tanks are all over the place with no rhyme or reason in almost every aspect, and xenon tanks are so terrible the ion engine needs massively inflated stats to compensate and even then still loses out to an ant engine stack of equal or lesser mass in all but the most dizzying 10min+ burn time missions.

and that's not even getting into the volume ratios can of worms

Link to comment
Share on other sites

Xenon tanks are made of cast iron because of the low delta-V requirements of the stock game in general. For the delta-V it takes to get into orbit in RSS (~9.5 km/s assuming average piloting) I could launch from Kerbin (3.5 km/s), reach and land on the Mun (1.5 km/s) then return to Kerbin (1 km/s), then get BACK to Kerbin orbit (3.5 km/s). For the amount it takes to reach the Moon in RSS (~13 km/s) I could do the same for Duna, or get to Jool/Moho, land on something and return. 

If Xenon tanks used 9 to 1, you could basically torch-ship to anywhere in the stock system. 

Monoprop tanks are meant to be pressure vessels (type "Servicemodule" in Realism Overhaul, so are always going to be heavier than your standard Aerozine/NTO tanks connected to a pump-fed engine, which is the closest approximation to what Squad has in stock. I have no problem with them being heavier than LF/O tanks - maybe a little lighter than they are now, but not by much. 

Considering that Squad intentionally lets you clip parts together, I don't think volume ratios are high on their priority list. I would much rather see game engine, memory usage and physics optimizations - parts are easy in comparison and are already handled by dozens of talented mod developers. 

Link to comment
Share on other sites

33 minutes ago, MaxL_1023 said:

Xenon tanks are made of cast iron because of the low delta-V requirements of the stock game in general. For the delta-V it takes to get into orbit in RSS (~9.5 km/s assuming average piloting) I could launch from Kerbin (3.5 km/s), reach and land on the Mun (1.5 km/s) then return to Kerbin (1 km/s), then get BACK to Kerbin orbit (3.5 km/s). For the amount it takes to reach the Moon in RSS (~13 km/s) I could do the same for Duna, or get to Jool/Moho, land on something and return. 

If Xenon tanks used 9 to 1, you could basically torch-ship to anywhere in the stock system.

You make the mistake of assuming the tanks design is intentional. Kerbal has never had a proper balance pass its part stats are more fractured than its art styles making excuse for kerbals oddies just doesn't work beyond your own personal play.

Also if xenon tanks used 9:1 you could bring down dawns stats to closer to realistic levels while still keeping it balanced and practical to play with except now it can be compact enough and cheap enough to actually be worth more than a gimmick @Streetwind (the guy who balances near future propulsion) worked this all out just before 1.2.
 

33 minutes ago, MaxL_1023 said:

Monoprop tanks are meant to be pressure vessels (type "Servicemodule" in Realism Overhaul, so are always going to be heavier than your standard Aerozine/NTO tanks connected to a pump-fed engine, which is the closest approximation to what Squad has in stock. I have no problem with them being heavier than LF/O tanks - maybe a little lighter than they are now, but not by much. 

Where have they said anything like this ever? Squad gives no insight into its part design and balance process and making baseless excuses helps no one but yourself in feeling better about this games sorry state.

Anyway if they were assuming all monoprop tanks are for pressure fed engines then all monoprop type tanks would still be consistent between each other which they frankly are not even in the slightest. You're just making excuses for squads negligence of production quality and you should stop.
 

33 minutes ago, MaxL_1023 said:

Considering that Squad intentionally lets you clip parts together, I don't think volume ratios are high on their priority list. I would much rather see game engine, memory usage and physics optimizations - parts are easy in comparison and are already handled by dozens of talented mod developers. 

Ah the old "I'm just here to sink the discussion so it doesn't get in the way of my own agenda" argument. two things as you make your exit. First not everyone on staff is a coder who can work on optimizing the code base some one has to do balancing and playtesting. Second modders shouldn't have to clean up a developers mess a modding community is not a crutch to be leaned on and when you try it'll suddenly buckle out from under you as the modders leave and try and find a developer that values them instead of just uses them.

Edited by passinglurker
Link to comment
Share on other sites

5 hours ago, passinglurker said:

but lfo tanks are generally a flat 9:1 across the board

Really?

When you take the total fuel (Lq + Ox) and divide it by the mass of the part, for most of the Squad tanks, you end up with numbers somewhere in the 6.98 and 10.0 (ignoring the oddball like the NCSadapter and Twin Boar) range, not 9.  

Monoprop is even wider, going from 3.2 to 7.5 

Liquid fuel and Ox both have the same weight per unit (0.005), while Xenon's weight/unit is 0.0001 and Monoprop is 0.004

 

KW Rocketry, on the other hand, is much more uniform, where most of the Lq/Ox ratios being in the range of  9.01-9.13 (there is one which is way off, looks like a mistake, it's 18.13, not sure if the mass or volume is off)

 

And frankly, I'm leaning very heavily in the direction of having the mass of the tank be related to the amount of surface area, which would make larger tanks more efficient (as in real life).

Link to comment
Share on other sites

1 hour ago, linuxgurugamer said:

Really?

When you take the total fuel (Lq + Ox) and divide it by the mass of the part, for most of the Squad tanks, you end up with numbers somewhere in the 6.98 and 10.0 (ignoring the oddball like the NCSadapter and Twin Boar) range, not 9.

9:1 mass ratio means for every 9 tons of gross mass 1 of those tons is dry mass so when you divide your fuel mass by your dry mass it should come out as 8. This figure is consistent across the main conventional fuel tanks (jebs junkyard, rockomax, and kerbodyne) odd shapes like space planes have a different ratio and its hard to say if any intelligent design went into finding thier figures or if they just went "as long as it's worse than rockets we're good"

1 hour ago, linuxgurugamer said:

Monoprop is even wider, going from 3.2 to 7.5

Yup its a mess and the ratios don't even cleanly scale up, but instead fluctuate up and down never mind the volume issues. It's a wild west so you can either do what you want, or you can raise the issue on the bug tracker and maybe squad will fix it for 5 minutes before reverting it again like last time cause it breaks a tutorial they can't be bothered to polish.

1 hour ago, linuxgurugamer said:

Liquid fuel and Ox both have the same weight per unit (0.005), while Xenon's weight/unit is 0.0001 and Monoprop is 0.004

Yeah I made sure to account for that in my own consistent mk1 monoprop tanks.

1 hour ago, linuxgurugamer said:

And frankly, I'm leaning very heavily in the direction of having the mass of the tank be related to the amount of surface area, which would make larger tanks more efficient (as in real life).

It would be cool to be like real life when its not overbearing (overbearing as in uliage requirements, limited restarts, light signal delay, etc... bleh...), but if @Streetwind's account is anything to go by you'd best be prepared for an influx of support requests as users don't read, don't notice, and when they do they don't realize its the mass ratio, or that mass ratios are a thing and think its a bug, but to be fair a mod like smurff seems to get by without complaint just fine probably because mass ratios is all it does so users know whats going on and what they are getting into when they download it.

Link to comment
Share on other sites

You have the definition of ratio wrong.  9:1 means that for every 9 of the first, there is 1 of the second, giving a total of 10.

If you doubt that, think what a 1:1 ratio means.  1 of the first for every 1 of the second, giving a total of 2

https://www.mathsisfun.com/definitions/ratio.html

Edited by linuxgurugamer
Link to comment
Share on other sites

Breathe, @passinglurker. It's just a game. :)

I'll take consistent mass ratios any day it's offered to me (and I'm admittedly sad that although my suggestion was initially picked up by the devs, it fell off the face of the schedule after the great exodus of 2016) but I'm not losing sleep over it, nor do I curse players or developers for some perceived shortsightedness. Overall the game works well enough to make this a "nice to have" feature rather than a "must have".

 

7 minutes ago, linuxgurugamer said:

You have the definition of ratio wrong.  9:1 means that for every 9 of the first, there is 1 of the second, giving a total of 10.

Perhaps, but then you'll have to berate the people who defined mass ratios, not those who use the definition :P

A mass ratio is 9 when 9 tons of wet mass is set against 1 ton of dry mass. Ostensibly that would be an 8:1 mathematical ratio of fuel versus not fuel. But since dry mass is also included in wet mass, it appears on both sides of the colon and therefore the thing actually ends up being (8+1):1.

 

Edited by Streetwind
Link to comment
Share on other sites

10 minutes ago, Red Iron Crown said:

Mass ratio is calculated the way it is because that's what's needed for the rocket equation. It's not fuel:dry mass, it's wet mass:dry mass. Still a ratio, just one defined a different way. 

Can you explain why?

Link to comment
Share on other sites

@linuxgurugamer:

The main reason is convenience in the calculation.

For a slightly more mathematically-inclined explanation, the reason is because the relationship of mass to delta-V given by the rocket equation uses the logarithmic function to take the endpoints of the continuous loss of mass (continuous over the expenditure of fuel, not over the passage of time) and then use those endpoints to describe the changes seen in delta-V over the course of that loss.  In the rocket equation, these endpoints are naturally defined by the mass of the rocket at the start of the burn and the mass at the end of the burn, and in the most extreme possible case, this means the mass of the rocket when full of fuel and the mass when empty of it.  However, the mass of the full rocket necessarily must include the mass of the empty rocket, so it appears on both sides of that relationship.

Your information about ratios is correct, but I respectfully suggest that your information is incomplete:  one must not confuse the colon of a ratio with an addition sign, because the colon of the ratio states only that a relationship exists--the exact nature of the relationship depends on the units of the values.  For example, in the link you gave, one example ratio given is 1:3 as a ratio of boys to girls.  Is it not equally valid to say that there is a 1:4 ratio of boys to children?  Each description has its place and its use.  One may even describe relationships that cannot add directly:  for example, after every rapid unplanned disassembly, let's say the KSC sacks ten engineers.  This implies a relationship--a ratio--of 10:1 unemployed engineers for each rocket failure, but that does not add up to eleven sacked-engineer-failed-rocket-hybrid-things, because although engineers and rocket failures are units, those units are not related by addition.  (Correction:  they don't add unless the KSC terminates employment with some gruesomely extreme prejudice.)

To that end, the 9:1 wet-to-dry ratio for fuel tanks is useful for comparing theoretical maximum delta-V of a spacecraft in a convenient, ready-to-plug-into-the-rocket-equation form.  The elegance of it is that if we take the rest of the rocket mass to be constant, we can use the mass fraction to decide at a glance whether one fuel tank is more or less efficient than another without having to go through the full calculation.  By contrast, using the 8:1 fuel-to-tank ratio requires an additional step.  It is still useful to compare tank efficiency, and one can even use it in the rocket equation (change wet mass to dry mass + fuel mass; it certainly isn't difficult), but given the fact that it is easier to get total rocket mass from KSP rather than fuel mass + part mass components, the 9:1 mass fraction is the convenient one.  Similarly, the version of the rocket equation most often used for KSP is not the one that relies on effective exhaust velocity--even though that is easy to intuit, KSP reports Isp, not Vexh, so it is more convenient to incorporate the conversion into the rocket equation and use the values that display on the screen.

Edit:  For completeness's sake, here's a version of the rocket equation that uses fuel-to-dry-mass:

ΔV = Isp * g0 * ln (1 + (mfuel / mdry))

Where:

ΔV is delta-V,
Isp is specific impulse,
g0 is acceleration due to gravity at Kerbin's surface (9.81 m/s2),
ln is the natural logarithm function,
mfuel is the fuel mass, and
mdry is the dry mass of the rocket.

Nothing prevents anyone from using the 8:1 ratio, but it has to be used correctly.

Edited by Zhetaan
Link to comment
Share on other sites

Thank you for the explanation.

For the spreadsheet we have, I need the raw numbers, not the "Mass Ratio" as just described.  This is needed in order to come up with a reasonable value for the weight of the skin.  But for calcualting thrust, etc, I now understand what you are say.

Basically, @passinglurker was talking about the Mass Ratio, while I was talking about the raw ratio.  Both valid, depending on what is being talked about

Link to comment
Share on other sites

13 hours ago, passinglurker said:

As for fuel volume ratios since squad has no consistency outside of most LFO tanks you are basically free to enforce your own standards if consistency is your aim. For example for the monoprop tanks in MoarMk1 I did +50% more units for a given tank volume (so if a Ft-400 hold 400 units of lfo then a monoprop tank of the same dimensions would hold 600 units.)

Since an offshoot of this project will be to provide a mod to rebalance the stock tanks, the tanks will be consistent across these mods:

  • Munar Industries
  • KW Rocketry
  • Stock

And, I'd be happy to share my formulas with anybody who requests.

My aim is to not nerf any existing parts, the end result will be to actually make them better and more efficient.

I do have a question for people:  One would expect that as technology advances, the mass of a given tank will be reduced as new materials are found.  So, I can approach this one of several ways:

First, come up with standardized values

Second, and here are the possibilities:

  • For the early tanks, implement part upgrades to reduce the mass of the tank.  Drawback is that each part will have to have the upgrade purchased individually
  • Ignore mass changes, and just have tanks further along the tech tree have better materials
  • Write a small mod which will adjust the mass of new tanks (ie: new in the Editor) as the tech tree advances

Comments?

Link to comment
Share on other sites

@linuxgurugamer, this seems like a perfect use for the Stock Upgrades feature, so you wouldn't have to write a new mod. I think mass of all tanks should slightly decrease with progression. Otherwise you only the larger volume tanks, late in the tech tree will benefit.

You could add buy the upgrades in tech tree nodes related to materials or production techniques. I was experimenting with something similar in my own game. The upgrade feature is somewhat limited in scope, but it should be  easy for simply changing masses with tech progression.

Link to comment
Share on other sites

1 hour ago, Nightside said:

@linuxgurugamer, this seems like a perfect use for the Stock Upgrades feature, so you wouldn't have to write a new mod. I think mass of all tanks should slightly decrease with progression. Otherwise you only the larger volume tanks, late in the tech tree will benefit.

You could add buy the upgrades in tech tree nodes related to materials or production techniques. I was experimenting with something similar in my own game. The upgrade feature is somewhat limited in scope, but it should be  easy for simply changing masses with tech progression.

The Stock upgrade feature is designed to work by purchasing (unless I read it wrong).   If it is possible to have it work automatically, please let me know.

What I would want is to have something be in the tech tree node, which, when purchased, would upgrade all the tanks which reference it by reducing the mass.

On the other hand, this may not be worth it.  It would mess around with people sharing craft files.

Link to comment
Share on other sites

11 hours ago, linuxgurugamer said:

And frankly, I'm leaning very heavily in the direction of having the mass of the tank be related to the amount of surface area, which would make larger tanks more efficient (as in real life).

There are offsetting factors, to the point at which even the RO crowd accepts procedural tanks as having mass proportional to volume.

The bigger your tank is, the more aerodynamic stresses it will endure, and you have to thicken the tank walls.

Pressurant is probably roughly proportional to volume.

There's fuel lines, there's sensors, there's wiring, a thousand other things that make the "tank mass proportional to surface area" model a poor representation of reality.

 

Link to comment
Share on other sites

4 hours ago, linuxgurugamer said:

On the other hand, this may not be worth it.  It would mess around with people sharing craft files.

Probably best to test it and how upgraded parts carry over when sharing craft files before jumping to conclusions. Not a lot is known about the upgrade system so far because the UI is a bit of a bear, but the system is getting some quality of life refinements in 1.3 so it could be wise to design with the system in mind from the start.

1 hour ago, Starman4308 said:

There are offsetting factors, to the point at which even the RO crowd accepts procedural tanks as having mass proportional to volume.

The bigger your tank is, the more aerodynamic stresses it will endure, and you have to thicken the tank walls.

Pressurant is probably roughly proportional to volume.

There's fuel lines, there's sensors, there's wiring, a thousand other things that make the "tank mass proportional to surface area" model a poor representation of reality.

If you ask me I think it's more important to make a set of rules and stick to them than it is for those rules to be an accurate reflection of real life. Both ways have merit a scaling mass ratio would give you incentive to progress through the tech tree, and a flat ratio is more intuitive being easy for new players to wrap thier heads around and at least be justifiable to veteran players.

Link to comment
Share on other sites

57 minutes ago, passinglurker said:

If you ask me I think it's more important to make a set of rules and stick to them than it is for those rules to be an accurate reflection of real life. Both ways have merit a scaling mass ratio would give you incentive to progress through the tech tree, and a flat ratio is more intuitive being easy for new players to wrap thier heads around and at least be justifiable to veteran players

I did ask, thank you.  

Link to comment
Share on other sites

12 hours ago, Streetwind said:

I just did. Dry mass apppears on both sides of the colon, not just on the right side. It's not "fuel mass against dry mass" but rather "(fuel mass plus dry mass) against dry mass".

You explained what, but not why.

Link to comment
Share on other sites

Stock tanks are not balanced, especially monoprop tanks are totally out of whack between their volumes.

As someone already said, for compatibility (procedural parts, fuel switchers, etc) only volume based factors are workable.

You can install SETIrebalance and Procedural Parts.

Then just stretch the procedural liquid fuel tank to the size of the tank you want to rebalance, select the tank content and use that value. And coincidentally, with SETIrebalance, stock tanks are balanced between each other as well. Also displays the volume/mass/units/costs and so on ratios.

 

edit: And I would stay away from the upgrade feature, as you said, it kills craft sharing. And compatibility with different tech trees.

Edited by Yemo
Link to comment
Share on other sites

This thread is quite old. Please consider starting a new thread rather than reviving this one.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...