[1.12.x] Kerbal Atomics: fancy nuclear engines! (August 18, 2024)

[Angelo Kerman]

Pretty cool that you're making a NSWR. For DSEV I considered a similar engine, and called it the Nuclear Salted Fizzy Water (NSFW).  I dropped it in favor of the Supernova though.

And cool idea about radioactivity! I can see your space program taking reputation hits for irradiating kerbals.

[Temeter]

1 hour ago, iZim said:

IRL, of course. But in game it means you need 4X more tank volume, so that big atomic cruisers i build(150 tons and more) will look like a nonsense, because of enormous size of a tanks, and will be a very problematic to launch. So using LF is the only option for me.

You mean it'll look awesome. Why would you want your atomic cruise to be small? It should block out the sun!

Edited April 13, 2016 by Temeter

[iZim]

20 hours ago, Temeter said:

You mean it'll look awesome. Why would you want your atomic cruise to be small? It should block out the sun!

Of course they're look awesome. Especially a liberator with big ronund tank. But in a middle of career i have a big problems delivering shiltoads of 2,5 m cryogenic tanks to orbit for LV-N propulsed interplanetary ship. Look at Constellation Mars vehicle - tanks are not THAT big.

Pic - size comparison of the same amounts of LF and LH2

Edited April 14, 2016 by iZim

[Temeter]

12 hours ago, iZim said:

Of course they're look awesome. Especially a liberator with big ronund tank. But in a middle of career i have a big problems delivering shiltoads of 2,5 m cryogenic tanks to orbit for LV-N propulsed interplanetary ship. Look at Constellation Mars vehicle - tanks are not THAT big.

Pic - size comparison of the same amounts of LF and LH2

They are that large in reality. Can't remember the exact numbers, but LH2 is like 6+ times the volume of Kerosene.

Edit: The numbers on wikipedia might even indicate 10 times the volume? Anyway, that stuff is huge. Just look at Saturn 5: The second stage is 2/3rd the size, but the lower stage is almost 5 times heavier (LH2 stages ofc contain the denser LOX too).

Edited April 14, 2016 by Temeter

[iZim]

On 14.04.2016 at 1:03 PM, iZim said:

13 hours ago, Temeter said:

They are that large in reality. Can't remember the exact numbers, but LH2 is like 6+ times the volume of Kerosene.

Edit: The numbers on wikipedia might even indicate 10 times the volume? Anyway, that stuff is huge. Just look at Saturn 5: The second stage is 2/3rd the size, but the lower stage is almost 5 times heavier (LH2 stages ofc contain the denser LOX too).

 

You are right, comrade, LH2 is few times less denser than kerosene. Russian wikipedia says that LH2/OX tanks 2-3 times taller than one with K/OX mixture, at same diameter. But still there is some mistake in calculations(maybe a feature of a stock tanks). Here is NERVA spacecraft designs.

Edited April 15, 2016 by iZim

[Temeter]

1 hour ago, iZim said:

You are right, comrade, LH2 is few times less denser than kerosene. Russian wikipedia says that LH2/OX tanks 2-3 times taller than one with K/OX mixture, at same diameter. But still there is some mistake in calculations(maybe a feature of a stock tanks). Here is NERVA spacecraft designs.

What do you base that idea on, tho? Gotta give some numbers, not just pictures.

Those H2 tanks in the pic are pretty large, looks like ~9m diameter tanks?

[aniron]

On 15/04/2016 at 0:46 PM, iZim said:

You are right, comrade, LH2 is few times less denser than kerosene. Russian wikipedia says that LH2/OX tanks 2-3 times taller than one with K/OX mixture, at same diameter. But still there is some mistake in calculations(maybe a feature of a stock tanks). Here is NERVA spacecraft designs.

You realize that the Copernicus was meant to launch in pieces on three separate boosters, in 10m diameter by 30m length fairings, right? And accompanied by two unmanned cargo vehicles sent a few years ahead carrying the hab/lander and MAV/ISRU modules?

The mission would require 7 separate heavy cargo booster launches not including the final crew-transferring launch.

[Nertea]

This is a test release for 1.1. It should not be considered fully operational.

I will not push the AVC file or the official update links in the OP until post-official KSP 1.1.

• KSP 1.1 repack
• Updated CRP version
• Updated bundled MM version
• Deprecated IFS in favour of B9PartSwitch
• Added part search tags for parts
• In basic mode, trimodal NTRs no longer generate power when firing
• Fixed missing AFTER tag in LiquidFuel extra patch

https://github.com/ChrisAdderley/KerbalAtomics/releases/tag/v0.2.0

 

[Temeter]

16 minutes ago, Nertea said:

This is a test release for 1.1. It should not be considered fully operational.

I will not push the AVC file or the official update links in the OP until post-official KSP 1.1.

• KSP 1.1 repack
• Updated CRP version
• Updated bundled MM version
• Deprecated IFS in favour of B9PartSwitch
• Added part search tags for parts
• In basic mode, trimodal NTRs no longer generate power when firing
• Fixed missing AFTER tag in LiquidFuel extra patch

https://github.com/ChrisAdderley/KerbalAtomics/releases/tag/v0.2.0

 

^Thank you, really missed Kerbal atomics (and cryo engines)! Really expand the games formula in an awesome way! <3 <3

[Nertea]

Just an FYI, I did not bundle the correct version of B9PartSwitch, the one that ended up in the zip is only good for pre-1209 releases. I recommend not downloading this until I fix it tonight. 

[Fraz86]

9 hours ago, Nertea said:

Just an FYI, I did not bundle the correct version of B9PartSwitch, the one that ended up in the zip is only good for pre-1209 releases. I recommend not downloading this until I fix it tonight. 

Looking at CryoTanksFuelTankTypes.cfg, I believe there are a few other issues to address:

• Oxidizer's tankMass should be 0.000625
• Based on the values in B9PartSwitch.cfg, I think tankCost is meant to be the dry cost per unit volume (I imagine it adds the fuel cost automatically), whereas LH2 and Oxidizer's tankCost is currently set equal to the fuel cost per unit volume
• I believe we need two more tank types (LH2Cryo and LH2OCryo) in order to give ZBO tanks their higher mass and cost; tankMass should be 0.0003125 for LH2Cryo and 0.0004375 for LH2OCryo

Edit - It should also be noted that B9PartSwitch.cfg appears to contain errors/misinformation, and thus should not be relied on as the basis for setting your variables. For example, based on stock values, rocket mass should be 0.000625, and spaceplane mass should be 0.0007125, in disagreement with the comment section at the top of the config:

// Structural values (should be set by user)
// Rocket
//   Mass = 0.0005
//   Cost = 0.25
// Spaceplane (re-entry shielded, maxTemp > 2000)
//   Mass = 0.000625
//   Cost = 0.375

 

Edited April 20, 2016 by Fraz86

[Nertea]

I'll have a look tomorrow. I probably made a mistake. Well, that's why it's a prerelease!

[blowfish]

5 hours ago, Fraz86 said:

Looking at CryoTanksFuelTankTypes.cfg, I believe there are a few other issues to address:

• Oxidizer's tankMass should be 0.000625
• Based on the values in B9PartSwitch.cfg, I think tankCost is meant to be the dry cost per unit volume (I imagine it adds the fuel cost automatically), whereas LH2 and Oxidizer's tankCost is currently set equal to the fuel cost per unit volume
• I believe we need two more tank types (LH2Cryo and LH2OCryo) in order to give ZBO tanks their higher mass and cost; tankMass should be 0.0003125 for LH2Cryo and 0.0004375 for LH2OCryo

Edit - It should also be noted that B9PartSwitch.cfg appears to contain errors/misinformation, and thus should not be relied on as the basis for setting your variables. For example, based on stock values, rocket mass should be 0.000625, and spaceplane mass should be 0.0007125, in disagreement with the comment section at the top of the config:

// Structural values (should be set by user)
// Rocket
//   Mass = 0.0005
//   Cost = 0.25
// Spaceplane (re-entry shielded, maxTemp > 2000)
//   Mass = 0.000625
//   Cost = 0.375

 

Those values are the amount of base mass and cost the part should have (per volume) in addition to what the tank adds.  But they were mostly just there for for my own reference (and I never bothered to remove them).  And they're very approximate values based on stock anyway.

But still @Nertea I think your configs might be making the tanks lighter than they should be, since you're using the provided tank types but then subtracting the base part mass.  I'm not going to say that you can't include the structural mass in tank types, but you should probably make your own set of tank definitions if you want to do things that way.

[Fraz86]

7 hours ago, blowfish said:

Those values are the amount of base mass and cost the part should have (per volume) in addition to what the tank adds.  But they were mostly just there for for my own reference (and I never bothered to remove them).  And they're very approximate values based on stock anyway.

But still @Nertea I think your configs might be making the tanks lighter than they should be, since you're using the provided tank types but then subtracting the base part mass.  I'm not going to say that you can't include the structural mass in tank types, but you should probably make your own set of tank definitions if you want to do things that way.

Ah, that makes a bit more sense. But then, shouldn't tankMass = 0 for LFO, LF, and OX? If the mass specified in the part config is left untouched, then shouldn't stock tanks already have the correct mass for those fuel types?

I do think this approach makes more sense overall, as opposed to zeroing out the part config mass. Nertea, I'd be happy to provide corrected tankMass and tankCost values for this implementation.

[Nertea]

4 minutes ago, Fraz86 said:

Ah, that makes a bit more sense. But then, shouldn't tankMass = 0 for LFO, LF, and OX? If the mass specified in the part config is left untouched, then shouldn't stock tanks already have the correct mass for those fuel types?

I do think this approach makes more sense overall, as opposed to zeroing out the part config mass. Nertea, I'd be happy to provide corrected tankMass and tankCost values for this implementation.

Yeah, now I'm a bit confused. If realDryMass = cfgMass + volume*tankMass + massOffset, then tankMass should be 0 for stock tank types, or else they will be too heavy. 

I think that zeroing the mass does make sense from a consistency point of view. I mean, for pretty much everything derived from stock, there is a fixed dry mass per unit, right, so any fuel tank mass can be given by realDryMass = numUnits*tankMassPerUnit. There is no offset parameter for a basic tank. It makes sense to me to make tankMassPerUnit the same as tankMass, so in B9PartSwitch terms, I want realDryMass = cfgMass + volume*tankMass + massOffset. In most cases I want massOffset to cancel out cfgMass, because the base "structural" cost of the tank is part of the tankMassPerUnit variable, and thus tankMass.

[blowfish]

@Nertea the idea is that the part's base mass should be recalculated to only include the "structural" mass.

[Fraz86]

1 hour ago, Nertea said:

Yeah, now I'm a bit confused. If realDryMass = cfgMass + volume*tankMass + massOffset, then tankMass should be 0 for stock tank types, or else they will be too heavy. 

I think that zeroing the mass does make sense from a consistency point of view. I mean, for pretty much everything derived from stock, there is a fixed dry mass per unit, right, so any fuel tank mass can be given by realDryMass = numUnits*tankMassPerUnit. There is no offset parameter for a basic tank. It makes sense to me to make tankMassPerUnit the same as tankMass, so in B9PartSwitch terms, I want realDryMass = cfgMass + volume*tankMass + massOffset. In most cases I want massOffset to cancel out cfgMass, because the base "structural" cost of the tank is part of the tankMassPerUnit variable, and thus tankMass.

Actually, there are important exceptions to the standard tankMassPerUnit (0.000625), namely, all spaceplane parts (0.0007125), multi-purpose modules (with fuel + X, e.g., batteries or MP), and your truss tanks. Granted, we could find a way to zero out config mass and still account for these variances, but I think it would be a more robust and simpler implementation to leave config mass untouched. It would also allow for better consistency with tankCost calculations, where we definitely can't zero out the config costs because they scale non-linearly.

Here are the tankMass values for this approach:

LFO, LF, OX, and LH2Cryo = 0 (these configurations should use the unmodified config mass)

LH2 = -0.000375

LH2O = -0.000225

LH2OCryo = 0.000125

With these values, there would be no need for any massOffset or modification of existing config mass values.

Edited April 20, 2016 by Fraz86

[Nertea]

Well, as long as the right numbers are arrived at, I don't particularly care how they're done. What are cost numbers in this method?

[Fraz86]

22 hours ago, Nertea said:

Well, as long as the right numbers are arrived at, I don't particularly care how they're done. What are cost numbers in this method?

Well, because of the silly way that KSP calculates cost (where dry cost = config cost - fuel cost), I would need to experiment with the fuel switcher to see how exactly it interacts with those calculations. If tankCost is modifying the total cost, including fuel, akin to the config cost, I think the following should be right:

LFO, LH2Cryo = 0 (default configurations; no modification of config cost)

LF = 0.341

OX = -0.279

LH2 = -0.0915

LH2O = -0.1665

LH2OCryo = [deleted - see below]

EDIT - It turns out there isn't a way to accurately calculate LH2OCryo's cost using the tankCost variable alone, due to non-linear scaling. Instead, we need both of the following:

costOffset = -0.2 * config cost

tankCost = -0.1116

EDIT 2 - If the part switcher doesn't have a costOffset variable, or if the use of this variable is deemed undesirable, I came up with an alternate solution:

Modify the part configs for all ZBO tanks, cutting their config costs in half (matching the costs of their stock analogues). Then, set tankCost at 1.0 for LH2Cryo, and 0.4884 for LH2OCryo. The resulting costs would scale a bit more linearly than other fuel tanks, but it would be reasonably similar.

Edited April 21, 2016 by Fraz86

[Fraz86]

23 hours ago, blowfish said:

@Nertea the idea is that the part's base mass should be recalculated to only include the "structural" mass.

Hmm, I get the idea - that, theoretically, dry mass comprises both a fixed "structural mass" and a variable "tank mass," but it strikes me as needlessly complicated. Config mass obviously already includes both, and we can easily leave the implied "structural" element constant by simply making adjustments based on differences in tank mass alone (with no need to modify the config mass). See my above numbers for examples. I believe that this approach is easier, more robust, and more precise in reproducing stock masses and costs.

[blowfish]

25 minutes ago, Fraz86 said:

Hmm, I get the idea - that, theoretically, dry mass comprises both a fixed "structural mass" and a variable "tank mass," but it strikes me as needlessly complicated. Config mass obviously already includes both, and we can easily leave the implied "structural" element constant by simply making adjustments based on differences in tank mass alone (with no need to modify the config mass). See my above numbers for examples. I believe that this approach is easier, more robust, and more precise in reproducing stock masses and costs.

You can use whatever approach you want, I was just trying to explain how I designed the default tank types.  They're based heavily upon B9's parts, for which there's usually a default tank-less subtype.  No one using B9PartSwitch is under any obligation to use this paradigm or use the default tank types.

E: However, I could probably come up with some MM to calculate the correct "structural" mass based on the default mass, tank type mass, and volume pretty easily.

Edited April 21, 2016 by blowfish

[Fraz86]

41 minutes ago, blowfish said:

You can use whatever approach you want, I was just trying to explain how I designed the default tank types.  They're based heavily upon B9's parts, for which there's usually a default tank-less subtype.  No one using B9PartSwitch is under any obligation to use this paradigm or use the default tank types.

E: However, I could probably come up with some MM to calculate the correct "structural" mass based on the default mass, tank type mass, and volume pretty easily.

Fair enough. Your approach makes sense if you're starting with structural parts rather than fuel tanks.

[sojourner]

Hey, just did a clean install of KSP 1.1. So far the only mods I've installed are the latest dev build of mechjeb and now the 0.2.0 release of Kerbal Atomics.  My issue, and this may be a total noob/airhead thing on my part, is that all of my fuel tanks no longer have any fuel either listed in the description or on the pad.  This includes the tanks that come with the standard game and the mod.  I've tried every combination of keys in the VAB while clicking on a placed tank to see if I could add fuel to it, but I don't even get a dialog box for the part.  What am I missing?

[Nertea]

1 hour ago, sojourner said:

Hey, just did a clean install of KSP 1.1. So far the only mods I've installed are the latest dev build of mechjeb and now the 0.2.0 release of Kerbal Atomics.  My issue, and this may be a total noob/airhead thing on my part, is that all of my fuel tanks no longer have any fuel either listed in the description or on the pad.  This includes the tanks that come with the standard game and the mod.  I've tried every combination of keys in the VAB while clicking on a placed tank to see if I could add fuel to it, but I don't even get a dialog box for the part.  What am I missing?

There's a bit of a bug that breaks everything in this and CryoEngines. It'll be fixes later tonight or early tomorrow.

 

On 4/20/2016 at 11:18 AM, Fraz86 said:

Well, because of the silly way that KSP calculates cost (where dry cost = config cost - fuel cost), I would need to experiment with the fuel switcher to see how exactly it interacts with those calculations. If tankCost is modifying the total cost, including fuel, akin to the config cost, I think the following should be right:

LFO, LH2Cryo = 0 (default configurations; no modification of config cost)

LF = 0.341

OX = -0.279

LH2 = -0.0915

LH2O = -0.1665

LH2OCryo = [deleted - see below]

EDIT - It turns out there isn't a way to accurately calculate LH2OCryo's cost using the tankCost variable alone, due to non-linear scaling. Instead, we need both of the following:

costOffset = -0.2 * config cost

tankCost = -0.1116

EDIT 2 - If the part switcher doesn't have a costOffset variable, or if the use of this variable is deemed undesirable, I came up with an alternate solution:

Modify the part configs for all ZBO tanks, cutting their config costs in half (matching the costs of their stock analogues). Then, set tankCost at 1.0 for LH2Cryo, and 0.4884 for LH2OCryo. The resulting costs would scale a bit more linearly than other fuel tanks, but it would be reasonably similar.

I'll add this in - I tested the masses and they're working fine, didn't check the costs but I will before I release. 

[sojourner]

Ah, thanks for the quick reply. I'm just happy to find out I'm not doing something wrong.