Nuclear engines useless?

[Strawberry]

5 minutes ago, jimmymcgoochie said:

Liquid hydrogen is not dense. It’s an extremely small, extremely light molecule that has a really low density and so you need a much larger volume of hydrogen to contain the same mass of hydrogen compared to methane. It doesn’t matter if it’s “unintuitive” because it’s completely realistic.

Okay I assumed that they used values for like compressed hydrogen instead of liquid hydrogen but did the math and hydrogen in game is actually more dense then normal. The small hydrogen tank has a volume of around 6 cubic meters, this means that the hydrogen is stored at a density of around 100 kg per cubic meter (100 g per l). Liquid hydrogen (at melting point) is around 70 grams per liter. I'll retract this point, stand by the rest of it.

Just now, LameLefty said:

SWERV is great! It’s only OP in the case of a pure sandbox, with all parts unlocked and usable from the get-go. In a true career-based campaign with an actual technology tree, you can be sure it’s going to be waaaay up there past the basic NERV, and require a lot of Science!(tm) to unlock for your late-game in-system bases, colonies and stations. 

Science consumption honestly shouldnt be a massive balance factor, having parts that just rot in the VAB is boring and straight upgrades make the game much less interesting and fun. While its fine for some parts to be overall stronger then others, everything should be a sidegrade to varying degrees.

[regex]

4 minutes ago, Strawberry said:

While its fine for some parts to be overall stronger then others, everything should be a sidegrade to varying degrees.

In order to support interstellar gameplay we're going to need engines that make the SWERV look like absolute trash, some balancing is going to be needed elsewhere than base stats.

[Strawberry]

1 minute ago, regex said:

In order to support interstellar gameplay we're going to need engines that make the SWERV look like absolute trash, some balancing is going to be needed elsewhere than base stats.

I definitely agree! That's why I support mechanics like heat and radiation, they add design constraints that make them not always the go to. 

[LameLefty]

5 hours ago, Strawberry said:

Okay I assumed that they used values for like compressed hydrogen instead of liquid hydrogen but did the math and hydrogen in game is actually more dense then normal. The small hydrogen tank has a volume of around 6 cubic meters, this means that the hydrogen is stored at a density of around 100 kg per cubic meter (100 g per l). Liquid hydrogen (at melting point) is around 70 grams per liter. I'll retract this point, stand by the rest of it.

Science consumption honestly shouldnt be a massive balance factor, having parts that just rot in the VAB is boring and straight upgrades make the game much less interesting and fun. While its fine for some parts to be overall stronger then others, everything should be a sidegrade to varying degrees.

LH2 is what it is. It’s already as dense as hydrogen gets except for its metallic form, which requires truly intense pressures of the kind found deep inside gas giants like Jupiter. As a deep-cryogenic liquid, the tanks provided in game are pretty okay from a mass standpoint. Big  Mylar-insulated Dewar containers like this are perfectly feasible.

As for tech tree philosophies, that’s a subject for a dedicated thread once a tech tree gets added. Lord knows there was plenty of that in KSP1. Having said that, something like a SWERV and eventually Orion nuclear pulse, MHD plasma drives, etc. are truly order-of-magnitude more capable than earlier technologies and will definitely NOT be a “side-grade.” 

Edited March 14, 2023 by LameLefty
typo

[Strawberry]

3 minutes ago, LameLefty said:

As for tech tree philosophies, that’s a subject for a dedicated thread once a tech tree gets added. Lord knows there was plenty of that in KSP1. Having said that, something like a SWERV and eventually Orion nuclear pulse, MHD plasma drives, etc. are truly order-of-magnitude more capable than earlier technologies and will definitely NOT be a “side-grade.” 

Orion drives are expensive, bulky, and imprecise, MHD plasma drives are power hogs, etc. They dont need to be perfect sidegrades and its perfectly fine for as you get along the tech tree for things to be overall stronger, but every part should maintain a niche 

[Wheehaw Kerman]

I think @Strawberry has nailed it.  Radiation and heat are going to be the constraints on SWERV.

[regex]

5 minutes ago, Strawberry said:

I definitely agree! That's why I support mechanics like heat and radiation, they add design constraints that make them not always the go to. 

We'll get heat but I wouldn't hold my breath for radiation, so we're going to need balancing elsewhere and that will likely be from progression. Unfortunately most games consider "sandbox" to be a "use everything willy-nilly" game mode rather than as a "set your own goals" mode, so extra balance considerations may not enter the picture.

[LameLefty]

1 minute ago, Wheehaw Kerman said:

I think @Strawberry has nailed it.  Radiation and heat are going to be the constraints on SWERV.

Heat yes, radiation? I doubt it. If you’ve ever played with Nertea’s mods in KSP1, I think you’ll get a very good idea of how things are likely to end up in KSP2 (especially now that he is a KSP2 dev). Heat rejection/radiator mass will cut into the “OP-ness” of the SWERV but not to the extent that it’s unusable. 

[Strawberry]

7 minutes ago, regex said:

We'll get heat but I wouldn't hold my breath for radiation, so we're going to need balancing elsewhere and that will likely be from progression. Unfortunately most games consider "sandbox" to be a "use everything willy-nilly" game mode rather than as a "set your own goals" mode, so extra balance considerations may not enter the picture.

Radiation was definitely planned at a point in development, the devs explicitly said so early on and we saw indicators of this such as shadowshields and radiation icons on the VAB. I still hope for radiation (even if Im iffy if theyll actually implement it) because the sidegrades of using compact objects versus lots of trusses is interesting and leads to really good looking crafts, and also promoting the use of lots of trusses leads to natural synergies with lots of radiators and crafts poking off of the side which is always cool. 

Edited March 14, 2023 by Strawberry

[regex]

1 minute ago, Strawberry said:

because the sidegrades of using compact objects versus lots of struts is interesting and leads to really good looking crafts

I think the idea of sidegrades is fine if we're talking pure gameplay but if you're creating a game that has ties to actual realistic/theoretical spacecraft design, some things are going to be miles better than others because that's how reality works. You would be stupid to use a gas core nuclear engine if you had access to a plasma Z-pinch fusion torch, all other considerations being equal. Yes, you'll probably need more heat rejection mechanisms but that's not a big deal because you have so much more isp to work with.

Quite frankly if you hand me a fusion engine and it ends up being a "sidegrade" to a conventional nuclear hydrogen engine purely in terms of the resulting craft I'm going to be pretty damn angry. If it costs more in terms of precision, resources, knowledge, or build time that's fine, but if it's nerfed to be balanced in a "use everything willy nilly" game mode then I would call that an extremely dumb choice on the developer's part.

[Strawberry]

13 minutes ago, regex said:

I think the idea of sidegrades is fine if we're talking pure gameplay but if you're creating a game that has ties to actual realistic/theoretical spacecraft design, some things are going to be miles better than others because that's how reality works. You would be stupid to use a gas core nuclear engine if you had access to a plasma Z-pinch fusion torch, all other considerations being equal. Yes, you'll probably need more heat rejection mechanisms but that's not a big deal because you have so much more isp to work with.

Fusion drives produce lots of neutron radiation, and tend to be very large, meaning that even for nuclear standards you cant make a small spacecraft for them. They also are almost always lower thrust then fission drives as well. They'd be good for bulk cargo and large transfers, but for "small" missions youd likely want to stick to gas core rockets. You can still have the philosophy of all engines should maintain a niche, and also still have engines be overall stronger and also be realistic on top of that.

While I dont think the game should lean too heavily on this I think one thing that will help loads with this balancing is resource as you mentioned, though the game should definitely create a demand in game for spending resources and also using your fast but not the most efficient torch drive ships, Id suggest tourism here but thats a topic for an entirely separate thread that I need to write up one day.

Edited March 14, 2023 by Strawberry

[regex]

2 minutes ago, Strawberry said:

You can still have the philosophy of all engines should maintain a niche, and also still have engines be overall stronger and also be realistic on top of that.

So long as you're not arbitrarily nerfing them for "gameplay", I agree, because that would be dumb.

[cocoscacao]

27 minutes ago, regex said:

You would be stupid to use a gas core nuclear engine if you had access to a plasma Z-pinch fusion torch

Paste this, out of context, somewhere on the web, and try to convince people that it came from a game forum 

[Mutex]

Did some experimenting and the NERV seems to fill a niche to me:

Spoiler

Left: 3053 m/s, 8.55 tonnes

Middle: 2492 m/s, 15.55 tonnes

Right: 2774 m/s, 8.77 tonnes

 

So in this case the NERV powered ship has the best dV and the lowest weight. Of course if you made the ship bigger and added more hydrogen, then using the SWERV would start to make more sense due to the higher ISP being more of a pro and the extra mass being less of a con. For smaller ships the lower weight of the terrier would probably make it the better option.

 

Edited March 14, 2023 by Mutex

[LameLefty]

If people want to know what to expect for later game content and in particular, nuclear engines, see Nertea’s longstanding (amazing) work with KSP1. In particular, I draw your attention to:

The existing nuclear engines in the game are direct descendants on his work above, and I expect the thermal control system aspects will  be as well, once worked into the game. 

 

 

[Strawberry]

15 minutes ago, Mutex said:

Did some experimenting and the NERV seems to fill a niche to me:

  Reveal hidden contents

Left: 3053 m/s, 8.55 tonnes

Middle: 2492 m/s, 15.55 tonnes

Right: 2774 m/s, 8.77 tonnes

 

If you want a ship with only 3k delta v you might as well use chemical engines. With higher delta v's the difference becomes a lot less worthwhile.

[Mutex]

23 minutes ago, Strawberry said:

If you want a ship with only 3k delta v you might as well use chemical engines. With higher delta v's the difference becomes a lot less worthwhile.

Here's three ships all with roughly 5500m/s dV (slightly weird shapes due to me trying to hit the same dV for all three ships):

Spoiler

Left: 5547 m/s, 12.07 tonnes

Middle: 5572 m/s, 20.05 tonnes

Right: 5552 m/s, 39.24 tonnes

 

If we go for much more dV then the SWERV definitely becomes the better option. But the NERV definitely fills a niche, which is all an engine needs to do, no engine should be the best in all scenarios.

EDIT: Seems it's just after 12-13 km/s where the SWERV takes over and becomes the better option, for this payload.

Edited March 14, 2023 by Mutex

[Strawberry]

5-6k range tends to be the tipping point yeah, but I tend to design nuclear ships for the 7-8k range as less then that might as well use chemical, and for that SWERV is better. Also delta v isnt everything and SWERV produces much more thrust then NERV, even in the picture you showed Id argue the SWERV is just better there as the thrust is much higher. I have tried many times to design a practical ship to where I use a NERV engine and I just cant do it

Edited March 14, 2023 by Strawberry

[FlazeTheDragon]

2 hours ago, Strawberry said:

Fusion drives produce lots of neutron radiation, and tend to be very large, meaning that even for nuclear standards you cant make a small spacecraft for them. They also are almost always lower thrust then fission drives as well. They'd be good for bulk cargo and large transfers, but for "small" missions youd likely want to stick to gas core rockets. You can still have the philosophy of all engines should maintain a niche, and also still have engines be overall stronger and also be realistic on top of that.

While I dont think the game should lean too heavily on this I think one thing that will help loads with this balancing is resource as you mentioned, though the game should definitely create a demand in game for spending resources and also using your fast but not the most efficient torch drive ships, Id suggest tourism here but thats a topic for an entirely separate thread that I need to write up one day.

I dont think the devs are overly concerned with balancing every engine via stats to be more or less a side grade considering were going to have metallic hydrogen and anti-matter engines in the future.

Realistically, it would be basically impossible to construct massive motherships with the current tech we have, so i dont think its realistic to constrain the game to a pure stat based balance.

That said, i do think the devs need to do something with the nerv, its honestly the only engine in vab that is completely useless rn. Maybe convert it to a metallic hydrogen engine in the future idk...

[LameLefty]

7 minutes ago, FlazeTheDragon said:

That said, i do think the devs need to do something with the nerv, its honestly the only engine in vab that is completely useless rn. Maybe convert it to a metallic hydrogen engine in the future idk...

Take a look at Nertea’s Kerbal Atomics mod for KSP1, which I linked above. It has a complete line of much better-balanced (and thus more usable) nuclear engines in a variety of scales suitable for outer system light probes all the way up to SWERV-type gas-core rockets, each one thoughfully balanced as to mass, ISP, and placement on the Community Tech Tree to which he was a contributor.

I think the NERV we have right now is either merely a carryover placeholder, or will later be supplemented by more advanced (lighter, better ISP) versions analogous to those from Nertea’s Kerbal Atomics. It’s also possible that by the devs are intentionally NOT going to fill out the lower end of the tech tree with better versions of existing tech (such as a sorely-missed  Centaur-equivalent LH2/LOX engine) but instead focus on the far, high-tech end of the tree with those advanced plasma and fusion torch engines focused on interstellar tech. 

[silent_prtoagonist]

5 hours ago, Tapeta said:

I think nobody is reading parts descriptions. SWERV is suppose to be a "GASCORE" nuclear engine. Solid core nuclear thermal engines were tested in RL, but nobody knows how the heck will gascore reactor work without spewing its intrails all over current orbit.

SWERV is completely different league than NERV.

I think the SWERV is a closed-cycle gas core , aka nuclear lightbulb, design. The fissile fuel is a gas, but it's still contained inside a sealed container made of transparent, high melting point quartz crystal. It cuts the specific impulse by about half (which is why I'm assuming it's this design, an open-cycle would have an ISP of more like 3000s), but it's slightly less insane and slightly more likely to ever actually exist. 

We can leave the true Kerbal-level insanity for the Orion and nuclear salt water drives that are planned. 

[Klapaucius]

I've noticed that KSP2 recalculates Delta V on the fly. If it you are taking a delta V calculation from the launch pad, you are not going to see the benefit, but as you get out of the atmosphere, the NERV engines start to show their efficiency. 

 

I made an all hydrogen SSTO, and I noticed that while taking off my delta V was 2082, but by the time I got to orbit, having used almost 3/4 of my fuel, my delta V was now 3,278.  I actually watched the DV indicator go up once I gained sufficient altitude. 

 

[obney kerman]

If you need a use for the SWERV I have one word for you, and one word only: Tylo.

[aNERVASman]

10 hours ago, Klapaucius said:

I've noticed that KSP2 recalculates Delta V on the fly. If it you are taking a delta V calculation from the launch pad, you are not going to see the benefit, but as you get out of the atmosphere, the NERV engines start to show their efficiency. 

 

I made an all hydrogen SSTO, and I noticed that while taking off my delta V was 2082, but by the time I got to orbit, having used almost 3/4 of my fuel, my delta V was now 3,278.  I actually watched the DV indicator go up once I gained sufficient altitude. 

 

KSP1 does this as well

[The Aziz]

17 hours ago, Strawberry said:

Also delta v isnt everything and SWERV produces much more thrust then NERV

It's a deep space engine. Thrust doesn't matter, otherwise nobody would be ever using ion propulsion. With the new maneuver calculation system and timewarp, you don't have to worry about thrust. Set up the maneuver, start burning, timewarp to save yourself some time, end the burn on planned trajectory. [early game bugs not taken into account]

Edited March 15, 2023 by The Aziz