Concerned about KSP2 aerodynamics

[MechBFP]

Well the arguments in here for having FAR implemented in KSP2 have definitely convinced me that would be a terrible idea. 
So mission failed successfully I guess?

[Dinlink]

6 hours ago, t_v said:

Just a question: Would you be able to make the same place with forwards-sweeping wings? A lot of the more interesting-looking plane designs in KSP 1 use wings at strange sweep angles. And, if you are interested in an even harder challenge, could you try to approximate @The Aziz's train in FAR? It doesn't have to be stable, it just has to be able to barely control its flight trajectory. 

The forwards-sweeping wing version of supersonic drone:

1) With a Goofy big tail (It flights like a charm, no SAS):

2) More serious version (a little bit roll unstable on take off, but stable cruise flight, no SAS):

 

The @The Aziz's train is just a rocket in disguise that takes off from the runway , you just need a TWR greater than 1, and hide some flywheels for control. i added 8 small flywheels inside the cargo bay which gave me total control of this pseudo-supersonic-train:

Pesudo-train sonic flight:

At mach 3 in low atmosphere:

 

I don´t know why so much PTSD with FAR... The rest of the game is about the same difficulty as making planes that flight perfectly with FAR... or even getting Grand Piano to flight (SAS and TWR>1).

Rules of thumb to make plane with FAR: same as Stock KSP, and a big tail, far back tail or multi-tail... in my experience take-off runway problems with FAR are normally related to having not enough tail to get laterally stable (roll/yaw). Second recommendation: use only trimming, because SAS tends to oscillate too much, a stable plane flights by itself.

FAR (and aerodynamics and airplane design in general) Just needs a guided tutorial like stock KSP. When I started playing I didn't took any tutorials at first, killed Valentina, felt terrible and closed the game... and didn't tried again for months... but when i tried again i did the tutorials, and now I have more than 600h of gameplay, playing career in hard mode. I have learned a lot about Orbital mechanics and rocket design well beyond my wildest imagination. I think with FAR would be similar for everyone if it had a well made guided tutorial and documentation.

 

Edited October 31, 2022 by Dinlink

[t_v]

3 minutes ago, Dinlink said:

The forwards-sweeping wing version of supersonic drone:

1) With a Goofy big tail (It flights like a charm, no SAS):

2) More serious version (a little bit roll unstable on take off, but stable cruise flight, no SAS):

 

The @The Aziz's train is just a rocket in disguise that takes off from the runway , you just need a TWR greater than 1, and hide some flywheels for control. i added 8 small flywheels inside the cargo bay which gave me total control in this pseudo-supersonic-train:

Pesudo-train sonic flight:

At mach 3 in low atmosphere:

 

I don´t know why so much PTSD with FAR... The rest of the game is about the same difficulty as making planes that flight perfectly with FAR... or even getting Grand Piano to flight (SAS and TWR>1).

Rules of thumb to make plane with FAR: same as Stock KSP, and a big tail, far back tail or multi-tail... in my experience take-off runway problems with FAR are normally related to having not enough tail to get laterally stable (roll/yaw)... And use only trimming, because SAS tends to oscillate too much, a stable plane flights by itself.

FAR (and aerodynamics and airplane design in general) Just needs a guided tutorial like stock KSP. When I started playing I didn't took any tutorials at first, killed Valentina, felt terrible and closed the game... and didn't tried again for months... but when i tried again i did the tutorials, and now I have more than 600h of gameplay, playing career in hard mode. I have learned a lot about Orbital mechanics and rocket design well beyond my wildest imagination. I think with FAR would be similar for everyone if it had a well made guided tutorial and documentation.

 

Thanks for this! The dual-tail design actually seems like a reasonable ask to get sort of stable flight with inherently unstable forwards-swept wings. I was worried that in order to get those designs, you would absolutely need to do something specific like the big tail, which would be almost as bad for creativity as not being able to do those designs at all. And naturally, with enough control wheels, you can get something with any amount of torque to fly, if you can really call it flying and not just powered freefall. I'll boot up the game and try again on my wacky designs; maybe I just need some extra vertical stabilizers

[BowlerHatGuy3]

I don’t really like how people are so narrow minded. With enough tutorials and a good ui/ux system, a FAR like aerodynamic model could be just as fun and intuitive as the one that in the game right now! You would just need to learn how it works and how to apply it to your spaceplanes and rockets. I also don’t get how a better aero system would limit your creativity. You would still be able to create crazy crafts, it’s just that they won’t be able to go hypersonic anymore and completely destroy the entire point of KSP.

Like seriously, isn’t the entire point of the game to take complicated concepts and make them digestible for the average person? A more realistic aero would do nothing but give newer players a better understanding of aero and give more experienced players (like myself) a more in-depth and challenging experience.

[t_v]

21 minutes ago, BowlerHatGuy3 said:

I don’t really like how people are so narrow minded. With enough tutorials and a good ui/ux system, a FAR like aerodynamic model could be just as fun and intuitive as the one that in the game right now! You would just need to learn how it works and how to apply it to your spaceplanes and rockets. I also don’t get how a better aero system would limit your creativity. You would still be able to create crazy crafts, it’s just that they won’t be able to go hypersonic anymore and completely destroy the entire point of KSP.

Like seriously, isn’t the entire point of the game to take complicated concepts and make them digestible for the average person? A more realistic aero would do nothing but give newer players a better understanding of aero and give more experienced players (like myself) a more in-depth and challenging experience.

I agree; I think it speaks volumes to the potential of FAR's system that it is pretty fun and intuitive despite its UI shortcomings. I have not had the best experience porting my plane designs over to FAR because the aesthetic I am going for simply does not work well aerodynamically, but @Dinlink has shown me that creative designs are still possible with some extra considerations.

What I would like though is a FAR like aerodynamic model, that includes the benefits of having sleeker planes go realistically faster and the voxel model making the interiors of cargo bays possible, but tones down the realism so that people like me and people who are even worse at designing planes than me (which there are a very small number of) can learn these realistic aerodynamic effects without hitting such a steep learning wall. I think this requires changes beyond just UI, I think this requires the effects to become easier for new players. 

[mikegarrison]

On 10/31/2022 at 2:40 AM, Master39 said:

The FAR experience for most people is "Your planes don't work anymore, close the game and get back after you studied elsewhere how aerodynamics works"

Well, I suspect most people aren't airplane engineers, but the FAR experience for me was "This rocket flies exactly like I expected it would based on how it looks" while the stock aero experience for me has always been "this monstrosity would be impossible in real life, but it's what KSP wants".

The more you actually know about aerodynamics, the more that FAR seems intuitive. The more you actually know about aerodynamics, the less that stock KSP seems intuitive.

(And this is *after* they fixed the total nonsense that they originally had of aero forces being proportional to velocity rather than velocity squared.)

Edited November 15, 2022 by mikegarrison

[Bej Kerman]

Like when it was implied precise landings being a necessary thing you have to do in order to build colonies up in the KSP 2 trailer (that was a fun discussion), I feel like people are scared of no longer being at the top of KSP's difficulty curve.

Orbital mechanics and basic aerodynamics, we've already conquered that. Precise landings and aerodynamics that works? That's further up the difficulty curve than where most of us are at so it should all be simplified.

That's the general sentiment I see being repeated with many experienced players, notably both here and back when we realized we'd have to start refining our pinpoint landing skills to be able to efficiently build colonies - imagine if a new player said that KSP 2 shouldn't have orbital mechanics because they were lower down the difficulty curve and thought having to learn everything in order to play looked scary. I just want to make it clear that KSP 2 having a longer difficulty curve shouldn't be something to be scared of. We're being put in the same situation we all were when we first discovered KSP and we're scared of having to relearn things, but we shouldn't be. I'd say having to relearn things in order to be able to play a higher quality game with a better aero model and richer progression should be welcomed. If you could learn orbital mechanics in KSP after being led to believe space travel is as simple as going in a straight line thanks to the likes of Star Wars, then you most certainly can adjust to a better aero model already having picked up some basic intuition from the KSP 1 model 

[Guest]

8 hours ago, Bej Kerman said:

Like when it was implied precise landings being a necessary thing you have to do in order to build colonies up in the KSP 2 trailer (that was a fun discussion), I feel like people are scared of no longer being at the top of KSP's difficulty curve.

Orbital mechanics and basic aerodynamics, we've already conquered that. Precise landings and aerodynamics that works? That's further up the difficulty curve than where most of us are at so it should all be simplified.

That's the general sentiment I see being repeated with many experienced players, notably both here and back when we realized we'd have to start refining our pinpoint landing skills to be able to efficiently build colonies - imagine if a new player said that KSP 2 shouldn't have orbital mechanics because they were lower down the difficulty curve and thought having to learn everything in order to play looked scary. I just want to make it clear that KSP 2 having a longer difficulty curve shouldn't be something to be scared of. We're being put in the same situation we all were when we first discovered KSP and we're scared of having to relearn things, but we shouldn't be. I'd say having to relearn things in order to be able to play a higher quality game with a better aero model and richer progression should be welcomed. If you could learn orbital mechanics in KSP after being led to believe space travel is as simple as going in a straight line thanks to the likes of Star Wars, then you most certainly can adjust to a better aero model already having picked up some basic intuition from the KSP 1 model 

I think you're getting me wrong.

If it's not fully explained in game it shouldn't be in the game extends to everything. Orbital mechanics and landing included.

The era of games that needed 50 tabs of tutorials opened in the browser ended 7-8 years ago and it should have never been a thing in the first place.

Not giving the player the needed information to play the game is not a matter of difficulty curve,  it's a failure in game design.

That's why tutorials and gameplay are more important than the simulation. This is not DCS, studying manual shouldn't be a requirement to play the game.

 

9 hours ago, mikegarrison said:

The more you actually know about aerodynamics, the more that FAR seems intuitive. The more you actually know about aerodynamics, the less that stock KSP seems intuitive.

Agree, but that knowledge shouldn't come from outside the game.

[mikegarrison]

5 minutes ago, Master39 said:

that knowledge shouldn't come from outside the game.

Well, I went to college to get that knowledge, so I guess I have to disagree.

But leaving that aside, KSP does not have to be a game for everyone. If you are playing a game like, say Bioshock or Skyrim, you expect the game to teach you how to play the game, because plasmids and magic and dragons and little sisters aren't real. The game has to teach you. If you are playing something like a super-accurate flight simulator, you should expect that IRL pilots are going to know how it works better than random people off the street. Different games, different audiences.

I see KSP as being in some kind of middle ground here. The game sort of teaches you how to play it, but it doesn't teach you how to play it well. You need to go to all kinds of out-of-game or mod helpers for that -- delta-V maps and calculators, for instance. It wasn't until the very last major update that we got *any* stock tool to do things as fundamental as plan for a transfer window. For almost the entire history of the game, there wasn't even a stock delta-V calculator.

But even more than that, the game can simplify things, sure, but it shouldn't do things that are just so completely wrong that if you build something that IRL should work better, it works worse in the game.

[Bej Kerman]

7 hours ago, Master39 said:

If it's not fully explained in game it shouldn't be in the game extends to everything. Orbital mechanics and landing included.

That's a completely irrelevant problem. The discussion here is that aerodynamics should be overhauled - FAR is a golden standard here. How the game goes about explaining it is a different problem altogether - besides, if it came to Take Two forcing players to watch tutorials elsewhere, we would be saying "improve the tutorials!", not "cut orbital mechanics". Things like this should be in the game, and fully explained. No point introducing scenarios where a new aero model isn't fully explained for the sake of instigating a pushback against said aero overhaul, because the problem would be the on-boarding and not the aerodynamics model.

Edited November 15, 2022 by Bej Kerman

[Guest]

7 hours ago, mikegarrison said:

But leaving that aside, KSP does not have to be a game for everyone. If you are playing a game like, say Bioshock or Skyrim, you expect the game to teach you how to play the game, because plasmids and magic and dragons and little sisters aren't real. The game has to teach you. If you are playing something like a super-accurate flight simulator, you should expect that IRL pilots are going to know how it works better than random people off the street. Different games, different audiences.

A simulator is not a game at its inception. Most simulators are terrible at being games and that's, arguably, one of the main reasons they're so niche.

 

7 hours ago, mikegarrison said:

I see KSP as being in some kind of middle ground here. The game sort of teaches you how to play it, but it doesn't teach you how to play it well.

KSP is a good game that happens to be a sim.

 

7 hours ago, mikegarrison said:

The game sort of teaches you how to play it, but it doesn't teach you how to play it well. You need to go to all kinds of out-of-game or mod helpers for that -- delta-V maps and calculators, for instance. It wasn't until the very last major update that we got *any* stock tool to do things as fundamental as plan for a transfer window. For almost the entire history of the game, there wasn't even a stock delta-V calculator.

That's acceptable for an amateur indie game, not for something as big as KSP2. I'm going to be playing KSP2 regardless, I'm going to put the effort in to learn how to fly on a new aero model. Most of the intended target for the game probably won't.

KSP2 needs to go past KSP1 audience to be profitable, and since dumbing the game down is not an option that means tutorials, tutorials and tutorials.

Especially if they make things work differently from KSP1.

 

1 hour ago, Bej Kerman said:

FAR is a golden standard here.

The problem is here.

For you FAR is a golden standard for a better aero models. For most people against it it's the worse possible example of an undocumented mod that changes the game in unexpected ways. And that experience, over time, became synonymous with "more realistic aero model".

I know a better aero model doesn't necessarily means a more difficult one. But for most people when you say FAR that's the picture you're projecting.

[Bej Kerman]

10 minutes ago, Master39 said:

For you FAR is a golden standard for a better aero models. For most people against it it's the worse possible example of an undocumented mod that changes the game in unexpected ways. And that experience, over time, became synonymous with "more realistic aero model".

 

To reiterate

That has nothing to do with anything here, it's a completely different problem that would be better suited to a thread specifically about on-boarding.

Here we are discussing the aero model itself.

[LHACK4142]

KSP isn't known because it's got Elite: Dangerous type physics. It's known for its reasonably realistic physics that demonstrate to people intuitively how things work. I see no distinction between saying "We can't have realistic aerodynamics" and "We can't have realistic physics". Aerodynamics is a very important aspect of rocket design, and it just makes no sense to me to say "Yes orbital physics, no realistic aerodynamics". KSP's physics (and real physics) say that you can't make an X-wing fighter with as much Isp and thrust as the real thing, and that's not a problem- nobody says that it kills the spirit of the game, but when aerodynamics say you can't make something, it's taking away the goofy spirit of the game?

1 hour ago, Bej Kerman said:

That's a completely irrelevant problem. The discussion here is that aerodynamics should be overhauled - FAR is a golden standard here. How the game goes about explaining it is a different problem altogether - besides, if it came to Take Two forcing players to watch tutorials elsewhere, we would be saying "improve the tutorials!", not "cut orbital mechanics". Things like this should be in the game, and fully explained. No point introducing scenarios where a new aero model isn't fully explained for the sake of instigating a pushback against said aero overhaul, because the problem would be the on-boarding and not the aerodynamics model

I strongly agree with this. Many people are saying that FAR is too hard, things behave weirdly, and there's no documentation. That's not a problem with the realistic aero, that's a tutorial problem. If there weren't any tutorials for stock KSP, people would say "Why does going forward make my rocket slow down?".

Also you can build and fly grand pianos in FAR.

[mikegarrison]

To be clear, I'm not advocating for FAR. I don't even play with FAR anymore, because of various reasons. But I think it would be better if the aero in KSP was intuitive, and intuitive aero means that what something looks like is what it flies like. If it looks streamlined, it should be OK. If it looks like a hedgehog, it should be draggy. If something is clearly not exposed to the airstream, it should not generate lift or drag.

In a game like KSP, it is not too much to expect that players will (even if by trial and error) figure out a little bit of the aerodynamics of rockets, just like they have to figure out things like delta-v and staging and TWR and that you burn at one point in the orbit to affect everywhere in the orbit except the one point you burn at.

[joratto]

On 10/31/2022 at 4:03 PM, The Aziz said:

So I won't be able to fly a train?

Because it doesn't match with very restrictive areodynamic rules?

 

You should not be able to fly things like this, because the exploits that make it easier to get something like this to fly (like wing surfaces generating lift whether or not they're buried deep inside the vehicle) can also make it harder for new players to understand how to build a plane that flies.

For example, you should not have to worry about the lift/drag of a part that's deep inside a fairing or otherwise visually covered up. There have been many occasions where I built something that SHOULD work in real life, but fails in ksp because some tiny, invisible, uncapped fuel tank was generating 10x more drag than the rest of my craft. You would expect those parts not to affect your aerodynamics at all, and it can cause great headaches to try to identify what unrealistic gameplay "feature" is responsible for flipping your SSTOs. That doesn't teach you anything except how to exploit the game.

[tstein]

On 10/29/2022 at 5:45 AM, Bej Kerman said:

I honestly can't see any conceivable reason Intercept wouldn't want to go as realistic as possible here.

Because as realistic as possible regarding aerodynamics,  usually result in only a top of the lien computer being able to handle the simulation. Even dedicated flight sims make shortcuts to avoid that. Effectively there is need to find a middle ground.

[tstein]

On 10/31/2022 at 11:20 AM, Bej Kerman said:

On 10/31/2022 at 11:20 AM, Bej Kerman said:

Literally wrong. I['m not saying I'm disagreeing, I'm saying this literally isn't how planes work. All you need to keep a little glider or a Boeing airliner in the air is to hold the yoke back a bit so that the ailerons keep the plane from nosing down. Autopilot functions like holding a heading or pitch and maintaining a certain speed are usually only used for insanely tedious and simple tasks like keeping a plane straight on an intercontinental path.

 

 

I think he meant things like the F-35 that sacrifices so much for stealth that it is very unstable without a computer making  minor corrections all the time.

[KSACheese]

Honestly, I get the argument for having a more advanced aero model. And, in some ways, I kind of agree. Introducing a more advanced model from the start probably wouldn't hurt the intuitiveness of the game for new players, as long as they are willing to learn. However, from my personal experience playing with FAR, it kind of annoys me. Not that I can't play with it installed, it just is kind of a hassle to play with sometimes. As moderately experienced player, if I find FAR annoying to deal with, I imagine it might be a bit more to someone who has no idea what they're doing. To me, the FAR-like flight model had so many factors that it was, at times, hard to keep track of. With the stock flight model, it's appropriate to just measure by the simple characteristics of aircraft - center of mass and lift - which makes it much easier for an inexperienced person to deal with.

If it were possible to merge the two, meet somewhere in the middle, I'd be much more willing to agree. But my experiences with FAR push me into believing that a more stock-like aero model would be more appropriate for the (hopefully) thousands of new players. 

Also, I just have more fun with the stock model, but that's just me :))

Edited November 21, 2022 by KSACheese

[Bej Kerman]

27 minutes ago, KSACheese said:

To me, the FAR-like flight model had so many factors that it was, at times, hard to keep track of.

KSP 2 shouldn't shy away from showing players the reality of designing an aircraft.

[KSACheese]

2 hours ago, Bej Kerman said:

KSP 2 shouldn't shy away from showing players the reality of designing an aircraft.

It's not about shying away from teaching them, it's inviting them. Motivation is a huge factor in someone's willingness to learn. If they find they like flying planes, they can find a way to fly them more realistically. But a player shouldn't be frustrated to a point where they have no interest in learning further, and I believe some percentage of the newcomers would by an aero model that's not appropriately designed to introduce them to the concepts.

You don't go from elementary to college right away, and why should that be the case with them?

[Guest]

On 11/18/2022 at 12:59 PM, joratto said:

For example, you should not have to worry about the lift/drag of a part that's deep inside a fairing or otherwise visually covered up. There have been many occasions where I built something that SHOULD work in real life, but fails in ksp because some tiny, invisible, uncapped fuel tank was generating 10x more drag than the rest of my craft. You would expect those parts not to affect your aerodynamics at all, and it can cause great headaches to try to identify what unrealistic gameplay "feature" is responsible for flipping your SSTOs. That doesn't teach you anything except how to exploit the game.

This is 100% in the realm of bugs, not realism of the aero model.

I don't think that people that don't want all the headaches FAR provides are here to defend parts closed inside fairings providing lift or causing drag.

[Vl3d]

On 11/4/2022 at 1:20 AM, t_v said:

What I would like though is a FAR like aerodynamic model, that includes the benefits of having sleeker planes go realistically faster and the voxel model making the interiors of cargo bays possible, but tones down the realism so that people like me and people who are even worse at designing planes than me (which there are a very small number of) can learn these realistic aerodynamic effects without hitting such a steep learning wall. I think this requires changes beyond just UI, I think this requires the effects to become easier for new players. 

I'll add my 2 cents to the discussion even though I don't have much experience with FAR. I can say this:

I would like aerodynamic forces and aero / G stress to be more important. Rockets spinning in the air should break apart (leading to more emphasis on Aborts), some parts that are not aerodynamically protected should be torn off,  Max-Q should be notified visually more clearly and have a greater impact on the ascent profile.

Generally I feel that KSP parts / materials / joints are too strong compared to aerodynamic drag forces.

Edited November 21, 2022 by Vl3d

[GoldForest]

1 hour ago, Vl3d said:

I'll add my 2 cents to the discussion even though I don't have much experience with FAR. I can say this:

I would like aerodynamic forces and aero / G stress to be more important. Rockets spinning in the air should break apart (leading to more emphasis on Aborts), some parts that are not aerodynamically protected should be torn off,  Max-Q should be notified visually more clearly and have a greater impact on the ascent profile.

Generally I feel that KSP parts / materials / joints are too strong compared to aerodynamic drag forces.

I think all that should stay in mods. KSP is about having fun, not learning about IRL physics. If your rocket and plane broke apart every time you launched, I don't see many people playing KSP like they do. 

I want to fly my cube to space and not have to worry about it crumbling away around my kerbals. 

[The Aziz]

2 hours ago, Vl3d said:

Max-Q should be notified visually more clearly and have a greater impact on the ascent profile.

We got Max Drag currently, and I think that's sufficient. You could keep wiggling your rocket around at low speeds without much happening, but steer off the prograde when the drag is the strongest and you're not going to space today. So I do what Falcons are doing, throttling down to reduce the drag a bit to keep my ability to steer the rocket on the way up when the chosen direction is critical.

[GoldForest]

23 minutes ago, The Aziz said:

We got Max Drag currently, and I think that's sufficient. You could keep wiggling your rocket around at low speeds without much happening, but steer off the prograde when the drag is the strongest and you're not going to space today. So I do what Falcons are doing, throttling down to reduce the drag a bit to keep my ability to steer the rocket on the way up when the chosen direction is critical.

You mean you don't just 100% throttle all the way and punch through max-Q and hope mechjeb can handle it?