KSP2 EA: Physics simulation quality tests

[Vl3d]

This post is simple and can be very useful for everyone. Seeing the new parachute physics made me think that maybe the physics systems got a lot of attention for KSP2.

Please come up with ideas of what and how to test in KSP2 to see if the STOCK physics simulation quality is improved. I will start with:

1. Air intakes should not have the same drag coeficient as nose cones.

2. How slippery the ground is (for wheels and landing legs) should depends on the material it's made of (not only on the gravitational acceleration).

3. Test wheels.

4. Properties of a part should vary according to where is placed, not according to where it's initially attached.

5. Clipping tanks and pods together should account for combined external surface, total internal volume and the weight of the walls: A clipped with B = A + B - (A intersected B)

6. Materials strength should be more realistic (don't take into account only G-Forces and pressure).

7. Very wide fairings should not be viable (atmospheric drag should apply greater resistance).

8. Any part that's floating in the air should require structural support (struts).

9. We should not be able to visibly pull apart two parts (like docking ports). Parts should not visibly compress and auto-clip into each other under high acceleration.

10.  Climbing ladders orientation should account for the direction of gravity (no more going down the latter upside down).

11. Lack of persistent rotation.

12. Clipping craft into each other when time warping.

13. Kerbal surviving high speed crash by jumping out of craft at the last second. Test collisions in general.

14. Test the stacked decouplers exploit and other kraken drives.

What other tests could we do day one?

Edited February 14, 2023 by Vl3d

[GoldForest]

1 hour ago, Vl3d said:

This post is simple and can be very useful for everyone. Seeing the new parachute physicsmade me think that maybe the physics systems got a lot of attention for KSP2.

It could have just been a pre-made animation and no wind physics. (Which have not been confirmed btw)

We've only seen one tiny 30 second video with parachutes in them, and that video only showed the chutes for 10 seconds of that clip.  It's kind of a jump, imo, to call that "Parachute physics."

Edited February 14, 2023 by GoldForest

[Vl3d]

30 minutes ago, GoldForest said:

It could have just been a pre-made animation and no wind physics. (Which have not been confirmed btw)

We've only seen one tiny 30 second video with parachutes in them, and that video only showed the chutes for 10 seconds of that clip.  It's kind of a jump, imo, to call that "Parachute physics."

Guess that's another thing to test on day one.

[Streetwind]

Whether wing aspect ratio and/or wing sweep matters. And while I doubt the devs would go there, might as well check for area ruling.

[Vl3d]

23 minutes ago, Streetwind said:

Whether wing aspect ratio and/or wing sweep matters. And while I doubt the devs would go there, might as well check for area ruling.

How do we do that?

[GoldForest]

21 minutes ago, Vl3d said:

How do we do that?

The greater the sweep, the faster the plane should go.

The more wing ratio you have, the more lift you should have. I.e. take off at slower speeds, also fly at lower speeds. Think gliders.

Also, in regard to testing parachute physics, I'm 90% certain that there will be none that effect the model and make it flap.

Edited February 14, 2023 by GoldForest

[AtomicTech]

4 hours ago, Vl3d said:

This post is simple and can be very useful for everyone. Seeing the new parachute physics made me think that maybe the physics systems got a lot of attention for KSP2.

Please come up with ideas of what and how to test in KSP2 to see if the STOCK physics simulation quality is improved. I will start with:

1. Air intakes should not have the same drag coeficient as nose cones.

2. How slippery the ground is (for wheels and landing legs) should depends on the material it's made of (not only on the gravitational acceleration).

3. Test wheels.

4. Properties of a part should vary according to where is placed, not according to where it's initially attached.

5. Clipping tanks and pods together should account for combined external surface, total internal volume and the weight of the walls: A clipped with B = A + B - (A intersected B)

6. Materials strength should be more realistic (don't take into account only G-Forces and pressure).

7. Very wide fairings should not be viable (atmospheric drag should apply greater resistance).

8. Any part that's floating in the air should require structural support (struts).

9. We should not be able to visibly pull apart two parts (like docking ports). Parts should not visibly compress and auto-clip into each other under high acceleration.

10.  Climbing ladders orientation should account for the direction of gravity (no more going down the latter upside down).

11. Lack of persistent rotation.

12. Clipping craft into each other when time warping.

13. Kerbal surviving high speed crash by jumping out of craft at the last second. Test collisions in general.

14. Test the stacked decouplers exploit and other kraken drives.

What other tests could we do day one?

Kraken Drives!

Specifically the docking port based ones.

[Kerbart]

4 hours ago, Vl3d said:

What other tests could we do day one?

• Ladder transitions: going from one ladder to another shouldn't yeet you of into yonder, or worse, your craft (although it'll be worth watching it go boom)
• Surface excursion: kerbals shouldn't spontaneously explode when touching the surface of a planet
• a payload inside a cargo bay that is closed after opening it in flight should no longer be exposed to aero forces
• If there's still such a thing a physical timewarp: exiting the atmosphere should cut aero forces, and not apply them "as long as physical timewarp is active"
• Flags remain standing after planting them
• Broken off parts should fall at sufficient velocity and not come down to a crawl just because they're aero surfaces — or even go up

None of this sounds like it should slip through internal testing but it did in the past, so it's something to look out for.

[The Aziz]

5 hours ago, Vl3d said:

Climbing ladders orientation should account for the direction of gravity (no more going down the latter upside down).

If my craft is in orbit pointing towards the planet, which way should the Kerbal face upon leaving the airlock?

[Vl3d]

1 hour ago, The Aziz said:

If my craft is in orbit pointing towards the planet, which way should the Kerbal face upon leaving the airlock?

If either way is acceptable, the regular legs first according to the pods orientation. But if it's in space is that really climbing?

[Bej Kerman]

1 hour ago, The Aziz said:

If my craft is in orbit pointing towards the planet, which way should the Kerbal face upon leaving the airlock?

Did they say anything about being in orbit? 

[Scarecrow71]

2 hours ago, Kerbart said:

Flags remain standing after planting them

I'd just like flags to get planted with correct orientation instead of having the graphics appear backwards.

2 hours ago, Kerbart said:

Ladder transitions: going from one ladder to another shouldn't yeet you of into yonder, or worse, your craft (although it'll be worth watching it go boom)

Ladders shouldn't go into the ground upon deployment.  They should stop at the ground, regardless of remaining length to deploy.  And they should either break due to applied force when this happens OR malfunction and reverse themselves like a garage door.

[MARL_Mk1]

I'm also on the side who thinks parachutes are animated and not affected by physics

I just wanna test if landing legs finally work properly...

[The Aziz]

Just now, Scarecrow71 said:

And they should either break due to applied force when this happens OR malfunction and reverse themselves like a garage door.

Oh gods no. 

[SolarAdmiral]

Just now, Scarecrow71 said:

I'd just like flags to get planted with correct orientation instead of having the graphics appear backwards.

Ladders shouldn't go into the ground upon deployment.  They should stop at the ground, regardless of remaining length to deploy.  And they should either break due to applied force when this happens OR malfunction and reverse themselves like a garage door.

Why couldn't the telescoping ladders just detect pushback from hitting the ground and stop where they are? That seems like something just as doable as reversing like a garage door. That way it would be more forgiving for different gravity bodies and different load weights where gear suspension might be differently compressed.

[Vl3d]

4 minutes ago, Scarecrow71 said:

malfunction and reverse themselves like a garage door

And elevators with cables that are 3 meters too short.

[desert]

Entering and leaving timewarp shouldn't affect landed vessels, they shouldn't jump or clip through the surface

[Bej Kerman]

1 hour ago, MARL_Mk1 said:

I'm also on the side who thinks parachutes are animated and not affected by physics

Why? The jank you see is indicative of a physics simulation and wouldn't happen with an animation.

[pandaman]

1 hour ago, Scarecrow71 said:

I'd just like flags to get planted with correct orientation instead of having the graphics appear backwards.

Ladders shouldn't go into the ground upon deployment.  They should stop at the ground, regardless of remaining length to deploy.  And they should either break due to applied force when this happens OR malfunction and reverse themselves like a garage door.

Choose a flag without text.  I made my own flag that has a symmetrical image so I don't have that issue in KSP1.

I'd rather ladders go into the ground than be a potential cause of 'glitching out'.  Whilst technically correct it's not something that would bother me if not fixed. 

[LHACK4142]

It's a relatively minor thing but it'd be nice if landing gear didn't bounce around for all eternity after a plane crash.

[pandaman]

15 minutes ago, LHACK4142 said:

It's a relatively minor thing but it'd be nice if landing gear didn't bounce around for all eternity after a plane crash.

In itself  a minor annoyance, but a symptom of a bigger issue I think.

[Domonian]

7 hours ago, Streetwind said:

Whether wing aspect ratio and/or wing sweep matters.

I really hope we get variable wing mechanics and physics... I'd love to make stuff that takes advantage of variable wing geometry.

[Bej Kerman]

16 minutes ago, Domonian said:

I really hope we get variable wing mechanics and physics... I'd love to make stuff that takes advantage of variable wing geometry.

For a video game series where airplanes and spaceplanes have been a massive mechanic since 0.17, aerodynamics that doesn't simulate the sort of phenomenon you might find with FAR installed would really be dropping the ball.

[Domonian]

26 minutes ago, Bej Kerman said:

For a video game series where airplanes and spaceplanes have been a massive mechanic since 0.17, aerodynamics that doesn't simulate the sort of phenomenon you might find with FAR installed would really be dropping the ball.

I wouldn't be surprised if they didn't model it, since we don't even know if we'll have the parts to make variable geometry wings. I would like to see it, but I don't expect to (at least without mods).

[Bej Kerman]

3 minutes ago, Domonian said:

I wouldn't be surprised if they didn't model it, since we don't even know if we'll have the parts to make variable geometry wings. I would like to see it, but I don't expect to (at least without mods).

Swept wings acting far different to delta wings is a basic feature in the book of anyone that's picked up on how prevalent winged vessels are in the trailers and previews. To undershoot aerodynamics would be to undercook a major feature of the game. It'd be like trying to hack in Rask and Rusk while only using two-bodies and convoluted SOI setups; we expect better of a much more funded game that's being developed full time and is meant to fix what the first game didn't get right.