Parts & Vessels Rovers are Hitting a Physics Glitch Every 1000m from the Location of Rover Being Loaded into the Game With Disastrous Results

[Anth]

Reported Version: v0.1.2 (latest) | Mods: none | Can replicate without mods? Yes 
OS: Windows 10 | CPU: Intel i9 9900K | GPU: Nvidia 3070ti | RAM: 32GB

Smaller Rover to Show Severity:

Spoiler

Video Evidence:
Smaller Rover Spinning out at Approximately 1000m.
https://youtu.be/kbh2DyfRzb4?t=26 (starts at 26s)

To Replicate:

1. Load SmallerRoverTestSave
2. Move forward to just after the first Kerbal at 25-30m/s. Don't Timewarp to get there.

Save File: SmallerRoverTestSave.json
Craft File: SmallRoverTestCraft.json

Larger Rover to show its at each 1000m interval:

Spoiler

Kerbals standing at 1000m, 2000m, 3000m, 4000m, 5000m, 6000m, 7000m, 8000m

Video Evidence:
Larger Rover Reacting to some sort of glitch at 1000m intervals.
https://youtu.be/ruS_RwW27xs (Use the time stamps to skip the travel between if desired)

Note: Didn't happen for the second to last Kerbal. Last one was unreliable because the rover flew over onto the beach

To Replicate:

1. Load LargeRoverTestSave
2. Be about 30m/s at each Kerbal. Don't Timewarp.

Save File: LargeRoverTestSave.json
Craft File: LargeRoverTestCraft.json

Additional information:

• Quicksave and Quickload and the 1000m glitch will reset the distance and will now happen from 1000m away from that new Load point.
• Physics Warp appears to resets the 1000m glitch point as well.
• Why the Large Rover didn't end up on its side or on its back is because it wasn't travelling fast enough.
• I tried it on Minmus and it did do it at 1000m but I couldn't get consistent results because the surface wasn't flat enough.
• The game is reporting the Kerbal 2000m (not 1000m) away as being on a crash trajectory each time the large rover glitches

[Lyneira]

I've ran across this issue countless times and always wondered if there was something on the terrain mesh that the wheels got hung up on. I never realized it was every 1000m, and that quicksave/quickload and physics warp resets the distance. From this, I get the distinct impression they're moving the origin of the physics simulation every 1000m and there is a bug in the transformation that causes a rotational impulse on the rover, either directly or through an erroneous collision with the terrain.

It brings me back to driving my rover at high speed on Ike's rough terrain, where I spent about as much time flying as driving and didn't see the glitch as often as on Duna, where I spent all my time on the ground. The glitch never happens while the rover is in the air, which suggests it affects wheel parts when in contact with the terrain.

[asdasdzxc]

It happened to me around every 1000m on Duna with 12m/s speed and I noticed you can't hit the rocks on Duna.  The rover just passed through the rocks, they're not physically existed on the game.

 

[BlackholeKS]

Can confirm still very much a major issue.

[Davidian1024]

Pretty sure I've experienced this.

[Anth]

5 hours ago, Davidian1024 said:

Pretty sure I've experienced this.

Everyone whose played KSP2 and has used rovers or planes have experienced this.

[cheese3660]

On 6/16/2023 at 6:17 AM, Lyneira said:

From this, I get the distinct impression they're moving the origin of the physics simulation every 1000m and there is a bug in the transformation that causes a rotational impulse on the rover, either directly or through an erroneous collision with the terrain.

I do believe this is the case

[Vl3d]

Reported Version: v0.1.4 (latest) | Mods: none | Can replicate without mods? Yes 
OS: Windows 11 | CPU: AMD Ryzen 5 7600X | GPU: Nvidia RTX 4070 | RAM: 32 GB DDR5

Rover using RoveMax M1 or TR2-L wheels suddenly / randomly swerves left (without user input) after coasting forward at high speed (+20 m/s). Reproducible by just holding (only) W for a while on the runway and going straight for about 20-30 seconds.

Included Attachments:

[Anth]

@Vl3d Merged your bug report with this one. It happens every 1000m

[Tyresan]

Reported Version: v0.1.4 (latest) | Mods: none | Can replicate without mods? Yes 
OS: Win 10 | CPU: I7 4790k | GPU: GTX 1080 | RAM: 32GB

 

Hello Support team,

I figured out that vehicles with wheels gets a kick sideways the driving direction every 50 to 120 seconds while driving around. This will happend in normal, as well in timewarp 2x and 4x.

The kick stays sideways even the vehicle was a few times expired this kick. So it is not depending on the cardinal direction.

I tried this out with planes and rovers. It seems the kick is more influense as lighter the vehicle is and can cause a instand turn of 30° (yaw). The tires are not moving in this eyeblink moment (as i observe). Typical droce speed are 15m/s to 20m/s

Further it seems this can be forced as i stay longer (30 sec. and above) in 4x physical time warp and exit the warp.

The point the force is applied seems the front of your vehicle, unless there is a Kerbal driven device or drone and unless SAS is on or off.

Best regards

Mark

Edited September 21, 2023 by Anth12

[Anth]

@Tyresan Merged your bug report with this one.

[Vl3d]

Reported Version: v0.1.5 (latest) | Mods: none | Can replicate without mods? Yes 
OS: Windows 11 | CPU: AMD Ryzen 5 7600X | GPU: Nvidia RTX 4070 | RAM: 32 GB DDR5

Bug seems to be present in 0.1.5.

Included Attachments:

[Dragonld]

Reported Version: v0.1.5 (latest) | Mods: none | Can replicate without mods? Yes 
OS: Windows 11 | CPU: Ryzen 9 5900x | GPU: RTX 4070 | RAM: 32GB

 

When going down the runway at high speeds with rover wheels, there seem to be bumps(?) causing the vehicle tu spin out of controll.

 

Included Attachments:

buggycarphysics.mp4

Edited October 27, 2023 by Spicat

[Spicat]

@Dragonld, merged your bug report.

[MechBFP]

I made a rocket powered rover to test this and the results are somewhat unpredictable. Sometimes it messes with the vehicle a lot and other times not so much. 

[The Aziz]

It's a global issue, ran into this on Duna.

[Anth]

2 hours ago, The Aziz said:

It's a global issue, ran into this on Duna.

Yeah, it happens everywhere.

[Anonymous Bug Reporter]

Reported Version: v0.2.0 (latest) | Mods: none | Can replicate without mods? Yes 
OS: Windows 11 Pro 64-bit (10.0, Build 22621) (22621.ni_release.220506-1250) | CPU: Intel(R) Core(TM) i7-9700KF CPU @ 3.60GHz (8 CPUs), ~3.6GHz | GPU: NVIDIA GeForce RTX 2070 SUPER | RAM: 16384MB RAM

When driving my rover on Duna, it periodically spins out. Almost like it "clips" something. Haven't tested on kerbin....this is making the duna monument mission very frustrating. 

Included Attachments:

[Vl3d]

I confirm it's still a problem. But it seems to happen at greater distances.

[mattihase]

so like 2km now?

[Kerbart]

38 minutes ago, mattihase said:

so like 2km now?

No, I was able to drive down the entire runway without a glitch. Only on the beach it had a spontaneous glitch. This might even be terrain related and not due the original bug.

[mattihase]

2 hours ago, Kerbart said:

No, I was able to drive down the entire runway without a glitch. Only on the beach it had a spontaneous glitch. This might even be terrain related and not due the original bug.

I also had some pretty glitchy movement near the ocean. was it at the transition point between land and wet? That seems to mess with kerbals' movement so I can imagine it also not being good for wheels.

Edited December 23, 2023 by mattihase

[Urus28]

I had this problem several times around the KSC and on Minmus, with the rover jumping in the sky on Minmus. I've been able to decrease the phenomenon by decreasing the spring value to minimum.

[MechBFP]

On 12/23/2023 at 2:06 PM, Kerbart said:

No, I was able to drive down the entire runway without a glitch. Only on the beach it had a spontaneous glitch. This might even be terrain related and not due the original bug.

I easily run into it on the runway, but that is with a rover that is going around 100 m/s.

[The Aziz]

Nice to see the thread right at the top as I was just about to post here.

So after some short testing - it does happen occasionally, but it's much less severe. Does not seem to happen every km, and, unlike previous versions where to vehicle could sink in the ground and get launched in the air, now it's only a slight nudge. Still needs work for flawless surface operations, but it's much better.

[MechBFP]

1 hour ago, The Aziz said:

Nice to see the thread right at the top as I was just about to post here.

So after some short testing - it does happen occasionally, but it's much less severe. Does not seem to happen every km, and, unlike previous versions where to vehicle could sink in the ground and get launched in the air, now it's only a slight nudge. Still needs work for flawless surface operations, but it's much better.

My vector powered rover (with wings configured to provide a constant down force through the centre of mass) gets thrown completely out of control.