Fix issue 925 (wheel rotation calculation)#945
Closed
tomo0613 wants to merge 470 commits into
Closed
Conversation
added missing `filter` parameters
FeatureId support
chore: update typedoc to 0.22
Add basic debug-render support
Simplify handles
The iteration order is now determined by the index part of the inserted handles.
Ensure that the iteration orders remain the same after deserialization
chore: update timestep jsdoc
chore: install wasm-pack via npm
* better extensions naming for rapier compat files * Add changelog
* feat: un-deprecate collider properties accessors * feat: add functions for getting a collider’s translation/rotation wrt. its parent rigid-body * Update to rapier 0.27.0-beta.0 * Release v0.18.0-beta.0 * chore: run prettier
* feat: update to rust rapier 0.27 * feat: add access to the internal performance monitoring from JS * feat: add function to get/set the max number of CCD substeps & run prettier * chore: update Cargo.lock * Release v0.18.0 * chore: re-run prettier
* chore: update to rapier 0.28 * Release v0.18.1
* Expose revolute joints with separate local axes * Update src/dynamics/joint.rs Co-authored-by: Sébastien Crozet <sebastien@crozet.re> --------- Co-authored-by: Sébastien Crozet <sebastien@crozet.re>
wasm-bindgen 0.2.109 removed the `new Function("return this")` fallback
for getting the global object (PR wasm-bindgen/wasm-bindgen#4910),
which required `'unsafe-eval'` in Content-Security-Policy `script-src`.
With this update, applications using @dimforge/rapier3d-compat can use
the more restrictive `'wasm-unsafe-eval'` instead of `'unsafe-eval'`,
blocking JavaScript eval() while still allowing WebAssembly compilation.
Refs: dimforge/rapier.js#366
…e#370) This implements the frame transformation capabilities requested in issue dimforge#271, allowing users to set joint frames after joint creation via: - setFrameX1() / setFrameX2() for rotation frames - setLocalFrame1() / setLocalFrame2() for full frame (anchor + rotation) These methods enable more flexible joint configuration in physics simulations, particularly for use cases like VistaKine where joint frames need adjustment after initial setup. Fixes dimforge#271
Author
|
video evidence of the bug: offroadJS_v2.rapier_current.mp4and the fix: offroadJS_v2.rapier_fixed.mp4 |
Author
|
As I was testing, found out that the fix is not complete. I'm comparing the ccde to cannon-es (also bullet port), and will update this PR |
Author
|
I added 2 extra conditions based on cannon-es implementation. this handles if a wheel should start spinning (if wheel in air and engine force is applied, or on ground but looses grip while engine force is applied [burnout]): this handles if a wheel should stop spinning: I also checked ammo.js and these issues are there as well |
This was referenced Jun 21, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
#925
explanation (by Claude):
The Bug
In
src/control/ray_cast_vehicle_controller.rs, the wheel rotation update uses afwd(forward) vector computed independently from the chassis rotation:But the friction impulse system in
update_friction()computes forward as:These two methods produce vectors that can point in opposite directions depending on the axle axis convention. For example, with
index_forward_axis = 2(Z forward) and a wheel axle along-X:forward_ws[i]=Y × (-X)=-Zfwd=+Z← exact opposite!The result:
delta_rotationhas the wrong sign, making the wheel appear to spin backward (and it cancels out to ~0). When steering is applied, the axle rotates, which partially corrupts/corrects the mismatch — which is exactly why rotation only appears when steering.