Hi @Justin,
Thank you for your detailed explanation of the problem. From your description, the instability during high-slip conditions is very likely due to numerical stiffness in the tire dynamics, rather than simply the global step size. The collapse of the Fx relaxation integrator under aggressive inputs is a strong indication of this.
Although I * do not have access to Simulink* to test your model directly, I have reviewed MathWorks documentation and open-access resources to provide guidance, so please follow recommendations below.
1. Solver Strategy: Stiff tire dynamics often destabilize with coarse fixed-step solvers (e.g., 0.01 s). Using a variable-step solver such as `ode15s` or `ode23tb` can help diagnose stiffness. Once confirmed, fixed-step settings can be refined for real-time deployment.
Reference: MathWorks – Choose an ODE Solver https://www.mathworks.com/help/simulink/ug/choose-an-ode-solver.html
2. Relaxation Length & Damping: Relaxation length dictates how quickly tire forces build under slip. If too short, integrators can collapse during aggressive transients. Increasing this parameter or adding damping improves stability.
Reference: Wikipedia – Relaxation length https://en.m.wikipedia.org/wiki/Relaxation_length
3. Subsystem Solver Consistency:Extremely small steps in atomic subsystems (e.g., 1e‑4 s) can cause conflicts if nested integrators inherit incompatible solver settings. Consistent solver configuration across the model hierarchy is recommended.
Reference: MathWorks – Control Execution of Atomic Subsystems https://www.mathworks.com/help/simulink/ug/atomic-subsystems.html
4. Transient Input Shaping: Abrupt throttle or steering inputs excite high-frequency dynamics, which can destabilize the tire model. Smoothing inputs via ramps or filters reduces stiffness while maintaining maneuver fidelity.
Reference: MathWorks – Modeling Vehicle Dynamics https://www.mathworks.com/help/vdynblks/ug/modeling-vehicle-dynamics.html
Summary of Actionable Steps
Step 1: Run the model with a variable-step solver to verify stiffness-related instability. Step2: Adjust tire relaxation length and/or damping to stabilize transient responses. Step3: Ensure consistent solver configuration across all subsystems. Step 4: Replace abrupt control inputs with smoothed or ramped profiles.
These steps reflect best practices from both MathWorks documentation and open-access literature. While I cannot directly test in Simulink, following this workflow should help you stabilize the model under high-slip conditions.
Good luck!
