Coulomb friction model

Category: Structural Analysis | Integrated 2026-04-06
CAE visualization for coulomb friction theory - technical simulation diagram
Coulomb Friction Model

Coulomb friction: Theoretical Foundations

What is Coulomb Friction?

🧑‍🎓

Professor, please teach me the basics of friction models in FEM.


🎓

Coulomb friction is the most basic friction model. The tangential friction force is proportional to the normal force:


$$ |\tau| \leq \mu \cdot p_n $$

  • $\tau$ — Friction force (tangential stress)
  • $p_n$ — Contact pressure (normal stress)
  • $\mu$ — Coefficient of friction

Sticking and Sliding

🎓

Two states:

  • Sticking — $|\tau| < \mu p_n$. No relative sliding.
  • Sliding — $|\tau| = \mu p_n$. Friction force reaches its limit and sliding occurs.

🧑‍🎓

So the transition from sticking to sliding is nonlinear, right?


🎓

Coulomb friction adds another layer of nonlinearity to the contact problem. It's a double nonlinearity: Contact (contact/no contact) + Friction (stick/slip).


Typical Values of Friction Coefficients

🎓
Material Pair$\mu$ (dry)
Steel-Steel0.15–0.3
Steel-Aluminum0.2–0.4
Steel-Rubber0.5–0.8
Steel-Teflon0.04–0.1
Concrete-Steel0.3–0.5

Friction Settings in FEM

🎓
  • Abaqus: FRICTION, PENALTY (Penalty method) or FRICTION, LAGRANGE
  • Nastran: BFRIC card
  • Ansys: MP, MU, cid, 0.3
  • LS-DYNA: *CONTACT with FS (static friction), FD (dynamic friction)

  • Summary

    🎓

    Key points:


    • $|\tau| \leq \mu p_n$ — Upper limit of friction force
    • Binary state of stick/slip — Adds additional nonlinearity to contact
    • Friction coefficient depends on material pair — Use measured values
    • Friction handled by penalty method — Penalty spring also in sliding direction

    Coffee Break Yomoyama Talk

    Coulomb's 1781 Experiment

    French military engineer Charles-Augustin de Coulomb systematically conducted hundreds of friction experiments in 1781 using combinations of wood, metal, and stone, and derived "Coulomb's law of friction," which states that friction force is proportional to the normal load and independent of contact area. Looking further back, it was discovered in 1967 that Leonardo da Vinci had illustrated the same relationship in his secret notebook from 1495, causing a surprise in the history of science.

    Computational Methods for Coulomb friction

    Penalty Method for Friction

    🎓

    Friction is also handled by the penalty method. Tangential "stick penalty":


    $$ \tau = k_t \cdot \delta_t \quad (|\tau| < \mu p_n) $$

    $k_t$ is the tangential penalty stiffness, $\delta_t$ is the tangential elastic slip. Transition to sliding occurs when $|\tau| = \mu p_n$.


    🧑‍🎓

    Does "elastic slip" mean there is a tiny slip before actual sliding?


    🎓

    It's an artificial "elastic slip" in the penalty method. Physically, there is zero slip in the stick state, but numerically, a finite tangential penalty stiffness causes a tiny slip. A larger $k_t$ results in smaller elastic slip.


    Static and Dynamic Friction

    🎓

    In LS-DYNA, static friction coefficient $\mu_s$ and dynamic friction coefficient $\mu_d$ can be set separately. When sliding starts, it transitions from $\mu_s → \mu_d$ ($\mu_d < \mu_s$). The transition is smoothed using an exponential function.


    Summary

    🎓
    • Tangential penalty method — Spring for stick state. Coulomb friction upon sliding.
    • Elastic slip — An artifact of the penalty method. Minimize by increasing $k_t$.
    • Static/Dynamic friction — Set FS/FD individually in LS-DYNA.

    • Coffee Break Yomoyama Talk

      Stick-Slip Determination

      In FEM implementation of Coulomb friction, it's necessary to branch each contact point into "stick" or "slip." The return mapping method introduced by Zienkiewicz et al. in the 1970s adopted a two-step operation: calculating a trial stress and then projecting it onto the friction cone, significantly reducing computational cost per iteration. This method remains the standard algorithm in ABAQUS and ANSYS today.

      Coulomb friction in Practice

      Friction in Practice

      🎓

      Problems where friction is important:


      • Clamping force in bolted joints — Load transfer via friction
      • Press forming — Friction between die and blank affects deformation
      • Brakes — Friction force = braking force
      • Pipe supports — Friction in sliding supports

      Uncertainty in Friction Coefficient

      🎓

      Friction coefficient has very high variability (±30% or more). Strongly depends on surface condition (roughness, lubrication, oxide film).


      🎓

      Countermeasures:

      • Sensitivity analysis of friction coefficient (two cases: $\mu_{low}$ and $\mu_{high}$)
      • Use measured values (test results)
      • Literature values are for reference only

      Practical Checklist

      🎓
      • [ ] Is the friction coefficient appropriate for the material pair?
      • [ ] Has friction sensitivity analysis been performed? ($\mu \times 0.5$ and $\mu \times 1.5$)
      • [ ] Is elastic slip within acceptable limits? (tangential penalty stiffness)
      • [ ] Is the stick/slip distribution physically reasonable?
      • [ ] Is friction not adversely affecting convergence? (First confirm convergence without friction)

      • Coffee Break Yomoyama Talk

        Brake Squeal Analysis

        Automotive brake squeal occurs when slight velocity dependence of the Coulomb friction coefficient generates negative damping. In a complex eigenvalue analysis using Nastran conducted jointly by Ford and TRW in the 2000s, simply changing the friction coefficient from 0.35 to 0.40 tripled the number of unstable modes, providing insights directly linked to pad shape optimization.

        Related Simulators

        Experience the theory through interactive simulators in this field

        All Simulators

        Related fields

        Thermal AnalysisManufacturing Process AnalysisV&V · Quality Assurance
        Rate this article
        Thank you for your feedback!
        Helpful
        More details
        Report error
        Helpful
        0
        More details
        0
        Report error
        0
        Written by NovaSolver Contributors
        Anonymous Engineers & AI — Sitemap
        About the Authors