【EDEM Physic Model Selection】Simulating Steel as a Particle Material in EDEM

Selena Lin

Hi All,

Did anyone use steel as a particle material in EDEM?

I have tried EEPA+Bonding V2+Type C Rolling Friction to simulate the property of steel.

1. EEPA (Elastic–Plastic Adhesion Model: This model accounts for plastic deformation and adhesion effects, which are essential for representing the contact behavior of metallic materials. Since steel undergoes both elastic and plastic deformation under force, EEPA helps capture these characteristics.
2. Bonding V2 Model: This model is used to simulate the metallic bond between particles, ensuring that they stay connected under stress rather than separating easily. It helps represent steel's cohesion and structural integrity.
3. Type C Rolling Friction: This friction model prevents excessive rotation of particles, which is important for mimicking steel's high resistance to rolling and deformation under load.

However, I observed that the material rebounded significantly when a force was applied in the Uniaxial Compression Test, which does not align with the expected behavior of steel.

Has anyone successfully simulated steel in EDEM? Any suggestions for improving the model setup?

Best regards,

Selena Lin

Stephen Cole

Hi Selena,

Typcially harder, stiffer mateirals do require lower time-steps. We usually recommend 20% of the Rayleigh time-step but as a general rule you may need to reduce this for stiffer material types.

For the compression test is the material bonded with V2 bonding during compression? The V2 bonding is elastic only so if the bonds don't break then you would see an elastic response from this model, even if EEPA is also included as the base model, however I would expect that the bonds should break with relatively little deformation for steel.

Selena Lin

https://community.altair.com/discussion/comment/191094#Comment_191094

Hi Stephen,

Thank you for your response.
The V2 Bonding model was applied during the compression test.

Following your suggestion, I'll try using the Linear Cohesion model instead of Bonding V2 or running the simulation without it, reducing the Rayleigh time step percentage to 15%.