Chengcheng WANG

• Post By Chengcheng WANG
• 3 months ago
• Post Type Public

Hello everyone,

I’m currently working on a thrust ball bearing simulation using Abaqus (implicit dynamics), and I’m encountering significant oscillations in the contact force (CFNM). I’d like to ask for advice on how to reduce or eliminate these oscillations.

🔧 Model Overview:

• Analysis type: Implicit dynamic analysis

• Simulation time: 0.01 second

• Material: All parts are steel with elastic–plastic behavior and I add some density + damping

• Ball settings: The balls are modeled as rigid bodies

• Contact pair: Balls and inner ring

• Contact formulation:

  • Surface-to-surface

  • Hard contact (normal behavior)

  • Friction coefficient: 0.2

  • Contact damping factor: 0.2

📌 Boundary and Loading Conditions:

• The balls rotate around the inner raceway (orbital motion)

• Each ball is subjected to a vertical force (toward the race)

• The raceway is fully fixed at the bottom

• Due to friction, the balls also spin around their own axes

❗ The Problem:

The CFNM (contact force in the normal direction) shows strong oscillations throughout the simulation. This instability leads to unreliable force results and makes it difficult to extract meaningful physical insights (see example plots below if needed).

✅ What I’ve Tried:

• Using soft contact (reduced oscillations, but I need to retain stiffness accuracy → hard contact is necessary)

• Applying smoothed initial loading using amplitude curves

• Mesh refinement in the contact area

• Adjusting contact damping (set to 0.2)

Unfortunately, none of these methods sufficiently reduced the oscillations.

❓ Questions:

1. Are there more effective ways to reduce or suppress contact force oscillations in implicit dynamic simulations?

2. Is my current contact damping value (0.2) too high or too low? Any empirical guidance?

Any advice, suggestions, or experience sharing would be greatly appreciated. I’m happy to provide figures or portions of the input file if helpful.

Thanks in advance!