Dear all,
I am working on an Abaqus simulation and I would greatly appreciate your help. My goal is to simulate the sit-down movement of a human model on a rigid plane.
I imported into Abaqus the lower part of the THUMS human model (developed by Toyota), from the navel down. The parts are provided as shell elements with an orphan mesh.
In the Assembly module, I organized the geometry into three main parts (bones, muscles, skin) using Merge Instance. Then I merged these into a single part, in order to avoid having too many interactions between the original ~21 components.
For the materials:
- I assigned density and elastic properties (Young’s modulus and Poisson’s ratio).
- I created three shell sections with different thicknesses: bone = 5 mm, muscles = 13 mm, skin = 1 mm.
- The full THUMS model represents an average male (height = 179 cm, weight = 78 kg). Since I imported only the lower body, I estimated its weight and applied the remaining body weight to a reference point placed above the model.
- This reference point was coupled to the whole body using a kinematic coupling.
For the boundary conditions and interactions:
- I created two rigid planes: one under the pelvis and one under the feet.
- General contact was defined between the rigid planes and the human model.
- Encastre boundary conditions were applied to the rigid planes.
- The analysis was organized in three steps: Initial, Gravity (gravity load applied), and Load (remaining body weight applied to the RP).
Problems I encountered:
- The simulation ends almost immediately with the following errors: PENETRATION ERROR TOO LARGE COMPARED TO DISPLACEMENT INCREMENT WARNING: DISPLACEMENT INCREMENT FOR CONTACT IS TOO BIG
- I also faced a warning about double-sided facets in General Contact: “The general contact domain for modeling contact interactions in Abaqus/Standard has double-sided facets… initial contact adjustments may be incorrect…”I tried to solve this by switching to a surface-to-surface contact definition, which improved the situation but the simulation still fails.
Has anyone worked on a similar setup?
- Do you have suggestions on how to properly define the rigid planes and contact conditions for this kind of human model?
- Are there best practices to ensure convergence and avoid penetration errors in sit-down simulations?
Any advice or tips would be very helpful. I will attach images of the model for reference.
Thank you in advance for your time and support!