Scale consistency on coupling SMC and DEM

[Leen]

Hello,

I am currently working on a project that combines Chrono’s SMC and DEM modules within the same simulation. I initially developed my robot model using SMC alone, following SI units consistently (meters, kilograms, seconds). In that setup, the robot’s geometry, inertias, and masses appeared correct and behaved as expected.

However, when I began integrating the SMC-based robot into a DEM environment, I noticed a severe scale mismatch: the robot appeared extremely small—on the order of a single DEM particle, or even smaller. Visually and physically, it looked almost like a “bug” relative to the granular terrain.

My DEM setup was inspired by the official DEM demos (DEM_ballCosim & DEM_fixedTerrain), and I am using the following JSON parameter file (excerpted below for clarity):

{

"sphere_radius": 1.0,

"sphere_density": 2.5,

"box_X": 200,

"box_Y": 200,

"box_Z": 60,

"time_integrator": "centered_difference",

"step_size": 5e-5,

"time_end": 5.0,

"grav_X": 0,

"grav_Y": 0,

"grav_Z": -980,

"normalStiffS2S": 1e8,

"normalStiffS2W": 1e8,

"normalStiffS2M": 1e6,

"normalDampS2S": 10000,

"normalDampS2W": 10000,

"normalDampS2M": 1000,

"tangentStiffS2S": 1e8,

"tangentStiffS2W": 1e8,

"tangentStiffS2M": 5e5,

"tangentDampS2S": 2000,

"tangentDampS2W": 2000,

"tangentDampS2M": 500,

"static_friction_coeffS2S": 0.5,

"static_friction_coeffS2W": 0.5,

"static_friction_coeffS2M": 0.3,

"friction_mode": "multi_step",

"rolling_mode": "schwartz",

"rolling_friction_coeffS2S": 0.1,

"rolling_friction_coeffS2W": 0.1,

"rolling_friction_coeffS2M": 0.05,

"cohesion_ratio": 0,

"adhesion_ratio_s2w": 0,

"adhesion_ratio_s2m": 0,

"verbose": 0,

"psi_T": 32,

"psi_L": 16,

"write_mode": "csv"

}

As an experiment, I scaled my entire SMC system from SI to CGS units. After doing so, the robot appeared to be correctly sized relative to the DEM terrain, and interactions looked correct. 

That being said, this approach was quite messy and time-consuming, and it feels like I may be compensating for something incorrectly rather than using the system as intended.

The reason I am reaching out is that I suspect I may be missing a fundamental assumption or requirement regarding unit conventions or scaling when coupling SMC with DEM. Specifically:

• Is DEM implicitly always assuming CGS units (as suggested by gravity ≈ −980)?

• Is there a recommended or canonical way to keep SMC and DEM consistent without fully rescaling one system?

Any guidance or clarification would be greatly appreciated. Thank you very much for your time and support.

Best regards,
Leen Said

[Ruochun Zhang]

Hi Leen,

Chrono is dimensionless. There's no assumed unit system. So you need to make sure to use a consistent unit system, rather than seeing it as an optional improvement. DEM's demos tend to implicitly use CGS units but that's never baked into the solver. And when you said you used SI for SMC, that is also implicit. So you see the problem: Your robot's size has the order of magnitude ~1, and the DEM particle's size is also 1, so surely they look about the same size. The easy fix would be to set the DEM simulation's parameters consistent with the underlying SI system instead (particle size 0.01, G 9.81 etc.).

Thank you,

Ruochun