Damping coefficient of force controller in EDEM

Hao Shi

Hi EDEM expert,

I have a question on the force controller in EDEM geometry motion setting. The documentation (https://help.altair.com/EDEM/topics/creator_tree_geometries/specifying_a_force_t.htm?hl=specify,force,controller) on this is not very clear:

"Select a Damping Coefficient value from the Damping Coefficient list. Damping Coefficient is calculated by CmV, where C - Damping coefficient m - Mass of the object and V is the velocity of the object."

I assume this is a viscous damping used to add extra viscous force during the force servo control to suppress the underdamped case/overcorrection in force. How to determine this value? Are there any extra reference/documentation on this?

Thanks.

Regards,

Hao

Eric Veikle_22252

Hao Shi said:

Dear Eric,

Thanks for the explanation. The definitions are clear. From the user's point of view, this simple force controller works perfectly for applying gravitational force to the geometry. If more complex motions/controls need to be implemented, coupling with MotionSolve is indeed a good option.

Just a suggestion: this damping coefficient is by default set to zero (so no damping added and gives underdamping behaviour) and there is no guidance on how to choose a good value, it would be nice to have some examples of typical values used.

Best regards,

Hao

Hi Hao,

The suggestions for providing guidance on a good damping value is going to be very system dependent, so it is tough to provide just one range.  However, for a EDEM PM-Flextire simulation in which I am applying a 10 kN vertical load through the pneumatic tire, a damping coefficient of 100 Hz is reasonable. I would expect that in this case I am over damping the system, but the response that I care about from the simulation (rolling resistance) is not effected by the damping, so overdamping to achieve faster steady state response is desirable to minimize run time.  The damping coefficient is not only a function of your system, but also you need to consider the objective of your simulation.
Kind Regards,

Eric

Eric Veikle_22252

Hello Hoa,

The damping used in the force controller is a simple implementation, exactly as the help documentation states.  The damping force applied through the force controller is equal to the damping coefficient [Hz] multiplied by the mass of the geometry [kg] multiplied by the velocity of the geometry [m/s].

If you would like to implement a more advanced force controller, one option would be to define it through the Coupling Interface - this will require creating a C++ code that would be compiled and run in parallel with the EDEM simulation.  Or, the method that I would recommend, take advantage the the coupling between EDEM and MotionSolve (Altair's MBD solver).  In motion solve you can implement more advanced dampers.  It is really easy to setup the MBD model using Inspire motion, which can then be exported and used to couple with EDEM.

Have a look at the YouTube video that walks through the process of setting up and executing a EDEM MS coupled simulation.

Kind Regards,

Eric

Hao Shi

Eric Veikle_22252 said:

Hello Hoa,

The damping used in the force controller is a simple implementation, exactly as the help documentation states.  The damping force applied through the force controller is equal to the damping coefficient [Hz] multiplied by the mass of the geometry [kg] multiplied by the velocity of the geometry [m/s].

If you would like to implement a more advanced force controller, one option would be to define it through the Coupling Interface - this will require creating a C++ code that would be compiled and run in parallel with the EDEM simulation.  Or, the method that I would recommend, take advantage the the coupling between EDEM and MotionSolve (Altair's MBD solver).  In motion solve you can implement more advanced dampers.  It is really easy to setup the MBD model using Inspire motion, which can then be exported and used to couple with EDEM.

Have a look at the YouTube video that walks through the process of setting up and executing a EDEM MS coupled simulation.

Kind Regards,

Eric

Dear Eric,

Thanks for the explanation. The definitions are clear. From the user's point of view, this simple force controller works perfectly for applying gravitational force to the geometry. If more complex motions/controls need to be implemented, coupling with MotionSolve is indeed a good option.

Just a suggestion: this damping coefficient is by default set to zero (so no damping added and gives underdamping behaviour) and there is no guidance on how to choose a good value, it would be nice to have some examples of typical values used.

Best regards,

Hao

Eric Veikle_22252

Hao Shi said:

Dear Eric,

Thanks for the explanation. The definitions are clear. From the user's point of view, this simple force controller works perfectly for applying gravitational force to the geometry. If more complex motions/controls need to be implemented, coupling with MotionSolve is indeed a good option.

Just a suggestion: this damping coefficient is by default set to zero (so no damping added and gives underdamping behaviour) and there is no guidance on how to choose a good value, it would be nice to have some examples of typical values used.

Best regards,

Hao

Hi Hao,

The suggestions for providing guidance on a good damping value is going to be very system dependent, so it is tough to provide just one range.  However, for a EDEM PM-Flextire simulation in which I am applying a 10 kN vertical load through the pneumatic tire, a damping coefficient of 100 Hz is reasonable. I would expect that in this case I am over damping the system, but the response that I care about from the simulation (rolling resistance) is not effected by the damping, so overdamping to achieve faster steady state response is desirable to minimize run time.  The damping coefficient is not only a function of your system, but also you need to consider the objective of your simulation.
Kind Regards,

Eric

Hao Shi

Eric Veikle_22252 said:

Hi Hao,

The suggestions for providing guidance on a good damping value is going to be very system dependent, so it is tough to provide just one range.  However, for a EDEM PM-Flextire simulation in which I am applying a 10 kN vertical load through the pneumatic tire, a damping coefficient of 100 Hz is reasonable. I would expect that in this case I am over damping the system, but the response that I care about from the simulation (rolling resistance) is not effected by the damping, so overdamping to achieve faster steady state response is desirable to minimize run time.  The damping coefficient is not only a function of your system, but also you need to consider the objective of your simulation.
Kind Regards,

Eric

Hi Eric,

Thanks, your case is indeed a very good example, although there is no single answer on which coefficient value to use, the example you provided offers a good insight into how to choose a value based on the damping category and objective of the simulation outputs.

Best regards,

Hao