Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

How can I calculate the coefficient of torsional viscous damping?

  1. Sep 14, 2014 #1
    I have some data from an electric motor found here:

    http://www.engelantriebe.de/pdf/DAT_HLR26_11-13_engl.pdf [Broken]

    However, it does not include information about the coefficient of torsional viscous damping which I need for a Simulink model.

    The units for torsional viscous damping are Nm.s/rad, and the following relevant information about the motor is given:

    Friction torque: 0.06Nm
    Mechanical time constant: 1.7ms
    Nominal speed: 4500rpm (471.24 rad/s)

    1) Can I just divide friction torque by nominal speed to get the coefficient of torsional viscous damping? Which would be 1.27 x 10^-4 Nm.s/rad (the same units)?

    2) I'm also wondering how torsional viscous damping from a load would be reflected back to the motor shaft through a reduction gearbox? In the sense that if we were concerned with the moment of inertia of a load J, we would divide J by the gear ratio squared. Would torsional viscous damping be divided by gear ratio squared or just the gear ratio?

    3) If the load is being reflected through a reduction gearbox which now has a lower speed than the motor speed, say 500rpm, would I use 500rpm for calculating the torsional viscous damping or the earlier 4500rpm in case 1?

    Last edited by a moderator: May 6, 2017
  2. jcsd
  3. Sep 15, 2014 #2


    User Avatar
    Staff Emeritus
    Science Advisor
    Homework Helper

    It's not clear from your post why you suspect this tiny electric motor of having torsional vibration problems.

    AFAIK, viscous dampers are usually put on internal combustion engines, where the torque pulses are not even, so that the torsional vibrations will not cause failure in the crankshafts.


    It seems like an electric motor would not have this problem.
  4. Sep 15, 2014 #3
    Hi SteamKing,

    I am not saying it is a problem, nor am I looking to put a damper on it. But for the sake of modelling it precisely it has a torsional viscous coefficient parameter. For the most part, it is very very small, usually 10^-4 or even smaller.

    In any case, I need to put the parameter in even if the motor itself has a small torsional viscous damping coefficient because the parameter required by the model requires the load itself! It will be reduced by the gearbox, but still it will have a bigger effect than the motor's own torsional viscous characteristic.
  5. Sep 15, 2014 #4


    User Avatar
    Staff Emeritus
    Science Advisor
    Homework Helper

    Are you saying that zero is not an acceptable value? Why don't you experiment and see if small values for the damping, say 10^-4, 10^-5, and 10^-6, make any significant change in whatever output you are looking for?
  6. Sep 15, 2014 #5


    User Avatar
    Science Advisor
    Homework Helper

    Are you talking about a simulink model something like this?

    It looks to me like the "damping" is mainly a fudge to make their time integration work. In the complete mechanical system the driveshaft is probably connecting two rotating objects with significant moments of inertia. You can estimate the torsional vibration frequency as the two rotating masses joined by a massless torsion spring. I would just set this damping parameter to give say 0.001 x critical damping for that vibration mode. That should stop the model "blowing up". If the output has spurious looking oscillations, increase the damping and try again.

    In real life the "damping" will be from sources like windage (air resistance) of the rotating components etc, not just the friction in the bearings and the internal damping of the materials.
  7. Sep 29, 2014 #6


    User Avatar
    Gold Member

    Electric motors definitely have torsional vibrations, particularly during start-up. Check a power systems book like Krause for this.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook