fix all property/global coefficientRollingFriction

Submitted by yyu on Thu, 12/19/2013 - 19:21

Hi everyone,

I have one question about cofficient rolling friction. In my case, I used three different material, called type 1 type 2 type 3. the particles (type 1) alway roll a long time to stop on the plane (type 3). the different rolling frictions from (0.1-10) were tried, But still have the same issue.
the following is the line:

fix m5 all property/global coefficientRollingFriction peratomtypepair 3 0.75 0.1 10.86 0.1 0.65 0.59 10.86 0.59 0.65
pair_style gran/hertz/history rolling_friction cdt #Hertzian without cohesion
pair_coeff

any comments are appreciated, thanks
yyu

richti83's picture

richti83 | Thu, 12/19/2013 - 20:32

The cdt is a type-A rolling friction model, the momentum is not damped so it allways wiggle (with little amplitude)
Try the epsd and epsd2 pairstyle (this are model typeC models), see AI et all, or Wensrich/Katterfeld "Rolling Friction as technique for particle shape" for a very good paper about this topic.

Also make sure you enabled rolling friction not only for the pair forces but also for the walls (see fix wall, you need to enable rollingfriction epsd(2) there too !).

I'm not an associate of DCS GmbH and not a core developer of LIGGGHTS®
but I have 7 years user experience using LIGGGHTS®, ParaView and LINUX

yyu | Fri, 12/20/2013 - 02:50

Hi richti,

thanks for your advisements. I tried the two ways, but still didnot work.
one way (epsd), even the coefficient rolling friction and viscousdamping were changed, nothing happened.
fix m5 all property/global coefficientRollingFriction peratomtypepair 3 0.75 0.1 0.86 0.1 0.65 0.59 0.86 0.59 0.65
fix m6 all property/global coefficientRollingViscousDamping peratomtypepair 3 0.25 0.1 1.26 0.1 0.35 0.39 1.26 0.39 0.65

#New pair style
pair_style gran/hertz/history rolling_friction epsd #Hertzian without cohesion
pair_coeff * *
ANOTHER (epsd2)
fix m1 all property/global youngsModulus peratomtype 2.5e7 5.368e7 1.8e7
fix m2 all property/global poissonsRatio peratomtype 0.25 0.22 0.3
fix m3 all property/global coefficientRestitution peratomtypepair 3 0.3 0.35 0.13 0.35 0.32 0.23 0.13 0.23 0.33
fix m4 all property/global coefficientFriction peratomtypepair 3 0.5 0.43 0.75 0.43 0.43 0.52 0.75 0.52 0.21
fix m5 all property/global coefficientRollingFriction peratomtypepair 3 0.75 0.1 0.86 0.1 0.65 0.59 0.86 0.59 0.65

#New pair style
pair_style gran/hertz/history rolling_friction epsd2 #Hertzian without cohesion
pair_coeff * *

Would tell me more , or I missed some parts?
thanks

richti83's picture

richti83 | Fri, 12/20/2013 - 09:44

I think you did not enabled rolling friction for walls:
check your fix wall command:

fix meshwalls all wall/gran/hertz/history mesh n_meshes 2 meshes cad1 cad2 rolling_friction epsd2


Default:

rolling_friction = 'off' cohesion = 'off' tangential_damping = 'on' absolute_damping = 'off' store_force = 'no'

I'm not an associate of DCS GmbH and not a core developer of LIGGGHTS®
but I have 7 years user experience using LIGGGHTS®, ParaView and LINUX

yyu | Thu, 06/02/2016 - 02:34

Thanks for your help. It works for rolling_friction. How about the CoefficientFriction ? I also tried to test it by different values and got the same answer using the example case (conveyor in the example folder). I looked for a similar way to find the "off" and "on" for CofficientFriction and can't finf from the manual or the document. Do you have any idear? Thanks.
Regards ,
Yaowei

richti83's picture

richti83 | Fri, 06/03/2016 - 08:58

As Conveyor example uses surfaces_velocity i would not expect any changes in Material transport. Do a simple AOR test with a cylinder filled with 7000 particles and move it up after particles settled.

I'm not an associate of DCS GmbH and not a core developer of LIGGGHTS®
but I have 7 years user experience using LIGGGHTS®, ParaView and LINUX