Make the wall smooth

Submitted by tshan on Tue, 11/23/2010 - 07:46

Hi, all

How can I make that there's no friction between walls and particles while there's friction among particles?

Could someone tell me this in detail?

Thanks in advance.

Thanks!

Forums:

Yansan | Tue, 11/23/2010 - 12:03

Hi tshan,

EDIT: I didn't read the last part of your post, that you want to have friction between the particles. So you should use Chris advice and forget mine. I have to read more careful next time.

Regards,
Yansan

tshan | Tue, 11/23/2010 - 12:13

Chris gives me the answer I want. I could not understand the manual relating to this.
Thank Chris.
Thank Yansan all the same.

Best,
Tong

cstoltz | Tue, 11/23/2010 - 11:30

You could just define two different materials in the input deck. Assign one to the particles, and the other to the walls, and set different interaction parameters for particle-particle and particle-wall.

For example, assigning material 1 to the particles and material 2 to the walls:

fix m1 all property/global youngsModulus peratomtype 2.5e8 2.5e8
fix m2 all property/global poissonsRatio peratomtype 0.25 0.25
fix m3 all property/global coefficientRestitution peratomtypepair 2 0.5 0.5 0.5 0.5
fix m4 all property/global coefficientFriction peratomtypepair 2 0.5 0.0 0.0 0.0

Regards,
Chris

tshan | Tue, 11/23/2010 - 15:06

Hi Chris,

If I use SI units, then what is the unit of youngsModulus in the fix property command?

If I use 7e7 to represent E=70GPa, is it mean the unit of pressure will be kPa?

Thanks.

Tong

cstoltz | Tue, 11/23/2010 - 16:35

Hi Tong,

In the SI units, all moduli should be in units of Pascals, so in my earlier post, 2.5e8 is 250 MPa. If you use 7e7, this will only by 70 MPa, not GPa. Similarly, your pressures will be reported in Pascals as well.

Regards,
Chris

tshan | Wed, 11/24/2010 - 12:50

Hi, Chris

I want to set the E=70GPa, but if dong this ,the dt must be extremely small and the simulation will last a long time steps. Is there a method to solve this?

Thanks!

Tong

cstoltz | Wed, 11/24/2010 - 13:01

Tong,

The best suggestion I can offer is to see if you can get away with reducing the modulus by a couple orders of magnitude to see whether or not the behavior of the system will really be impacted. I don't know what kind of problem you're doing, but in many cases, the macroscopic behavior of the system will not strongly depend on the modulus, and you might be able to reduce it significantly. As dt scales inversely with the square root of the modulus, reducing it by 2 orders of magnitude will yield a 10x speedup. Try running with modulus of O(1) and O(10) MPa and see if the macroscopic behavior changes. If so, keep increasing until you reach consistent behavior or the actual limit.

A second alternative is to force the density of the particles larger as this will also allow for a larger timestep. Again, you'll need to check to see if the density change impacts the system behavior.

Regards,
Chris