yade-dev team mailing list archive

[Bruno Chareyre <bruno.chareyre@xxxxxxxxxxx>]

Hello

Is the pseudo code below for developping a new contact law? It is 
probably useless since there is one law doing what you need already : 
the "CohesiveFrictional" class.

There are probably two mistakes in your code :
1/ For sure : what you define as Fs= - Ks*double_ds is in fact a dFs (an 
increment), not the total force.
If you check the current contact laws, you will see something like Fs = 
Fs - Ks*double_ds. With your equations, Fs is in fact a viscous force, 
proportional to the current relative velocity.

2/ Not sure : the fact that rotational velocity is constantly increasing 
may be due to a sign error somewhere (e.g. in the normal orientation), 
or perhaps an inapropriate timestep (too large). I don't feel like the 
first mistake could produce this result.

Bruno


Nice Man a écrit :
> Hi,Dear all.
>  
> I would like to use cohesive contact model to simulate a rock, the 
> small rock particle is modeled by cementing them and the force assumed 
> by the cementing is cohesive force--which can assume normal and shear 
> force.  but I meet problem in moment calculation and the rotational 
> speed calculation:
> just for 2D case ( 'x' is cross(), Normal_directior_unit is from P1 
> center to P2 center ):
> vector_relative_velocity=(Particle1.v+Particle1.omega x 
> (Particle1.r*Normal_directior_unit) ) -(Particle2.v+Particle2.omega x 
> (Particle2.r * -Normal_directior_unit) ) ;
> vector_relative_shear_velocity=(vector_relative_velocity * 
> vector_Shear_Direction) * vector_Shear_Direction;
> double_relative_shear_velocity=(vector_relative_velocity * 
> vector_Shear_Direction)
> double_ds=double_relative_shear_velocity*dt;
>  
> Fs= - Ks*double_ds;
> vector_Fs=Fs* vector_Shear_Direction;
>  
> Now moment is :
> delta_Moment= (Particle1.r * Normal_directior_unit) x vector_Fs; //(Fn 
> is not calculated here,and also since Fn pass through the center of 
> the mass, so it has no effect on the rotation)
>  
> Total_moment += delta_Moment;
>  
> so
> Rotational_velocity += ( Total_moment / Inertia)*dt;
>  
> My problem met is , after several iterations,
> my rotational velocity reachs 1e4 rad/sec, or even more!!! this is 
> totally incorrect.
> So what is the calculation problem in this cohesive condition? ?  or 
> can I do the moment calculation and rotational speed calculation like 
> this since there is cementing cohension between particles and the 
> cohension will strict the free to rotate of the particle now? In the 
> CohesiveFrictionalContactLaw, I see the moment calculation.
>  
> need your help. Thanks in advance.
>  
> Kan
> ------------------------------------------------------------------------
>
> _______________________________________________
> Mailing list: https://launchpad.net/~yade-dev
> Post to     : yade-dev@xxxxxxxxxxxxxxxxxxx
> Unsubscribe : https://launchpad.net/~yade-dev
> More help   : https://help.launchpad.net/ListHelp
>   


--

_______________
Chareyre Bruno
Maitre de conference

Grenoble INP
Laboratoire 3SR - bureau E145
BP 53 - 38041, Grenoble cedex 9 - France
Tél : 33 4 56 52 86 21
Fax : 33 4 76 82 70 43
________________