[mimi0201]

how to control particle number? i tracked whole lot particles (over 1000 particles) and the file just crashed because of the large particle information.

[`e`]

Are you injecting 1,000 particles per time step? If you run a simulation for 1,000 time steps then that would yield one million particles which require tracking (assuming none have escaped). Desktop machines less than 3-5 years old should handle these particles OK, but for older machines or many more particles, you may require a cluster/HPC.

[mimi0201]

Originally Posted by `e`

Are you injecting 1,000 particles per time step? If you run a simulation for 1,000 time steps then that would yield one million particles which require tracking (assuming none have escaped). Desktop machines less than 3-5 years old should handle these particles OK, but for older machines or many more particles, you may require a cluster/HPC.

well, i have no idea how to control particle number, just simply click "track" and see how many particles have been generated. Do you know if there is any way to control particle numbers, like i just want 50 particles? my simulation is steady state so does it relate to the time step?

[`e`]

If you're simulating steady state then there won't be a sequence of particles injected, only one set of particles for each injection. You can modify the maximum number of particles simulated by modifying the injections (Define > Injections). You can scale down this number of injected particles from the results section (Display > Graphics and Animations... > Particle Tracks) by specifying a nonzero integer for "Skip".

[hwet]

I think you are looking at making a group injection in which you can specify the number of particles.

[mimi0201]

Originally Posted by `e`

Ok, let's work through this reflection code with an example.

Suppose our particle exists in 2-D and has a velocity vector p = (3,-2) m/s. The wall is parallel to the x-axis and therefore has a normal unit vector of n = (0,1). Units are absent to keep simplicity in the following text.

/* Compute normal velocity. */
for(i=0; i<idim; i++)
vn += P_VEL(p)[i]*normal[i];

vn = p dot n = (3,-2) dot (0,1) = -2 (the same speed our particle was travelling towards the wall, "normal component")

/* Subtract off normal velocity. */
for(i=0; i<idim; i++)
P_VEL(p)[i] -= vn*normal[i];

p = p - vn*n = (3,-2) - (-2)*(0,1) = (3,0) (now the normal component for our particle is zero)

/* Apply tangential coefficient of restitution. */
for(i=0; i<idim; i++)
P_VEL(p)[i] *= tan_coeff;

Suppose the tangential coefficient is 0.5:
p = p*tangential_coefficient = (3,0)*0.5 = (1.5,0) (the tangential speed is reduced but the direction remains the same)

/* Add reflected normal velocity. */
for(i=0; i<idim; i++)
P_VEL(p)[i] -= nor_coeff*vn*normal[i];

Suppose the normal coefficient is 0.5:
p = p - normal_coefficient*vn*n = (1.5,0) - 0.5*(-2)*(0,1) = (1.5,1) (the normal speed is reduced and changes direction)

/* Store new velocity in P_VEL0 of particle */
for(i=0; i<idim; i++)
P_VEL0(p)[i] = P_VEL(p)[i];

Some DPM solvers may use the velocity of the particle as it enters the current cell for calculating the next velocity or position.

As we expect, our reflected velocity, (1.5,1) m/s, has been reduced and directed away from the wall. This code works.

A stupid question, just wondering that why the first part uses the dot product for P_VEL(p)[i]*normal[i], but the rest just use normal multiply??

/* Compute normal velocity. */
for(i=0; i<idim; i++)
vn += P_VEL(p)[i]*normal[i];

/* Subtract off normal velocity. */
for(i=0; i<idim; i++)
P_VEL(p)[i] -= vn*normal[i];

[pakk]

Originally Posted by mimi0201

A stupid question, just wondering that why the first part uses the dot product for P_VEL(p)[i]*normal[i], but the rest just use normal multiply??

/* Compute normal velocity. */
for(i=0; i<idim; i++)
vn += P_VEL(p)[i]*normal[i];

/* Subtract off normal velocity. */
for(i=0; i<idim; i++)
P_VEL(p)[i] -= vn*normal[i];

Because in the first part, a velocity component is calculated (vn), and in the other parts, velocity vectors are calculated (P_VEL(p)).

[mimi0201]

Originally Posted by pakk

Because in the first part, a velocity component is calculated (vn), and in the other parts, velocity vectors are calculated (P_VEL(p)).

Oh ok got it. Cheers!!! One more question is that which statement indicates whether the particle hits the wall and then reflects?

[`e`]

Originally Posted by mimi0201

One more question is that which statement indicates whether the particle hits the wall and then reflects?

The fate of the particle is dependent on if the velocity is above some critical capture velocity and is implemented with the conditional statement ("if(vmag > capture_vel)") where the particle trajectory is either continued:

return PATH_ACTIVE

or the particle trajectory is terminated:

return PATH_ABORT