def gravity_shift_closed_box():
manager = CVManager(10, 0.5, -0.5)
particles = []
walls = generate_closed_cube_box(1, [0, 0, 0])
def get_gravity(self):
time_fact = 0.25 * self.time * self.time
return [9.81 * math.sin(time_fact), -9.81 * math.cos(time_fact), 0]
for y in [-0.18, -0.07, 0.1, 0.21, 0.32, 0.43]:
for x in np.arange(-0.4, 0.41, 0.2):
for z in np.arange(-0.4, 0.41, 0.2):
pos = np.array(
[x + 0.05 * (rand() - 0.5), y, z + 0.05 * (rand() - 0.5)])
particles.append(
LowMemParticle(len(particles),
pos,
np.array([pos[0], 0, pos[2]]),
diameter=0.1,
get_gravity=get_gravity))
wall_cols = []
for p in particles:
for wall in walls:
wall_cols.append(
AAWallCollision(p,
wall,
restitution=0.8,
friction_coefficient=0.4,
friction_stiffness=5e4))
timestep = 0.0005
last_time = 0
max_time = 30
logger = Logger(particles,
"accel_gravity_shift_closed_box",
"../../run/accel_gravity_shift_closed_box/",
ignore_warnings=True)
logger.log(0)
bar = progressbar.ProgressBar(redirect_stdout=True, max_value=max_time)
for t in np.arange(0, max_time, timestep):
bar.update(t)
manager.add_particles(
particles
) # Kernel to set CV particle lists. (Pass over all particles).
p_cols = manager.get_collisions(
) # Kernel to get collisions from CVs. (Pass over all CVs).
delta_t = t - last_time
for col in p_cols + wall_cols:
col.calculate(
delta_t
) # Kernel to calculate all collisions. (Pass over all collisions).
for p in particles:
p.iterate(
delta_t, implicit=True
) # Kernel to iterate all particles. (Pass over all particles).
last_time = t
manager.reset()
logger.log(t)
bar.finish()
def test_cv_speed(self):
y_sets = [[0.43], [0.32, 0.43], [0.21, 0.32, 0.43],
[0.1, 0.21, 0.32, 0.43]]
ns = []
times = []
for y_set in y_sets:
start = time.time()
manager = CVManager(12, 0.5, -0.5)
particles = []
walls = generate_closed_cube_box(1, [0, 0, 0])
for y in y_set:
for x in np.arange(-0.4, 0.41, 0.2):
for z in np.arange(-0.4, 0.41, 0.2):
pos = np.array([
x + 0.05 * (rand() - 0.5), y,
z + 0.05 * (rand() - 0.5)
])
particles.append(
Particle(len(particles),
pos,
np.array([pos[0], 0, pos[2]]),
diameter=0.1))
ns.append(len(particles))
print(len(particles))
wall_cols = []
for p in particles:
for wall in walls:
wall_cols.append(
AAWallCollision(p,
wall,
restitution=0.8,
friction_coefficient=0.4,
friction_stiffness=5e4))
timestep = 0.0005
last_time = 0
max_time = 2
for t in np.arange(0, max_time, timestep):
manager.add_particles(particles)
cols = manager.get_collisions()
delta_t = t - last_time
for col in cols:
col.calculate(delta_t)
for col in wall_cols:
col.calculate(delta_t)
for p in particles:
p.iterate(delta_t, implicit=True)
last_time = t
manager.reset()
run_time = time.time() - start
times.append(run_time)
print(run_time)
plt.plot(ns, times)
plt.show()
def test_cv_manager(self):
manager = CVManager(10, 0.5, -0.5)
particles = []
walls = generate_closed_cube_box(1, [0, 0, 0])
for y in [0.1]:
for x in np.arange(-0.4, 0.41, 0.2):
for z in np.arange(-0.4, 0.41, 0.2):
pos = np.array([
x + 0.05 * (rand() - 0.5), y, z + 0.05 * (rand() - 0.5)
])
particles.append(
Particle(len(particles),
pos,
np.array([pos[0], 0, pos[2]]),
diameter=0.1))
wall_cols = []
for p in particles:
for wall in walls:
wall_cols.append(
AAWallCollision(p,
wall,
restitution=0.8,
friction_coefficient=0.4,
friction_stiffness=5e4))
timestep = 0.0005
last_time = 0
max_time = 5
bar = progressbar.ProgressBar(redirect_stdout=True, max_value=max_time)
for t in np.arange(0, max_time, timestep):
manager.add_particles(particles)
cols = manager.get_collisions()
bar.update(t)
delta_t = t - last_time
for col in cols:
col.calculate(delta_t)
for col in wall_cols:
col.calculate(delta_t)
for p in particles:
p.iterate(delta_t, implicit=True)
last_time = t
manager.reset()
bar.finish()
particles_to_paraview(particles,
"cv_test",
"../../run/cv_test/",
ignore_warnings=True)