Dear Farima,
you can use Python scripts to do that quickly and precisely. The first thing to do is to write a script that sample N=100 (x,y) coordinates within the radius of the cylinder according to a distribution that you want (random, uniform, etc.). Then you can create trajectories as polylines for each sampled point. Something like that:
import s4l_v1 as s4l
import numpy as np
from s4l_v1 import Vec3
R=10 # Radius (mm)
Z=100 # length of the cylinder (mm)
N=100 # N axons
## Creates a folder in the Model Panel
folder=s4l.model.CreateGroup('Axon Trajectories')
points=[]
cnt=0
while cnt<N:
# Creates Randomly Distributed Points
x=2*R*(-0.5+np.random.rand());
y=2*R*(-0.5+np.random.rand())
if (x*x+y*y<R*R):
axon=s4l.model.CreatePolyLine([Vec3(x,y,-0.5*Z),Vec3(x,y,+0.5*Z)])
axon.Name='Axon_'+str(cnt)
folder.Add(axon)
cnt+=1
