Young's modulus

I have a question regarding to the usage of a realistic Young's modulus in my simulations.
Up to now I have used a value of 5e6 N/mm² for my particles because this value is used in most of the examples. Now I want to implement a realistic value of 2.1e11 N/mm² (Steel). I know that I have to reduce the timestep, but I don't know exactly which correlation I have to consider.
In the documentation of the command "fix check/timestep/gran" there are two equations and with my parameters I get the following values for my dt_max:

dt_r_max = PI*r*sqrt(rho/G)/(0.1631*nu+0.8766)
r = 0.01 m
rho = 7850 kg/m³
nu = 0.3
E = 2.1e11
G = E/(2*(1+nu)) = 8.077e10
dt_r_max = 1.058e-5

Is this correct?

When I execute my simulation with dt = 1.058e-5 there is so much more kinetic energy in there and all of my particles suddenly have a positive z-velocity and fly out of my simulation box (file "bowl_E21e11.png") contrary to my simulation with E=5e6, dt=1e-4. Are there any explanations?

Furthermore it would be great to speed up my simulation. How do I find a good value for the "neighbor" and "neighbor_modify" commands? And where can I read something helpful about improving parallelisation?

In my simulation 20000 particles are set into vibration in a bowl. Here is the code, the STL (file extension changed to .txt in order to prepare the file for uploading it into the forum) of the bowl and a picture of the setup for your imagination.

===========================================================================

### Initialization

units si
dimension 3
newton off
processors 3 3 1
boundary f f f
atom_style granular
atom_modify map array
communicate single vel yes

#Read Particle Start Position
read_data restart/init20000.rstrt

###Settings

neighbor 0.03 bin
neigh_modify delay 0

### Material and interaction properties

fix m1 all property/global youngsModulus peratomtype 2.1e11
fix m2 all property/global poissonsRatio peratomtype 0.3
fix m3 all property/global coefficientRestitution peratomtypepair 1 0.8
fix m4 all property/global coefficientFriction peratomtypepair 1 0.15
fix m5 all property/global characteristicVelocity scalar 2

### physics

pair_style gran/hooke/history
pair_coeff * *
fix grav all gravity 9.81 vector 0 0 -1.0

### walls

fix cad all mesh/surface/stress file geometrie.stl type 1 scale 0.001 move 0 0 0
fix wall all wall/gran/hooke/history mesh n_meshes 1 meshes cad

### numeric

fix integrate all nve/sphere
timestep 1.058e-5

### Output

dump dumpmesh all mesh/vtk 200 post/mesh*.vtk id stress
dump dmp all custom 200 post/dump*.stream id type x y z ix iy iz vx vy vz fx fy fz &
omegax omegay omegaz radius

### Monitoring

thermo_style custom step atoms ke cpu
thermo 1000

### Vibration

# Variables Definiton

variable x_amp equal 0.002
variable alpha_amp equal 0.5
variable frequency equal 60

# derived variables

variable omega equal 2*PI*v_frequency
variable x_dot_amp equal v_x_amp*v_omega
variable alpha_amp_rad equal v_alpha_amp*(PI/180)
variable alpha_dot_amp equal v_alpha_amp_rad*v_omega

variable x_dot equal v_x_dot_amp*cos(v_omega*(step*dt))
variable y_dot equal v_x_dot_amp*sin(v_omega*(step*dt))
variable z_dot equal 0
variable alpha_dot equal v_alpha_dot_amp*sin(v_omega*(step*dt)+PI)
variable beta_dot equal v_alpha_dot_amp*cos(v_omega*(step*dt))

variable startPos_x_dot equal 0
variable startPos_y_dot equal (-1)*v_x_amp
variable startPos_z_dot equal 0
variable startPos_alpha_dot equal v_alpha_amp_rad

### Reach Start Position

fix startPosX all move/mesh mesh cad linear/variable v_startPos_x_dot v_startPos_y_dot v_startPos_z_dot
fix startPosAlpha all move/mesh mesh cad rotate/variable origin 0 0 0 axis 1 0 0 omega v_startPos_alpha_dot

run 10000

unfix startPosAlpha
unfix startPosX

### Start Vibration

fix xVibration all move/mesh mesh cad linear/variable v_x_dot v_y_dot v_z_dot
fix alphaVibration all move/mesh mesh cad rotate/variable origin 0 0 0 axis 1 0 0 omega v_alpha_dot
fix betaVibration all move/mesh mesh cad rotate/variable origin 0 0 0 axis 0 1 0 omega v_beta_dot

run 600000

===========================================================================

Thank you for your help!

Best regards,
Jan