shockwave simulation

Dear Lammps users,

I want to simulate a shockwave propagation in metals (such as Cu and
Ta) ...Basically i construct two regions using the command "region"
the first region is named as "piston" and second one is named as

region piston block INF INF INF INF INF 3
region bulk block INF INF INF INF 3 INF

then i create the corresponding groups "group piston region piston"
"group bulk region bulk" then I run the simulation in the (NVE)
ensemble "fix nve" and i use periodical boundady conditions at the
transverse directions of the shock wave, and "shrink-wrapped
" in the wave propagation direction i.e "boundary p p s" . Then i
define the velocity of the piston to an arbritary value i.e "velocity
piston set 0 0 2.766" then i set the forces of the region "pistion"
to zero i.e "fix 5 piston setforce 0.0 0.0 0.0" .
the atomic postions are dumped as .cfg i visualise the atoms using
"atomeye" ....I also use dump_modify to "color the atoms" in the
different groups ...

When i visualise the results I observe the "piston" remains same
postion, and my guess is that it should move , because i have an
initial speed, when i run for many time steps the box change it length
(it stretch), can someone tell me if my model or assumptions are
correct ?

Thank you
Oscar G.