OpenMp for dummies

I wrote this simple C code where I initialize a set of variables which are then copied them into a structure; this structure (which changes size every cycle of a for loop) is then passed to a function which cycle on its elements; my goal is to parallelize (using OpenMp) the code and I tried by adding the omp.h header and the this line: #pragma omp parallel for private(i)

#include <math.h>
#include <stdio.h>
#include <stdlib.h>
#include <time.h>
#include <memory.h>
#include <string.h>
#include <omp.h>

typedef struct variables{
   float Vx;
   float Vy;
   float Vz;
}velocity;

velocity *pv;

void func(velocity *pv,int id){
    float V[3];

    V[0]=1*pv[id].Vx;
    V[1]=2*pv[id].Vy;
    V[2]=3*pv[id].Vz;

    pv[id].Vx=V[0];
    pv[id].Vy=V[1];
    pv[id].Vz=V[2];

}

int main(){
    int j,jcoll,i,Np = 100, Nmax=2;
    float Vx[Np],Vy[Np],Vz[Np];
    srand(time(NULL));
    float r; int q;

   for(i=0;i<Np;i++){
      r = rand()%10;
      Vx[i]=1*r;
      r = rand()%10;
      Vy[i]=2*r;
      r = rand()%10;
      Vz[i]=3*r;
   }

   for(jcoll=1;jcoll<=Nmax;jcoll++){
      pv = (velocity*)malloc(jcoll*sizeof(velocity));

      for(j=0;j<jcoll;j++){
         pv[j].Vx = Vx[j];
         pv[j].Vy = Vy[j];
         pv[j].Vz = Vz[j];
      } 

      #pragma omp parallel for private(i) // Is this correct?
      for(i=0;i<jcoll;i++){
         func(pv,i);
      }


      free(pv);   
    }

   return 0;
}

Can anyone check if what I wrote makes sense or should I try to do something else in order to do a correct parallelization?