1. changed dataTransfer path definition from const string to CMake Definition
2. removed the middle barrier->wait() in DragForceGroup.hpp, seems to have saved around 55 sec execution time... for the test_2dshpl.py case

Results:
==============================================================
Before:
No SPAWN: CSubLatticeControler::initMPI()
No SPAWN: slave started at local/global rank 19348608 / 1
Loaded foam case: C:/SVWork/DEMDev/esys-particle-win_dem_cfd/trunk/example/2dshpl...
Loaded C:/OpenFOAM/bin/icoFoamESyS.dll...
Loaded dll data transfer function from
C:/SVWork/DEMDev/esys-particle-win_dem_cfd/trunk/buildvs2008/output/Release/datatransfer.dll...
totalSteps=5000
join completed, start clear...
data_trasfer from DEM side took: 0.024sec
data_trasfer from FVM side took: 0.038sec.
cleared.
N_total= 5000 , N_int= 10 , NDF= 10
mass_p= 0.523598775598
I got particles!!
foamCaseName= 2dshpl
endTime= 0.025 , deltaT= 5e-05 , writeInterval= 1
I got the mesh wall!!
simulation took 1244.220 sec (20.737 min)
----------------------------------------------------------------------------------------------------------------------------
After:
No SPAWN: CSubLatticeControler::initMPI()
No SPAWN: slave started at local/global rank 736384 / 1
Loaded foam case: C:/SVWork/DEMDev/esys-particle-win_dem_cfd/trunk/example/2dshpl...
Loaded C:/OpenFOAM/bin/icoFoamESyS.dll...
Loaded dll data transfer function from
C:/SVWork/DEMDev/esys-particle-win_dem_cfd/trunk/buildvs2008/output/Release/datatransfer.dll...
totalSteps=5000
join completed, start clear...
data_trasfer from DEM side took: 0.014sec
data_trasfer from FVM side took: 0.062sec.
cleared.
N_total= 5000 , N_int= 10 , NDF= 10
mass_p= 0.523598775598
I got particles!!
foamCaseName= 2dshpl
endTime= 0.025 , deltaT= 5e-05 , writeInterval= 1
I got the mesh wall!!
simulation took 1188.452 sec (19.808 min)