Different simulation results for restarting VIV problem using z-dependent ooutput in VCSMapping solver
I have been running a 3-D VIV problem of a rigid body using VCSMapping solver and found out that if the z-dependent motion information (in *.mot), the mapping field *.map and instant field .chk/.fld are specified as initial conditions for restart, the force in the motion direction shows large fluctuation at the initial few steps. These large initial fluctuations cause inconsistent evolutions for all signals compared with the original ones (see the green line in figure 1 as attached).
After changing the motion parameters as the spanwise-averaged values, the fluctuation on F_y is getting smaller and the results largely follow the original trend (see “2nd run” in figure 1).
The above is possibly due to the mismatch between the cable motion (z-dependent while using *.mot) and the coordinate mapping (independent with z) during the initialisation in ForcingMovingBoday.cpp.
Fluctuating_in_forces_during_restarting_simulation_in_VCSMapping_solver.pdf