Relaxation parameters for adjusting unknowns in iterations. This can stabilize the calculation. By example, a relaxation parameter of 0.1 means that the corresponding unknown is now completely updated with the iterative change, but only 10 percent of the change is actually applied in a iteration.
If enough iterations are used, the relaxation parameters with not influence the final solution.
You should specify a relaxation parameter term for each principal unknown which is present in the calculation (see the start of the data part description for a list of principal unknowns; these are velocities, temperature, etc.).
The relaxation is used for all timesteps. See also control_options_relaxation.