Rename (and correct a bug into) the SFH quenching module to sfh_quenching_trunk
and add a sfh_quenching_smooth version where the SFR is linearly decreasing
after the quenching.